MBA Marketing Strategy & Analytics Masterclass

1. Course introduction: So MBA in 2023 is expensive and there are very few people who can afford to do that. And that's why I've made this course on MBA marketing and business strategy, where I will help you understand all the concepts that I've learned during my MBA. If you watch all the videos and you will complete all the assignments that I'll be giving it to you. Then I'm sure you will have a very strong business foundation by the end of this course. I have divided this course into five different section. In section number one, we will build a strong business foundation. And in this section we will understand about some basic terminology and concepts of marketing. So we will start by understanding the definition of marketing. Then we will understand about marketing mix. Some people also call this as the four P's of marketing and seven Ps of marketing. And we'll be using a couple of examples to understand both of these concept. Then we will understand about STP analysis. Stp is also known as your segmentation, targeting, and positioning. Then we will talk about swot analysis, also known as strength, weakness, opportunity, and threat. And in the end, we will understand about product life cycle. In section number two, we will talk about business expansion and growth strategy. If you work for a company that have dozen of product and those guys are selling those products into multiple market than this section will be super-important for you. Because as a business executive, you need to formulate deed strategy by using Ansoff matrix, BCG matrix, so that you can communicate some blue ocean strategy to your team members. In section number three, we will talk about branding strategy. And in this section we will understand how companies will sell you the exact same quality of product at an expensive price using these planning strategy. In this section, I'm going to take couple of examples of companies like Apple, Nike, and Adidas so that you can understand about concepts like brand slogan, brand mantra, and BAB Morgan. Then apart from branding, in the last two sections, we will talk about business model and business strategy. And in those section, we will understand about the unique business model of all these unicorns, startup and Fortune 500 companies. So I have a lot to cover in this course. And if you are ready for that, let's dive in. 2. What is Marketing ?: So let's start this course by understanding what exactly is marketing. Marketing is the understanding that we humans are biased and by using our biases, selling us what companies wanted to sell. Now I know this definition may sound confusing to some people. So let me try to oversimplify this by giving you a small example. Let's say if I show you these two different images of two different product and this product have all of these following details. Let's say product number one and product number two, both have the exact same rating of 4.6. Both of these products have almost the same review. That is 2000s, and they are both selling at the same price. So my question is, which of these two product will you purchase? Now, you can pause this video for a while and think about the logical reason behind your answer. Now I know many of you might be thinking, well, product one and product two have almost the same rating. They have somewhat the same amount of review. And both of them are being sold at a price of $5. So I can pick any one of them. I mean, it's not really a big deal. But still majority of you will feel that the product one is having a good design. And that's why I will go with product number one. Because for majority of people, I guess product number one looks good. And this is where you did a mistake because you should have chosen the second product. And you may ask why. Now if you closely look at product number two, despite of having a normal packaging, this product is still able to get the exact same rating as product number one. And both of them have the exact same price of $5. So there might be something good in product number two. That's why a lot of people are purchasing this product. That's why in the marketing definition, I've told you that we humans are biased and these businesses are using our biases to sell us what they exactly wanted to sell. In short, a product that doesn't market well must be a better product if that product sells equally well, which is product number two in this case. Now product packaging is one of the part of marketing. There are multiple dimensions to it. And that's what we're going to study in this specific course. So I'll give you a single line definition of marketing. Then. Marketing is a way of creating, communicating, and delivering value to your customer better than your competitor. In this definition, you have three important thing to understand. The first one is about creating ten, communicating, and then delivering values. Now the way you will create value is by solving the problem of your customer. And that's why you will see in multiple marketing campaign, accompany will always try to highlight benefit of their product because they have created something that is solving a small problem. So the overall product, the packaging, the product have super-important in the value creation. Then you have the communication. So in your specific country, there might be people speaking different language. So how do you exactly communicate with those people in their specific language? And third one is delivery. Because in today's world, you have social media. You have newspaper, and some other media. So how exactly will you deliver the exact same value using these different medium? 3. 4P's of Marketing in Marketing Mix: Hey everyone, In this video, we will understand about marketing mix. The marketing mix is also known as four P's of marketing for product and seven Ps of marketing for service. But let me oversimplify this concept for you. So imagine you are making a pudding. To make a delicious pudding, you need sugar, flour, milk, and water. And then you have to mix all of these four product in the right quantity. And then you have to bake this specific product for right amount of time so that you can make a delicious pudding. Now, if I will give you a funny example, you can look at this specific image. So if a company wanted to have good marketing strategy, they need to ensure that they have the right kind of product that they're selling at the right price, at the right place. And they are also giving some form of promotions as well. So that's what a marketing mix looks like. You need to have a good strategy related to product, price, place, and promotion. I'll give you a definition. Marketing mix describes the different kinds of choices organization must make in the whole process of bringing a product or a service to the market. Now for the marketing mix of a product, you need to focus on product, price, place, and promotion. So you have to sell the right kind of product at the right price on the right place by giving some promotion. And that's what a marketing mix of product looks like. Let's start with product. I'll ask you a simple question. Why do you prefer Starbucks coffee? Well, your simple answer would be, the product quality is really good. And if I lost you, what do you mean by product quality? You may say that I can find different variety of coffee in Starbucks. The quality of the product is really nice. They are having these amazing design inside the store or inside the coffee shop. And they also have to use high-quality cup. And it can also highlight couple of more feature, the brand image starbucks have. They also have nice packaging and they have all of these different sizes of cup in their coffee shop. And you can give so many reasons related to product. So if I'll talk about Starbucks, they have a really nice product. And that's the first marketing mix. That is the product. That is the first ingredient of marketing the product. Let's talk about, please. Now, in order to provide a good customer experience, Starbucks will ensure that they are present across all the different channels. So in your country, you might be using a food tech platform where you can order a specific coffee online. And you might be just walking down the street to drink a coffee. So they are present on all the different types of channels. So whether you are purchasing a coffee online or if you're drinking a coffee by walking into a coffee shop, they are present everywhere. The second one is coverage. Starbucks is present in all major countries and in almost every single city. And that's the kind of coverage those guys have. Then we have location. So you may find a Starbucks coffee shop at the airport, near a railway station, or inside the shopping mall. They are presented almost all the places where you may need a coffee. The third ingredient of marketing is price. So if you look at Starbucks, they have a lot going on in the pricing part. They will list the product at a specific price. Then you can put some discount coupon and you may feel happy that I got the coffee at an amazing discount. And they have all of these different payment method. They provide payment flexibility to customer. You can be using your credit card or debit card. You can also be used in cash. So they have all of these pricing strategy where you can bundle different product together and you will get some discount. Or you can buy three or four coffee together and you will get some discount. So they have all of these different pricing strategy where you can bundle all of these different products. You can get their loyalty card and you can do a lot more things related to the pricing of the product. And that's why the fourth ingredient of this marketing mix is promotion. Starbucks do a lot more advertisement. They really promote the product where celebrities are drinking their coffee. And they tried to maintain their brand image lot on the social media platform where these celebrities and influencers are drinking the coffee of Starbucks, starbucks user, interesting social media growth strategy. So if you go to Starbucks, they will normally ask your name so that they can write that specific name on your coffee cup. So even if you spell your name right, they will still put the wrong spelling of your name on your cup of coffee. Because in that case, you will take a picture and you will post that picture on social media saying that that these guys can't even write the right spelling of my name. 4. 7P's of Marketing in Marketing Mix: So now that you understand all the four important ingredient of marketing mix, that is your product, price, place, and promotion. Let's talk about three other ingredient that you need to understand in case of a service. At the starting of this video, we had a discussion about pudding. If you wanted to make a delicious putting. In that case, you have to mix the sugar floor, milk and water in the right quantity. And then you have to bake this product for the right amount of time so that the cake is delicious. And that's what a marketing mix look like. If you want it to have a good marketing strategy of a product. In that case, you have to make sure that then you have to have a good mix of product, price, place, and promotion. So you need to have a good product and you are selling that product at the right price, at the right place by giving some discount or offer. That's what a good marketing mix of a product looks like. But imagine you're going on a date to enjoy this putting. In that case, you are availing or service. When it comes to service, you need three extra ingredient. You need people, unit processes, and you need physical evidence. And that is the marketing mix Forest Service. So if you look at pudding as a product, then you need to focus on the ingredient of this specific product, that is product, price, place, and promotion. But the time you started enjoying a service, let's say you are going out on a date to enjoy putting in a restaurant. Or let's say you're going with one of your colleague to enjoy a cup of coffee. In that case, you are enjoying a product and a service. When it comes to service than the company needs to focus on the people, the processes, and the physical evidence. If you look at Starbucks, the people in the Starbucks coffee shop are super nice, normally greet their customer. And they always try to behave in a certain way. So Starbucks will always focus on their people. They will give them a lot of training so that they can deliver the best customer experience. Apart from people, Starbucks and other restaurant also have a certain set of processes. In technical term, we call these processes as SOP, or standard operating procedure. You train your people to perform a specific process in a specific physical place. This can be a restaurant or a coffee shop, or a hotel. In case of a service, you have three extra ingredient, people, processes and physical evidence. So let's talk about three extra ingredient that we have in case of a marketing mix of service. That is your people processes and physical evidence. Let's start with people in Starbucks. They will ensure that all of the people who are working in the coffee shop are professional. They are spilled and they have a positive attitude. Then we have process. So they will normally try to process your order faster. They will also ask you for any short of customization If you need in your coffee. And they will also take feedback so that they can improve the process. The third one is physical evidence. If you go to a Starbucks or any coffee shop or any good fast-food restaurant, you may find a good modern furniture. All of the people might be wearing some uniform. They also have the sign board where you can read the price of the product. They have these non-smoking sign. There are couple of other things that these restaurant always do. So in case of a product, the four important ingredient is product, price, place, and promotion. When we are talking about a Service, in that case, the marketing mix of a service will have three extra ingredient. That is people, processes, and physical evidence. So if you are trying to understand about the marketing mix or the marketing strategy of a product, you always need to focus on these four important parameters. That is product, price, place, and promotion. But anytime you have a service, let's say a food delivery service, or a restaurant, or a salon, or something that is related to service. In that case, you need to bring these three additional ingredient. That is people, processes and physical evidence. And that's all about marketing mix of a product and a service. Let me give you a small assignment. And the main purpose of this course is to give you assignments so that you can solve these assignment and you can test out your knowledge. Now, I want you to download the PDF file of this specific assignment and you can print out this PDF file, or let's see, you can draw out this specific assignment on a blank sheet of paper. And you can solve the specific problem. You need to create a marketing mix of a service. That's a maybe you can create a marketing mix for Apple Store. And you can write about the product, the price, the place, the promotion, the people processes and physical evidence. 5. Marketing Management: So now that you understand about marketing mix in this video, let's talk about marketing management. Because if you work in a company, then there is a very high chance that you'll be managing couple of people and product. So what exactly is marketing management? Marketing management is the art and the science of choosing a target market and then acquiring new customer. Making sure that you are retaining those customer and growing the customer B's using these creative deliveries and communication messages so that you can provide superior value to the customer. Now, if you closely observe this definition, you will realize that you have to do a lot of work. So you first have to acquire these new customer, then you have to retain these new customer. And then you have to make sure that you are constantly growing at a specific rate. And for all of these process to happen, you have to create new marketing campaign. And then you have to deliver those campaign to the end-user. And that's what a marketing management looks like. You have to do a lot of work. Now, if you go to big companies, you may realize that they have a single marketing manager for all of these roles. There might be one marketing manager who is constantly looking after the performance marketing effort. There might be one person who is looking at the retention and the growth rate of a company. These role might be assigned to different people based on the size of company. But broadly, in marketing management, you are constantly thinking about how can you introduce new products into the market by picking up a smaller segment? And how can you expand your business by doing both of these activities. Finally, you will convince customers to buy your product with the help of unique delivery. So you have four major function of marketing management. Introducing new product, finding an interesting segment that is looking for a solution and expanding your customer base. And you will do all of these three process by delivering your communication with the help of some unique messaging or unique delivery. And we'll talk about all of these things in a minute. So let's start with new product. So for simplicity, in this video, I'm going to take an example of a company like Colgate. So as you know, in marketing mix, we had a discussion about product, price, place, and promotion. So when you're introducing a new product into the market, you can look at all of these four different ingredient of a product. So you can introduce a new product based on the quality of the existing product. You can introduce a new product based on the price that you are targeting and also based on the police. In different countries, you can introduce different product. Let's start with price. So Colgate introduce a most affordable product and that's the cheapest one. In some developing countries, especially some southeastern countries and African countries. Then they introduce a product for middle-class people. Want a little better quality of toothpaste. And then we have a product for upper-class people. They have Colgate visible white. If you are someone who can spend some extra bucks to use a high-quality toothpaste. In that case, you can buy Colgate visible white. So they have these three different product for three different class of people. And the reason they did that is because different people have different disposable income. I mean, if a person is earning less amount of money every single day, then he may not be interested in spending a lot in choosing a toothpaste. I mean, he just won the most affordable one. And same goes with the people who are in the middle-class and upper-class. So they have these three different product for three different class of people. And that's introducing a new product based on the price. The next one is introducing a product based on a specific segment. Now Colgate is a big company and they have a lot of investor and they are also a listed company. And that's why their aim is to make sure that they are growing at a specific rate every single year so that they can increase the stakeholder's value. And that's why they are constantly introducing a new product. By solving a pinpoint of a specific segment. Let's look at all the different segment you have in the market. The first segment is for kid. And I'm sure you have seen these different shapes and sizes of these toothbrush and toothpaste. I mean, the main aim here is to make these toothbrush or toothpaste or a little more interesting by using these different cartoons or that kids will enjoy brushing their teeth. So that's their first customer segment. I mean, they have a lot more stock keeping unit or SKU or product in this specific segment. The next one is for you. And we have seen all those three different products for different class of people. But they have a Colgate Max Fresh product, especially for youth. So if you are someone from the age group of 18 years to 25 or 30 years. In that case, you can go with Colgate Max Fresh. In the end. They have a product for adult people who have sensitive tooth as well. So they have a Colgate sensitive product for them. And not only these three different segment, I mean, they have product for almost every single segment that you can imagine. Now they did some research and they figured out that they need a product for every single segment. Because normally in a household you just have one toothpaste and every single family member is using just that toothpaste. And that's why they have introduced all of these different types of product so that people will purchase these products and they will use different toothpaste. And that's the whole purpose of maximizing the revenue or stakeholder's value by introducing a product based on the price or affordability, Then we have market expansion. So this is purely based on the type of country you are in. So Colgate have different types of product in different countries. So if you look into some African countries, they have this basic Colgate's strong teeth paste and that's the most affordable one in African country. They are not really launching these premium expensive toothpaste. They have this Colgate strong tea toothpaste. In some Asian countries like India or Indonesia, they have this toothpaste that is named Colgate with Shakti. This specific toothpaste have a lot more Ayurvedic ingredient because people are more inclined towards these Ayurvedic ingredient. And that's why in some Western countries they have Colgate Max ways. So you can see that in three different continent or countries, they have three different product. The reason Colgate launched Colgate beads at t is because one of the Indian brand was expanding very fast in Ayurvedic toothpaste category. And that's why they launched Colgate beads up thin India. So now that you understand how exactly a brand can launch a different product based on price, based on the customer segment. Let's understand how exactly that brand can deliver a unique value with the help of their advertisement. Let's start with the first campaign that is Colgate MOM. Now if you look at a normal household, you will realize that majority of the shopping or products are purchased by women, especially our mom. And those products are used by almost every single family member. That means the customer of the product is different from the consumer. You and me are the consumer of the product, but the customer of the product is our mom. And that's why Colgate obsessively focused on targeting these small. And that's why you may have seen these emotional ad from these marketing campaign like Colgate mom, where these different humans are describing how their kids teeth are stronger just because they are using colgate. And that is one of their marketing campaign. And that is a unique way to deliver value. Because if you are able to convince The woman that this product is good for all the kids or all the family members you have, then they'll probably purchase it. The second campaign they did was on youth. So they showed the benefit and the future of the product. Like you will avoid bad breath if you use the product, a lot of goals would be attracted towards you and all of those things. In the end, they had this unique delivery for their premium product that is, Colgate visible white. And they use these different actresses or these different influencer in order to show them that your teeth will look sparkling white if you use the premium Colgate visible white product. They use these different marketing campaigns to make sure that they have a unique delivery and they are delivering the value to the customer much better than the competitor. So that's all about the marketing management. Or if I summarize the video, marketing management is the art and science of choosing a specific market and then constantly acquiring new and new customer by making sure that you have good retention. And you're constantly growing in that specific market. And the way you do that is by communicating a superior value to your customer. So that's all for this video. I have a small assignment for you, and I highly recommend you to complete this assignment because that's how you're able to test your knowledge, whether you are able to grab this concept or not. So you have to do a marketing management assignment for a brand like Coca-Cola. So you have to write about how does Coca-Cola communicates superior value to the customer. You have to list out all of their product based on different customer segment, based on different price, and maybe based on different continent. 6. Intro to STP (Segmentation, Targeting, and Positioning): Hey everyone, My name is now beep. And in this video, we're going to talk about STP analysis. Now, there is a famous scene in the business world that if you try to be everything to everyone, you will become nothing to anyone. And that's why in this video, we will understand about the S-T-P process. Now, STP stands for segmentation, targeting, and positioning. So if you're launching a brand or a product into the market, in that case, instead of targeting all the people in that specific market, you have to pick a very small segment of people. And then you have to target them across different social media or newspaper or different medium. And then you have to position your product. And that is the overview of your S-T-P process, also known as segmentation, targeting, and positioning. So let's understand why do we need STP? And I'm gonna give you a small example so that everyone can understand. So let's say you are an entrepreneur and you strike up with this idea of creating toothbrush out of bamboo shoots. So instead of picking a very small segment that is environment friendly, Let's say you're planning to target all the toothbrush user by positioning your product is the environment friendly alternative to these plastic toothbrush? But the reality is a normal customer don't really care about sustainable living. That's why you have to do an STP analysis where you have to segment a specific class of customer and then you have to target them. Because in reality, 85% of your audience doesn't really care about sustainable living. No matter how good or bad your product or your idea is, you have to segment your market into all of these categories and then you have to target them. And that's the whole idea of doing STP analysis. You first have to segment your customer. Then you have to find unique ways by which you can target them across different media. And then you have to position your product over the period of time. So if I'll give you a high-level overview of S-T-P process. You first have to segment your customer. Then you have to target them by highlighting the benefit of your product. How will your product is different from all the existing product into the market? After targeting, you have to position this specific product in the mind of people. And obviously this process can take up two ears so that people can understand that this is a good alternative to these plastic tube pressed. And I think I should give it a shot. That's why positioning will take a lot more time so that you can convince people that this is a good alternative to the existing product into the market. And over the period of time, you also have to work on the implementation side. That means you have to work on pricing, product, place, and promotion. And I hope you already have a good understanding about marketing mix. So that's the high level overview of STP analysis. Now, let's take this framework of STP analysis and let's try to implement this SDP into the idea we had. You had an idea of launching a bamboo toothbrush, which is an alternative to the plastic toothbrush. So you first have to segment the customer. Then you have to find unique ways so that you can reach to those specific customer. And then you have to position your product. In a market. You might be able to find almost 15% people that are really interested in trying out products that are eco-friendly. But the main problem is how exactly do you reach out to them? So you have to find different channels and different community where all the people who are doing some environment friendly work are connected to each other. Then you have to educate them about the adverse effect of plastic. And that's how you will be able to position your product in the long run. Now, obviously, the whole process can take up two ears or sometime decades to convince people that this product is a good alternative to the plastic product that you have. And that's why some people will always try to find these so that they can target top 1% people. Instead of targeting all these 15% people that are interested in environmentally-friendly product. So if I summarize the complete video, than STP is a marketing approach where you segment your audience and then you try to target the best-fit audience segment for your product. And in the end, you position your product to capture your target segment effectively. This is all about the basic outline of S-T-P process. In the next video, we will understand how can you pick these different customer segment and how can you find different ways by which you can target those segment? So I'll see you guys in the next video. 7. STP Analysis: Hey everyone, My name is now deep. And in this video we will understand about segmentation, targeting, and positioning. So in the last video, we had a discussion about the outline of STP analysis. And in that video we had a discussion that what exactly is STP analysis and why do we need to do it? In this video, we will pick a small problem and then I will help you understand how can you exactly do some experiment in order to target different customer segment? Now I know I might be going a little slow in this course. But the main purpose of this course is to make sure that all of you have a strong foundation. So in this video, let's pick a small day-to-day problem and then we will understand how you can experiment with different sales channels. And in the next video, we will take our industry specific problem. Let's say you wanted to sell toys in the market. In that case, you have to find different unique ways by which you can sell these toys into the market. Now the false assumption you took was that these kids have these elder brother and those people might be going to these different universities. And that's why I have to visit these universities so that I can sell these products with their elder brother. The second assumption was that all of the appearance might be visiting these corporate officers. And that's why I should stand in front of a university or a corporate office. And that's how I'm able to sell all of these toys. So basically, in this case, we are targeting a wrong segment because in case of toys or customer and the consumer are two different people. A customer is someone who is buying these products, and consumer is someone who is playing with these product. Now customer can be the elder brother or appearance of the kid, but the customer is the good itself. Just picking the customer is not sufficient in this case, you realize that I'm only targeting costumer, but not the consumer. And that's why you changed your sales channel. Now you're selling in front of a school, or let's say you somehow got a permission to sell all of these toys inside the school. But still you are not able to sell these toys well, because you're targeting is not good. So you went back home and you brainstorm this problem with one of your friend. And your friend told you that you have to target the customer and the consumer at the same time. And that's why you decided to change your sales channel. And this time you are standing outside of school, especially when the students are coming out with their parent. And you're standing in front of a pediatric hospital. And in this case you have the right segment and you're doing the right targeting. Because when you stand outside a pediatric hospital, in that case, you have your customer and your consumer. And the consumer who are these kids have the emotional influence on these customer so that they can purchase these toys. Because normally these kids started crying and that's why these parents have to buy these toys so that they can make their kids happy. And that is your right segment and write targeting. You are able to get a place where you have both your customer and consumer. And these consumer have some influence on these customer so that they can purchase the product. So this is the analogy that I have used in order to explain how can you pick your sales channel and how can you target your customer or let's say consumer. So let's understand about the definition of STP analysis. So STP is a marketing model that redefines whom you market your product too. And how. Some people also call this as a step formula. That means if you are able to segment your customer really well, and if you are trying your best in order to target those customers at all the possible seams channel or marketing channel. In that case, you will end up with a good positioning of a product into the market. So in short, this step formula will make your marketing communication more focused, relevant, and personalized to your customer. So let's talk about the objective of STP analysis. And then I will give you these different example and framework so that you can segment your customer and then you can target and position your product. Let's talk about the objective of STP analysis. The first objective is that STP analysis will help companies identify attractive market segments. Now once you have a single or a group of market segment, in that case, STB analysis will also help you choose a target marketing strategy. The last objective of STP analysis is that it will help company's position their product for maximum competitive advantage. And we'll talk about all of these objective in the next video. 8. Segmentation in STP Analysis: So in the last two videos, we had a discussion about the outline and the objective of STP analysis. And this video will go deep into segmentation. And in this video we will understand how can you segment the market based on the specific attributes. Now, I understand that I'm going a little slow in this specific topic. And the reason is that SDP analysis is the core concept of marketing. And that's why you have to have a strong foundation. So segmentation is the process of segmenting the audience into smaller group based on specific attributes. This segmentation will give you a better clarity on who benefits the most from your product and how, if you want to split a big market into these smaller group, in that case, you can do a geographic segmentation or a demographic segmentation. Or maybe you can go ahead with psychographic or behavioral segmentation. Let's start with geographic segmentation. Now to understand geographic segmentation, I'm going to take an example of a brand like McDonald's. So if you know someone who work at McDonald's, they have a famous tagline that you need to think global and act local. And that's why if you closely observe a single product in McDonald's, that's burger across these different countries. So let's say in United States, you will find McDonald's selling a beef burger. But in India, you may not find them selling a beef burger. Instead, they are selling our ALU Tikki burger. And similarly in Philippines they are selling make spaghetti burger. So you can see that a single international brand selling a different type of product in different countries. And that's a really good example of geographic segmentation. Then we have a demographic segmentation. If you're segmenting a market based on different age, group of people in that specific market, or based on their gender or education level. Or maybe the family size or ethnicity or income group. In that case, That's a really good example of demographic segmentation. So if you look at these different FMCG brands like Unilever or PNG, these brands may launch different flavors or different types of product based on the income group of people, or based on their education level, or even based on ethnicity. And that's a good example of demographic segmentation. Then we have psychographic segmentation. If you're segmenting a market based on the interest, lifestyle, or subconscious motivator than it says psychographic segmentation. In the end we have behavioral segmentation. Let's go deep into segmentation and let's solve a real-world problem. Let's say you started working in a company and those guys are launching a new product that is plant-based milk. And let's say you are in the marketing team of that specific company. Now, obviously the first step of marketing campaign is to make sure that you are targeting the right segment. So you first have to find these different customer segment in the market because you don't really want to target general public. So you have to find all those segment of people who want to move away from the database product. So let's say this is your product and you have to find these different customer segment. Let's say after doing some market research and maybe talking to these different customer, you were able to find these two interesting customer segment are first segment is Segment number a. And this segment have all those people who are looking for a daily free alternative for lifestyle purpose. These are all high-income group people. Apart from this, you were also able to find a different customer segment, that is segment. And this segment have all those people who are lactose intolerant and they are looking for some other option. So lactose intolerant are all those people who are not able to digest dairy products, specifically milk. And in case if you don't know, the milk is normally digested by an enzyme called rennin. Rennin is present in maximum quantity, especially in kids. And as you grow older, the production of renin goes down in your body. In short, you have to segment of customer, segment E, have all those people who are looking for dairy free alternative for lifestyle porpoise. And these people have high-income. In segment P, you have all those people who are lactose intolerant and then looking for other option. So the first part of a marketing campaign is to make sure that you are going from undifferentiated mass marketing campaign to a micro marketing campaign. Because you don't really want to target abroad segment. You have to target of very narrow segment. And that's why we normally go from undifferentiated mass marketing campaign to a micro marketing campaign. Now, that doesn't mean that you don't really have to target a broader segment. If you look at a product like Coca-Cola. So whether you are a five-year old kid or a 65-year-old guy. Anyone can drink Coca-Cola. And that's why Coca-Cola will always try to have a undifferentiated marketing, also known as mass marketing. But because we are launching a duty-free alternative and we have a limited amount of marketing budget. In that case, we have to pick a smaller segment and we have to target them narrowly. And that's why we are doing a micro marketing. So I hope you understand the difference between these two. 9. Targeting in STP Analysis: So let's pick the same example to understand targeting in STP analysis. Targeting is the stage where you decide which segment you created during these segmentation fees is worth pursuing. So let's understand about the criteria to choose a target segment. So you have to pick a customer segment that have enough potential so that you can justify the marketing effort. This is because you might be putting a lot more financial and human resources in order to sell a product into the market. The next criteria is difference. So you have to ensure that you have enough measurable difference between all of these to different customer segment. Otherwise you will be unnecessarily duplicating the effort for these two different customer segment. The third criteria is reachability. Reachability means is your customer segment accessible to your sales and marketing team? I mean, which marketing channels will you use to make sure that you are reaching out to your customer? The fourth one is profitability. So in order to acquire a customer, you might be spending some amount of money, and that is your customer acquisition cost. And once you have a customer with you, than those people might be purchasing some product from your brand. And the frequency and the duration of their porches will give. And the frequency and the duration of their purchase will lead to customer lifetime value. And if your customer lifetime value is more than your customer acquisition cost, then you are making profit. So you have to check whether you will have profitability by targeting that specific customer segment. And in the end, you have benefits. So do you have enough benefits so that you can target that specific customer segment? So in our case, we have two different customer segment, segment E are all those people who are looking for a dairy free alternative. And segment P are all those people who have lactose intolerant. For segment a will go with cruelty free value proposition. And for segment B, we will go with dairy free value proposition. So let's look at both of our segment. Segment a is looking for a dairy free alternative. And all the people in this segment have high-income and they have some online presence. While for segment B, these people are lactose intolerant and they may or may not have a high-income. So all the people in this specific category are medium to low-income. And some of them might be present online, while few of them are also present offline. That means they might not be using any mobile phone or they might not have a social media presence. Now the reason we're focusing on income and channel is because if a specific segment is not using a mobile phone or if they are not connected with internet. In that case, how exactly will be target? Let's say as a marketing manager, if you have a choice that you need to pick just one segment out of these two different customer segment, then you should choose segment number e. This is because people in segment a have high-income and that's why these people can easily afford a premium product. And also all these people have some online presence. They might be using a social media app or they might be consuming some form of content using Internet. And that's why you can easily target these people. So to conclude this video, in this case, you have to pick a customer segment a, because this will have all the high-income group people who are ready to pay a premium price for quality lifestyle change in product. 10. Positioning in STP analysis: Hey everyone. Now that you understand everything about segmenting your customer and targeting them. And this video, let's talk about positioning. Positioning is the process of thinking about your product from the customer's perspective. So broadly, we have three different ways by which you can position your product. The number one is consumer-based positioning. So in consumer-based positioning, we tried to understand the pinpoint of the customer. And then we will align those pinpoint with the benefits of the product. Then we have a competitor base position. If you already have enough competition in the market. So let's say if you have multiple brands selling soy milk in the market, in that case, it's difficult for you to just highlight benefit and align the pinpoint of the customer. You have to also highlight all the benefits that your product have and how it is better than the competitor. And that's the competitor based positioning. And in the end, we have a price based positioning. And in this case, we will try to justify the cost of the product. So let's say if you're selling your product at a premium price, you also have to justify that premium price, that why we are selling our product at an expensive price. And apart from these three, we can also have a benefit based positioning and a prestige base position. So if you're selling a luxury product, whether it is a clothing item or a luxury forge, or a smartphone, then you have a prestige space positioning. So in the end, positioning is all about performing a competitor analysis and figuring out the value proposition of your brand. And then how can you communicate that value proposition with your customer? So let me give you a small framework. If you want to position your product into the market. In your marketing campaign, you first have to highlight what exactly is your product. Then you have to talk about the job that specific product do. And then you have to discuss about the outcome of your product. And finally, you have to ask yourself, why people should care about that specific outcome and why it is super-important to take an action. Now, let's start by highlighting what exactly does your product. So we will pick the exact same problem that we were solving in the last video, where we are trying to sell our soy milk in the market. So let's start by describing about the product. So soy milk is a great dairy free alternative. That's a description of your product. Then you have to tell people what job does this product do? We will be writing things like this product contains no fat. This have zero cholesterol and it also tastes amazing. Then we have to communicate the outcome of this specific product. If you consume or if you drink this specific soy milk. In that case, you will get omega-3, omega-6 fatty acid, and this will help you build strong bones. Now this is a good value proposition in order to target mom because our mom is really concerned about our health. And that's a good value proposition because majority of our moms are doing grocery shopping. And that's why, if you highlight the value proposition that this product contains omega-3 and omega-6 fatty acid. And these two are healthy fat that are super important in order to build strong bonds. In that case, your mom will end up purchasing this product somehow. Then you have to highlight why people will purchase this product. Because normally kids avoid drinking milk. And that's why, if you highlight the value proposition that this specific product comes in six different flavor and your kids will love it. In that case. That's a good way to tell people that. Why don't you try all of these six different flavor? In the end, you have to create urgency so that people will at least explore about your product. So you can highlight things like buy online on our website, or you can try out this product on Walmart or Target. And that's your urgency. So when you're creating a marketing campaign, in that case, you have to go from top to down and you have to use this positioning framework so that you will be able to position this product on the mind of people. So you have to start with giving the introduction of a product, and then you have to end the positioning by creating urgency about your product. And that's your positioning in STP analysis. So let's quickly summarize this video by understanding the benefits of STP analysis in marketing. The first benefit of STP analysis is that it will improve your engagement. So you have precisely targeting a segment and it is more likely to engage and convert from your marketing campaign. The second benefit is reduction in marketing cost because you have picked up very smaller segment and your precisely targeting it really well. In that case, you are not wasting your budget figuring out different marketing channels and different customer segment. And that is why your customer acquisition cost will be very less. And if your customer acquisition cost is less, and if your customer lifetime value is high, then you are generating profit. The third benefit is that you can create more robust product. Now because you have a clear understanding about the customer segment, Those people can give you instant feedback. And based on that specific feedback, you can also improve the product. These are the few benefits of doing STP analysis. In the next video, I'm gonna give you a small assignment and you have to complete that assignment so that you can test your knowledge. So this is the time you will test your knowledge by completing assignment. I'm gonna give you a small assignment where you have to find a customer segment. And then you have to find different ways by which you can target those customers segment. And in the end, you have to position about your brand in the mind of the customer. So Marriott Hotel have all of these different brand. I guess they have more than 20 different brands. So you have to find out the different customer segment that these people are targeting with these brand. And how can they position their individual brand in the mind of the customer? You have to find out the customer segment for, let's say Marriott Hotel. And how can they target and position about Marriott Hotel in the mind, of course, customer. Similarly, you have to find out the customer segment for executes t0 and how can the target and position about executes t0 in the mind of their customer. You have to complete this assignment by yourself. You can solve this assignment on a piece of paper. And I'm also going to attach the assignment and the solution in this specific video. 11. What is value Proposition?: Hey everyone, My name is now deep. And in this video, we will discuss about the value proposition. Now, before we discuss about what exactly the value proposition is, let's first understand about the structure of value proposition. A value proposition have a structure like this. So in your value proposition, you will first highlight your target customer, what your value proposition is for. And then you will highlight what kind of product your customer need or the opportunity you have in the market. Then you will mention about your product and what exactly your product is and all the benefits that your product will bring to the market. So that's the high-level structure of your value proposition. Let's look at the example and let's understand about the value proposition. So if I'll give you a small example, let's say in the last few videos, we were discussing a lot about the dairy free alternative, and we were discussing about soy milk. So let's take that soy milk blend as an example. So the value proposition of soy milk brand will go something like this. So for people who have active lifestyle, I'm looking for dairy free alternative. In that case, our product is good because it is healthy and our product contains omega-3 and omega-6 fatty acid that can improve your brain and muscle health. So that's the value proposition of a brand like soy milk. And we had a discussion about that specific product in the last video. Now, in this value proposition, you can see that we have covered for that is all the people who have an active lifestyle and who are looking for d roughly alternative. So we have also covered h2 and then we have r. That means what exactly our product to and in which category or product line. And then we have highlighted couple of benefits. So that is a high level overview of a value proposition. Now, before we talk more about value proposition, let's understand what value proposition is and what value proposition is not. So value proposition is a simple statement that summarizes why a customer would choose your product or service. So let's understand what value proposition is and what value proposition is not. Let's start with what value proposition is. Value proposition is exclusive. That means, how well does it highlights the competitive advantage of your brand and how it can separate your part. Also, value proposition is being focused. That means in value proposition you have to mention how your product can fix the customer's pain point and how it can improve their life. Also, value proposition needs to be specific. This means that you have to highlight the specific benefit that your customer will receive from the product. So these things will help you understand what your value proposition is. And then the last slide, we had discretion about that. Let's understand what value proposition is not. So value proposition is not a description of your brand. That means you don't really have to talk about what exactly your product is, how it is made up of these different ingredients and all of that. And obviously if those ingredients have some benefit, in that case, you can highlight that. Also, value proposition is not information. That means you will not talk about your product, your company, or your vision, mission, or who are all the people who started this brand and all of that stuff. Because value proposition is not the inflammation. Also value proposition is not a slogan. That means you don't really have to write some slogan or some catchy phrases for your brand. I mean, if you want it to do that, you can look at other brands attribute. So now that you understand what exactly value proposition is and what it is not, now, you might be thinking, fine, I understand the difference, but how do I create a value proposition of my brand? Let's say you might be working in a company as a brand manager or maybe as a marketing manager. In that case. Is there any framework that you can use in order to create the value proposition of a brand. So in this video, I'll give you a high level understanding or maybe an overview of a small value proposition building framework. So you first have to start with the market. You have to choose a specific group of customer that you are targeting. We had a discussion about this specific concept in the STP video, where we had a discussion about how do you go around segmenting your market into smaller groups and how exactly you can find different sales channel in order to target those customers. And obviously you have to position your product as well. So lets the first part of value proposition, where you choose a specific group of customers that you are targeting. Then the second part of this value proposition building framework is value. That means you first have to mention at least three to four benefit of your product. And when we talk about value, than value is nothing but benefits minus cost. And that's how the customer will look at your product. So you have to make sure that you are giving maximum value out of your product. So value in terms of benefit of your product and in terms of prestige. So if you're selling a premium product, the reason people pay for a premium product is because they are getting more than just benefit. They're getting a perceived value. Let's say if there's a premium brand or let's say there is a premium smartphone brand. The Nevada from benefit or uses. People also carry a luxury or a value along with the product. So that's number two. You have to find out all the ways by which you can build a value of your product. The third one is we are offering. And in that you will highlight. For this, you will use your product or service mix that you are selling. In this case, you have to work on product, price, place, and promotion. And we had a discussion about that specific concept in the marketing mix topic. Then we have benefits. So you have to write down at least three to four benefits that your product will provide. And then you have to highlight those benefits whenever you are targeting different customers segment. So if you are targeting moms in that case, you can highlight benefits like so Emily can help your kids improve their brain and bone cells. And when you're targeting a high-income group people, none, maybe you can highlight the specific benefit. Does the product have like this product will have omega-3 or omega-6 fatty acid. Then we have differentiation and, and that you will distinguish your product. Then we have a differentiation. So if you have a competitor in the market, then you can differentiate your brand in terms of price benefit quality, or GMO free on GMOs, genetically modified organism. So many brands use these hybrid ingredient or genetic modified ingredient that are not good. So you can differentiate your brand in terms of the ingredient to use, or in terms of benefit or in terms of flavor or quality. And different brands use different techniques to do that. In the end, you have your proof. And normally people use. Now normally brands use a third party organization or a health authority in order to show a proof that our product is approved by this, this, this, and it is used by 1 million people and all of them are happy and recommended by dentists and all of that. So these brands use these different techniques so that they can prove to the customer that this product is good for you and it is approved by all these health care professionals. So now that you have a good understanding about value proposition, now let's quickly do a small assignment. So in this assignment you have to create a value proposition canvas of a brand like Tesla. But you might be thinking, well, what exactly is a value proposition canvas? So first, let me help you understand what exactly a value proposition canvas is. And then you want to complete this assignment where you will create a value proposition canvas of a brand like Tesla. Now, you have to do all this exercise by yourself. I mean, if you do a small Google search, you can always find the solution of this assignment. But I will highly recommend you not to do that. Because doing this assignment will help you understand how do you or your concepts are. And these things are super-important. In the value proposition canvas. You first have to write about the basic detail of the brand and the product. So you will write things like the company name, the ideal customer, the different products that company has, and the substitute that are available in the market. Then on the right side, you have to mention about the customer. So in that specific section, we will discuss all the details about the customer. And on the left side, we will write things about the product. So let's start with customer. So whenever our customers thinking about purchasing a product, the first thing those people have in mind is that what all benefits to get from the product. So in this section you will talk about all the benefits or gains of a customer and what is their expectation and desire from a product. That in this section, you will talk about all the pins. So before a customer buy a product, let's say they may have some negative emotion or some undesired cost or some risk in mind. Since this section you will write all the risk or pins that the customer have in mind before he think about purchasing a product. And in this section you will talk about the jobs. That means what is the minimum expectation the customer have in mind from a specific product? So you have all these three section where we will talk about the gains, the pains, and the job. Then in the left section where we will discuss about the product. The first one is about the product and service. So you will highlight the list of product and service where your value proposition is built around. You will highlight all the different product and service. Then you will discuss about all the gains that these customer will get out of this specific product. We will discuss how your product or service can create customer. Again. In this section we'll talk about the pain relievers. So you can see that in the customer section they had some pain or negative emotion or some risk in mind. So in this product section, we will discuss how that product is. Eliminating the customer pains or negative emotion or undesired cost or situation that those people have in mind. This is the high level overview of a value proposition canvas. Now, you have to do assignment where you will highlight all the gains, pains, and the benefit of a product. For a brand like Tesla. I'm going to attach the assignment and the solution. You can download the assignment and just try completing that assignment by yourself. 12. Whta is SWOT Analysis?: So almost a few days back, I was scrolling through Instagram. Although I don't really spend so much of time using it. But I saw a video and in that specific video or person was seeing that many people fail in certain task in life because they are unable to channel their strength in the right direction. And as a result of that, they loose direction in life and become frustrated. Now some of you will say that, Well, why are you using that specific video in this course? Well, just like people have strengths and weaknesses, organization also have strengths and weaknesses. And that's what we're going to study in this course or in this video. In this video we'll talk about strength, weaknesses, opportunities, and threat of a company. Now, if you look at yourself or if you look at any person, you may have some strengths and some weaknesses. So let's say if communication is your strength as a person, in that case, you should choose domain like journalism or human resource. And you should avoid becoming a software developer or something else. On the other side, if your strength is mathematics, in that case, you should become a financial analyst or a statistician instead of becoming a human resource manager. So you have to make sure that you are generalizing your strength and you are avoiding your weaknesses. Or at least we are working on them. Now in this video, we'll talk about swat analysis. So let's understand what exactly is swot analysis and why as a business manager, we need to perform swat analysis. Swot analysis is an examination to identify its internal strengths and weaknesses, as well as its external opportunities and threats that will affect the business growth. Let's start with strength. So strength are all the things that you do well in your business. Now some business or companies use people as their strength. Some companies use financial resources as their strength. While some companies have operation management as their strength. So strength can be anything that you do well, then you have weaknesses. Weaknesses are all those things, all those departments of your business where you need to improve. So maybe you do not have enough people to work on a specific task. Or you may not have the best technology team in your company. All of these are your weaknesses. Then you have opportunity. Now the strength and weaknesses are always internally in your company. That means you have a full control on your strength and weaknesses inside your company and external factor cannot affect your strength and weaknesses, then you have your opportunities. And opportunities and threat can be influenced by external factors. If your company wanted to expand into a new market or they wanted to launch a new product. That's the opportunity. That means all the goals that you want it to achieve or you're looking forward to it. That's the opportunity. In the end, you have your threat. And threats are all the obstacles that you face. So let me quickly summarize all these things for you. So strength are all the things that you do well as a company. And these can be human resource or financial capital that you have with you. And these can be your strength because you can leverage your people or financial resources. If you have, then you have a sauteing weaknesses in your company. And you have to use either your human resource or financial resource to make sure that you are generalizing your weaknesses into strengths. Then we have opportunity. And let say, if you are working in a company like Apple or Google, and they have a lot more free cashflow in their company, or they have deep pockets so they can invest a lot more capital in the new opportunity or in new areas. In that case. In this area, you will understand how you can leverage your strength in order to create opportunity in the large market. You can take advantage of some trends. Let's say Google is investing a lot more capital in cloud computing and in the machine learning space. Similarly, Microsoft and Apple, along with Facebook, is investing a lot more capital in metaphors and we are space. Then we have threats where we have to understand about how new businesses can disrupt the existing business. So in the recent few example, if I talk about one of the product of OpenAI, that is Chet GPT-3. They are somewhat disrupting search engine. So people are using Chet GPT-3 to quickly get the answer. Instead of using Google, that's your thread. Your thread and weaknesses have internal origin. And your opportunities and threats are external origin. And strength or all those things that your organization do particularly well. And strength will help you distinguish your company from your competitor. So things like human resource manufacturing or financial resource can be one of your strength. Then you have weaknesses where you need to work on. And then you have opportunity and these have the external origin. So let's say if a government is creating a platform and as a company, you can build a product on the top of that specific platform. These are all the opportunities you have, or let's say if the government is changing rules and regulation and they are setting up some sort of compliances in a specific industry. And that's the opportunity you have to focus on. In the end, you have your threat. And threat are all those things that can negatively affect your business. So maybe you have more competition into the market, or there might be a supply chains shortest in the future due to some pandemic or something. These are all the possible threat. If you are able to identify threats or opportunity really well, before then your competitor, then you are able to create much better impact in the market. So in the next video, Let's do a swat analysis of a company like Amazon. And in that video we will understand how people at Amazon are doing a swat analysis in order to understand how they can channelized their weaknesses or how they can grab these opportunities and maybe avoid these threats in the future. 13. SWOT Analysis fof Amazon: So in the last video we had a discussion about swot analysis. And in that video we had a discussion about all the strengths, weaknesses, opportunities, and threats of a company, and how they can generalize these weaknesses, or let's say these opportunity in order to make sure that their business is constantly growing and they are generating profit. Now, let's try to apply swat analysis in case of Amazon. And in this video, we'll understand about strength, weaknesses, opportunities, and threats of a company like Amazon. Let's start with strength. Now, we all know that Amazon is a customer oriented brand. And I guess they have written this specific keyword into their mission statement as well. All the things that they do is very well aligned with the value that they provide to the customer. Then if you look at the e-commerce vertical of Amazon, then they have an amazing network effect. Now, if you don't know about network effect, let me try to explain that in a single line. So one of the reason you use a specific app, like WhatsApp or Facebook or Instagram, is because your family member and your friends are using the exact same map. That's a good example of network effect in case of social media. But if you talk about Amazon, so in case of Amazon, if you have more number of users using your e-commerce platform, then you would be able to attract more number of sellers selling their product on your platform. And in that case, you will have a network effect because you would be able to grab more number of user, because you have more number of sellers on your platform. And sellers will only sell on Amazon because it is the only product that have the maximum number of users. And that's a good example of network effect. In the coming videos, we have a dedicated video network effect. So I'm not gonna go deep into this specific topic. Also, one of their strength is large number of acquisition that these guys have done so far. I mean, from acquiring Whole Foods to couple of Robo-Taxi platform, Amazon do a lot more acquisition. And that is one of their strength because they know how exactly to generalize their acquisition well into their main business. Now let's talk about couple of weaknesses. Now, one of the business model of Amazon that is e-commerce is easily imitable. I mean, you can easily duplicate it. Like let's say if you want it to start at e-commerce brand, then you can simply create a store with the help of Shopify or Woo Commerce. Or you can use any e-commerce store builder. And you can create your own website and your own e-commerce store. Now I understand that it is typical to duplicate the network effect or it is not easy to drive traffic on your website, but it is super easy to create your own store. And that's why you can easily duplicate the e-commerce business model of Amazon. In fact, in some countries they have a lot more competition, especially in Southeast Asian countries, there are multiple alternatives. Amazon, like Flipkart and other platforms. Also one of the weaknesses is that Amazon doesn't have any control on their seller. Obviously, Amazon have a lot more negotiation power. So normally when these sellers sell their product on Amazon, in that case, amazon charged them around 25 to 30% commission. And that will include the logistic cost, handling the packaging and the reverse shipment and all of that. But still, I guess 30% is a very huge amount. And that's why many sellers also don't really sell their product on Amazon. Apart from that, they are also making losses and couple of areas. Although their Cloud computing business is doing really well, that is AWS. But I guess they are making a huge losses in the e-commerce business. Let's talk about the opportunity. Now. One of the opportunity that Amazon has recently started focusing on is the omni-channel expansion. Now, omnichannel is when you are expanding into more than one channel. They have recently started expanding into the physical store. And they are somewhat competing with Walmart or LDs. Also, they can do more acquisitions, So that is one of their opportunity. I mean, if a company is up and running and if they are doing good and they are somewhat overlapping with the business of Amazon, then most rapidly Amazon will try to acquire that company. Also, one of the opportunity that they have is to attract the Netflix customer. And they are doing that with the help of Amazon Prime umbrella. So just take a Prime membership, you will get Prime video. You will get prime music, one day delivery and so many other benefit. I guess that's their value proposition. They are trying to grab a little market share in case of OTT platform. Or maybe they are trying to do that. Not really sure how much will they succeed in this. But these are all the opportunities that Amazon hat. Let's talk about threats. So controversies related to the diverse of the founder of Jeff Bezos, related to him quitting as a CEO and joining the space race with Elon Musk and all of that. And what are their threat is aggressive competition from some of the other e-commerce players, especially in Southeast Asia or in Asia in general. And also a brand can easily. One of the major threat is that other brands can easily duplicate this business model, especially the e-commerce business model. So these are all the strengths, weaknesses, opportunities, and threats off Amazon. Now, let me give you a small assignments so that you can practice what you have learned from this video. So you have to do a swat analysis of Google. You can download this PPT file and you can complete this assignment by yourself. So that's all for this video. I'll see you guys in the next one. 14. What is Holistic Marketing: Hey everyone, My name is now deep. And in this video, we will discuss about holistic marketing. Now, holistic marketing is the integration of value exploration, value creation, and value delivery activities with the purpose of building long-term mutually satisfying relationship with all the key stakeholders. Now I know this definition may sound confusing. So let's understand the meaning of holistic marketing with the help of this diagram. So in short, holistic marketing is nothing but the management of all the important key aspect of the business. In a single line. Holistic marketing is the marketing management across different department. So if we discuss about relationship marketing, then in your business, you might be having these different distributor or supplier. And in that case, you have to build a relationship with those people. Then we have integrated marketing. So in your business, if you are selling across different distribution channel, let's say you might be having an e-commerce website or let's say if you have offline store, in that case, how do you make sure that you have a constant communication across all your sales channel? And that's your integrated marketing. Then we have performance marketing. So as a brand, you might be running these marketing campaigns or ad campaigns across social media. So how do you make sure that you have enough revenue, enough sales from those marketing campaigns? And you are doing all those things ethically. And how can you build a community around all those things. So that's your performance marketing. Then we have internal marketing. And as a company, it is super important to make sure that all your employees are also aligned with the vision and the mission of the company. So let's break down all these things are in the next slide. Let's start with relationship marketing. Relationship marketing aims to develop deep and enduring relationship with all your stakeholder, like your customer employees, marketing partners, like your distributor or supplier. And finally your financial community. Like your shareholder, your investor, and your analyst, who are talking about your brand, or who are analyzing the financial health of your company. Now, let's pick a small example because I love taking example. And I guess example can also help you understand the topic in a much better way. To understand relationship marketing, Let's take an example. I'm going to take an example of a company like Pepsi. So let's talk about the distributor of Pepsi. And Domino's Pizza is the biggest distributor of Pepsi. I mean, they have exclusive tie up with the brand. We also call this as a strategic partnership. And in Domino's Pizza, They are only allowed to sell PepsiCo product. And these people are there distributor or channel partner, whatever you call it. Then we have partner. So PepsiCo have multiple product like leaves, or they are also into cookies and odds. So their partners, let's say a farmer is the partner of PepsiCo. Normally, these brands have tie up with these NGOs or community members and they will buy all the potato these farmers are producing. These people are their partners than they have supplier like Walmart who is helping companies like PepsiCo and distributing these product to the end customer. So sometime if PepsiCo end up producing more of a certain product, walmart will give discount and push that product to the end consumer. So Walmart will do volume purchase and they will push these product to the end consumer. And that's what your relationship marketing looks like. If you work in a company like PepsiCo, in that case, you have to make sure that you are constantly building stronger relationship with your distributor, with your partner, supplier, or other stakeholders like investor or employees. And that's your relationship marketing. Let's talk about integrated marketing. So let's say if you are working for a company like Warby Parker or lend Scott, these are D to C IVR brand. And they have these different sales channel. So if you look at a company like Warby Parker, they might be having an offline store near you. Or you can also order your spectacles online. That's why they have a try at home program. And you can also use their online app where you can use the augmented reality technology in order to try out these spectacles on your fees. So if they have all these different sales channel, then they must deliver a consistent experience across communication, product, and service to make sure that all of these things are intact. That means whatever quality you will get in the offline store, you will get the exact same quality at your home and in the app as well. And the marketing or the communication across all these different sales channel will exactly be seen. And that's your integrated marketing. Third one is your internal marketing. And marketing. Activities outside your organization is as important as inside because it makes no sense to Promise excellent service before the company's stock is ready to provide those excellent service. And that's why internal marketing is super-important. Now, when we talk about internal marketing, we are basically talking about how do you communicate goals, aims, and objectives with all of your team member, or let's say all of your employees. So in that case, we will discuss about culture, staff motivation. We will do couple of training. We also have to establish a reward program. Then we have internal marketing. And if you look at a company, marketing activities outside the company is as important as insight. Because in reality it makes no sense to Promise excellent service before the company stuff is ready to provide. And that's why we have to focus on internal marketing. Now, when we're talking about internal marketing, we are basically talking about goals, aims, and objectives. And that's why in internal marketing, you need to make sure that your employees are motivated. You are giving them proper training and SOPs, also known as standard operating procedure. And you're also rewarding them time-to-time. And you're also ensuring that they are maintaining a quality standard and they also have a culture. I mean, that's the responsibility of a Chief Human Resource Officer in your company. So that's your internal marketing. Because you have to do couple of marketing activities inside your company as well. In the end we have performance marketing. And performance marketing is all about the understanding of financial and non-financial return. So performance marketing is all about finding a sales channel and optimizing the budget for these different social media channels or sales channel. And in the end, figuring out the ROI of the business. And optimizing on these three parameters. You can also go deep into performance marketing by looking at CPM, CPC, CPA, and LTV. Now, we will not discuss about all of these things in this video, but I have a dedicated course on all of these SAS metrics, or I would say these digital marketing metrics. And that course we had a discussion about course formulae or cost-per-click, cost per action and customer lifetime value and customer acquisition cost. Will not discuss about these things in this video. And that is all about your holistic marketing. Now, I have a small assignment for you. And in this assignment you have to write about the holistic marketing of a company like Starbucks. And in that case, you can cover things like relationship marketing, integrated marketing, performance marketing, and internal marketing. And you can refer to this video and you can take some help. And you can complete the assignment by yourself. 15. What is product life cycle ?: So I was a big Nokia, and Nokia used to be one of my favorite brand. So at the starting of my career, I bought their first phone, that was Nokia 11 double zero. Then I purchased Nokia 3110, then Nokia N9, T7. And I remember I also bought a Windows phone somewhere around 2015, 2016. So I was a big Nokia fan. But the reality is products like people have life cycles. And that's what we're going to study in this video. And this video, we'll talk about product life cycle. Product life cycle is the four-stage process that a product have to go through from boat to that. And you can understand the product lifecycle by looking at this diagram. So on x-axis you have time, and on y-axis you have sales or revenue of a product. And you can see that you start your journey from point a and then you go from point a to point E. Point E is the decline. And after that, you will have to, after e, You have to either spin-off or you have to kill the product. So let's start with process number e, that is the development. So at this process, you are investing a ton of money in developing a new product. And in fact, Nokia was also investing a ton of money developing new and new product for people. And they started off by developing big font so that people can communicate. So they were investing a lot of money in developing flip phone. You can open your phone or slide it like a laptop. So that was the development process where you invest a lot of money in solving the pinpoint of your customer and you are developing a new technology here. That is your stage number one in the product life cycle. Then the second stage is introduction. So at this stage, you are building a buzz around the product and you are creating awareness in the market for that specific product so that people can purchase it. And that's your introduction stage in the product life cycle. Then you have stage number three, that is the growth stage. And at growth stage you are growing at the exponential rate. You are killing it in the market, and you're generating massive amount of profit from that specific product. Now, at this stage, because you have a bigger market and you may not be able to fill all the gaps in the market. And this is the best time where you will have some sort of competition or some new players were entering into the market. Then we have our maturity stage. Now, all those competitor or small players who entered the market when you were growing exponentially, those people have now started dominating the market. And at maturity stage, you have enough competition in the market and your growth is stagnant and your sales will eventually decline or stabilize. And in the decline stage, your profits are declining and your sales is reducing. And now this is the best time you should innovate new product so that you can create a new market altogether. This is the product life cycle for almost every single product that you can imagine. I mean, this is not just about Nokia, this is about almost every single technology or physical product you can imagine. Now, let's understand product life cycle by looking at Nokia as a mobile brand. Because, I mean, who else doesn't know about Nokia? So this will be super interesting and it'll be easy for you to understand. So on x-axis you have time, on y-axis you have profit or sales. Now the fourth stage of product life cycle is development, where you're investing a ton of money in research and development. And you're quickly innovating on product and you are solving customer pinpoint at this specific stage in the product life cycle, that is development. Once you have launched your product, then this is the introduction stage where you are introducing a new product in the market. Now, Nokia introduced a new product. I mean, as far as I remember, back in 2010, I mean, Nokia is a very old brand. They had a product in 1995 and in 1980s as well. But as far as I can remember, I was the first user of their first mobile phone. That was 1110. I mean, it was a small mobile phone. Then they started launching couple of more product and that was their growth stage. So at this stage they were launching products like nokia 3110, Nokia N9, T7 series. Then they started innovating a lot. They started investing a lot more money in research and development. And then after the growth stage, they started launching multiple products. So they had a flip phone. They were launching our form that looks like a laptop. So that was their maturity stage. And this is somewhere around 2,014.2013 ish. And in the end, they started the brand was at the peak when they had the n series or the ECUs in their category. But after the Microsoft acquisition, their sales started declining. And at that point of time they had an Windows Phone. And in fact, I bought this Windows phone as well. Although the UI was super nice for these windows phone, but somehow they were not able to adapt well in the market. And developers were not that interested in developing apps for these windows phone. And because of that, Microsoft had to kill the operating system. And the soul, the smartphone division, or I guess the whole company to a Chinese brand or a conglomerate. I don't exactly remember the name, but that is the lifecycle of Nokia mobile brand. They started off with introduction stage, then they were growing at exponential rates somewhere around 2010. Then they were innovating really fast and they were launching the school product. You can flip these smartphone, you can fold them like a laptop. Then their sales started declining because they were not able to convince enough developer to build apps for their platform that has been dose before Windows, they were using operating system called Symbian. I mean, a lot of people might not be aware of that, but I was a big Nokia phone back then. But that's a product life cycle of Nokia mobile brand. I don't know why I'm so excited about Nokia, but this is a really interesting way by which I can help you understand how exactly does a product life cycle looks like for a product. Instead of Nokia, you can also take couple of more examples of different brands like Ford. And that's why Let's understand about couple of things about product life cycle. So if you look at this product life cycle and if you look at sales and profit, when you are developing a product, that means at the product development stage, you're investing a ton of money in research and development. And at that point of time, you don't really have enough sales and you are investing some of your profit into research and development. Now once the product is ready, then you have to invest a lot more capital so that you can manufacture the product. And at the introduction stage, you have to invest a lot of money in manufacturing and distribution of the product. And after the initial launch, once your product hit a certain stage of growth, in that case, you will start generating profit. So you can see from this diagram that, that the product is generating good amount of profit at the growth stage and at the maturity stage. And the profit will start declining once you reach to the decline stage. So Nokia invested a ton of capital in the product development stage. They were doing a lot of research and development. Now, once they launch the product, then they have to do manufacturing and distribution. So they have to bear more losses. Once the product is there in the market after the initial launch, they were generating massive amount of profit. And these profit will stagnant for a while and they will decline once you reach the decline stage. So that's your sales and profit in the product development lifecycle. If you want to understand the competitive landscape in the product life cycle than at the R&D stage or at the introduction stage, you have the monopoly in the market because you are the only player in the market who is selling a smartphone. But once you reach a certain inflection point in the growth stage than competition started coming. Because now enough gaps in the market that these competitors can fill. So the code for innovative form will look like this. First, you will innovate a lot. You do a ton of research and development. And once you are generating enough profit than the innovation or the experimentation will go down. Now the competition code will look like this. So once you are at the growth stage than the competition, co, will increase exponentially. And obviously, once the market is mature and enough, people are using a specific type of product and the competition will decline. So if we talk about the introduction, growth, maturity, and decline stage, then at the introduction stage, you have some innovative idea until the competitor started copying. Then you have your growth stage. Where all of these commutators started replicating your product or they are developing a new feature on the existing product by their own. Then you have the maturity stage where the product has standardized or innovative product have now become a commodity. Now these competitor or company have to work on cost or economies of scale so that they can produce the same product at the cheapest rate possible. And the term for that is economies of scale. And probably we will discuss about economies of scale and economies of scope in the coming videos. And in the end, we have a decline stage where the competition is increasing. And the one that has the minimum cost to produce a product will eventually win the market when you don't really have other brand advantage. So that's all about the product life cycle. Let me give you a small assignment so that you can practice what you have learned from this video. I'm attaching this assignment and you can download and complete the assignment of product life cycle for Ford. 16. What is Value Chain Analysis?: Hey everyone. In this video, we will talk about value chain analysis. So our value chain analysis will describe all the business activities it takes to create a product from start to finish. So things like designing production distribution and so on. Now to understand how a business create value will be using this specific diagram. So let me take the laser pointer. Now if you look at this specific diagram, you have some primary activities, things like inbound logistics, operations, outbound logistics, marketing and sales and service. All of these are primary activities. Then you have some supporting activities. So things like form infrastructure, human resource management, technology development, and procurement. These are all the supporting activities. And then as a company, you are giving out some value to the customer, and then you're also generating some profit and these are all your margin. Now, the meaning of doing a value chain analysis is to explain the relationship between these five dynamic forces that can affect the industry performance. Also, with the help of value chain analysis, you can assess the structural attractiveness of analyst industry. So let's understand about this specific diagram. Let's start by understanding about what exactly is this margin. So as a business, you are creating some value. And if you subtract your cost of creation of value, that's your margin or profit. So let's understand about value chain analysis. Now, we will first start with primary activities and then we will understand about supporting activities. And as you know, the amount of value that you are creating. If you subtract the cost of creation, that's your margin. Let's start with primary activities. Now, the first primary activity you do in a business is to buy all of these raw materials so that you can produce a product. So inbound logistic means all the raw material or all the products that are coming into your warehouse. That's your inbound logistic. To manage inbound logistic unit, a procurement team, you need to make sure that you have real-time inventory data. You also need to make sure that you have a supply chain control tower where they can see the location of all your distribution facility. You need to have some trucks for transportation these warehouses, and you need people to handle the material. And all of this is a part of inbound logistic. In fact, in some companies they have a dedicated team or department for inbound logistic and their responsibilities to make sure that everything is happening normally in inbound logistic. Apart from inbound logistics, you have to do some operation. And this will include day-to-day operation, whether it is manufacturing a product or if you're a distributor than storing the product in your warehouse for a period of time and then shipping it back to a retailer or to a customer. And these include all of your operation. So for operation department, they need to have access to the real times sales and inventory data so that they can supply you the amount of product you need in the coming week or in the coming month. And the operation team have some standardized model in their system. Then you have your outbound logistics. Outbound logistics is when the product is going out of your warehouse. So let's say if you are a distributor of a specific company, in that case, you might be shipping some product every single week or every single month to a retailer so that, that person can sell that product to the end consumer. So that's your outbound logistic. Now, their main role is to do order processing. They are also delivering the product to the retailer and delivery and processing is their main function. And companies have special outbound logistics department or team to handle this. So we are generating value as a business and that's why we have to understand the core pillars of that. So inbound logistics, operations, outbound logistics, marketing and sales. These are all the pillars of your business. In the end, you have your marketing and sales. And as a company, I'm sure that you are focusing on product, price, place, and promotion. So working on pricing, making sure that you have the right communication and marketing to the end consumer. You are producing new product based on their need. And you are selling those products at low prices. These are all the things done by a normal marketing and sales team. In the end, you have your services. Whether you are a hardware company or a software company, you always have a customer support or customer service team will ensure that they have the delivery, the installation, the repair of a product. And anytime they have any issue related to a product, they can always raise a ticket or they can always reach out to the repair center. So these are all the primary activities of a business in order to create value for the end consumer. So from procurement, that is your inbound logistics to operation, to outbound logistic, which means shipping out the product from your warehouse to marketing and sales and to service. These are all your primary activities. Now, in order to perform these primary activities, you need some supporting activities. At first, you need a procurement department who can ensure that you have the form material or semi-finished good in your warehouse all the time. If you are someone who is manufacturing a product. So the procurement department may need a real-time inventory data. They need to communicate with supplier and they need to purchase supplies and materials. For that. Normally companies use our ERP software like SAP or Oracle. Now, in order to make sure that you have a proper implementation of these ERP solution and you are using technology in your business. In that case, you need a technology department for technology development. So these people will make sure that you have integrated supply chain system. The all the department have real-time sales information so that they can plan their manufacturing AND operation. Then you have human resource management. Human resource management team will ensure that they are helping their employees in professional development. They are building a strong relationship with them. And then they are doing performance appraisal based on a proper evaluation parameters they have. And then obviously the main role of human resource management is to make sure that your salary is credited on time. Because that's the most important thing. Then in the end we have formed infrastructure. I think I should have covered all of the supporting activities from top to bottom instead of going from bottom to the top. So form infrastructure means that you have a good management team in your company. You are closing all your books and you have a good capital inflow in your company. That means you have a good cash-flow and making sure that you don't really have any legal issues and maybe planning for other normal day-to-day operation. So that's the underlying foundation of your companies. And that's how a company's able to create value. They have to perform all of these five primary activities. And for that, they need all of these supporting activities, department or people. And that's how they are able to create value and profit or margin. 17. Value Chain Analysis of Tesla?: Now I know that this value chain analysis may not be very much useful for a lot of people. So let's try to implement this specific concept to solve a real-world problem. So let's say you want to work in Tesla. In that case, you first have to do a value chain analysis on how exactly a company is creating value and also generating profit. For that, you can do a value chain analysis and you can use this template. Obviously, in though, last few minutes, I'm gonna give you an assignment. But before that, let's understand about the value chain analysis of a company like Tesla. So we will go from top to the bottom. At the top, you have all your forms, primary activities. And at the bottom you have the opportunities of reducing the cost. So let's start with firm's primary activities. So if you look at Tesla is a brand, they manufacture these electric car. Now if you're a car manufacturing company, then you will first start with design and engineering to make sure that you are building a high-quality product. Now, once you're done with your design and engineering and obviously it's ongoing process, then you will do purchasing material and component from all of the suppliers to have. Once all those components are there in your manufacturing facility, then you will do assembling. Once the car is ready. After that, you will do testing and quality control. And once your product is completely ready, in that case, we will do some sales and marketing and finally, distribution on dealer support. So these are all the primary activities you have to do as our electric car brand or as a car brand, then you have your total cost and importance. So over here we will write about the total cost of doing this specific operation. Now, you can write the monthly cost or yearly cost on this specific area. But the main purpose here is to understand Fitch activity is costing you a lot. And which one can you outsource to one of your partners or strategic alliance? So I was not having enough time for this video, so I was not able to pick the exact number from the financial statement of Tesla. So these are all the random number. So let's say they might be putting $164 million for design and engineering. And they might be putting to 30 million dollar for distribution and dealer support. You have to mention all of their capital expenditure across all of these different primary activities. Then you have your cost driver. That will help you understand how they can reduce down the cost over a long period of time. No cost driver will help you understand which of the process is costing them a lot and how can they reduce down the cost structure? Now if you look at Tesla as a company, they don't have a bunch of models. They just have four to five different model. And that's why they are not investing that much off. Capital in design and engineering. Then they have purchasing material and component. Now the price of these components and material depend on the amount of your purchase. So if you're purchasing in very large quantity, in that case, you would be able to get the exact same product at a much cheaper price. So if their supplier is present in United State, in that case, they might be purchasing those components at a higher price. But if their supplier is theory in China than they would be able to purchase the same component at a cheaper price. Very obvious. Then they have assembly. So now the cost driver in case of assembly depends on the scale of plant, the capacity utilization and the location of plant. Then they have testing and control. I guess they have a list of more than 800 different types of tests that they do before releasing a product to the public market. Then you have your sales and marketing. And I guess they invest very less capital in sales and marketing because Elon is their brand ambassador or their chief marketing officer as well. Apart from being a CEO, these are all the cost driver across design and engineering to distribution dealer support. Then you have to write about the link between these activities. So let's say if you have a high-quality assembling process, just like Tesla, where robots are assembling your car. In that case, you can increase the cost of quality because these machines have some SOP and they don't really work like human, where if you give more work to human, the quality of the end product will reduce. Also they have some Gigafactory located in China. So that's how they are able to assemble couple of products that cheaper price because, because the plant is near the cluster of supplier and that's how they are reducing on purchasing and distribution cost. Also, Tesla have fewer model and that's how they are reducing on assembling coast because they don't really have to configure these assembling machine. Every single quarter, then you have your opportunity and reducing cost. If you just sell one model across all these different countries. In that case, you'd be able to reduce down your cost structure. And that's very obvious because in that case, you will be producing just one single product and then you are selling that product in almost every single country. Also, there are some components that you can manufacture inside your company. So if Tesla as a company, is able to manufacturer couple of component inside their Gigafactory. In that case, you'll be able to get rid of these few supplier and you can also produce those products faster. So that's the value chain analysis of a company like Tesla. Now, you can invest some time and you can complete this assignment by doing a value chain analysis of a company like Apple. And you can write about the firm's primary activities, the total cost and importance, all the cost drivers you have in the company. And then you can discuss a bit about link between these primary activities and all the opportunities of cost reduction. You can invest some time and you can complete this assignment by yourself. 18. What is a business Strategy ?: Hey everyone, My name is now deep. And in this video, we're going to talk about what exactly is a strategy. So have you played any of these games like Clash of Clans or chess? Or maybe you have read a book like The Art of War. All of these things are based on a strategy. Even before you start playing chess, you have to look at the other person. And while you're playing the game, you have to constantly mixed strategy. And maybe you have to predict the next move of the other person. And that's why I asked you that if you have played any of these schemes or if you have read a book like The Art of War, then you already understand how the other person is making a strategy based on your Move. Now, if you look at Clash of Clans as a game, The objective of this game is to build your village. And then you have to design your bees and you have to defend it using the resources that you have gained by attacking other people's religious groups. And all of these things are real example of how do you make strategy as a human being. Now, just like these games, you also have to make some business strategy as a business executive. Because I'm sure after watching this course, you'll be working in some company, making either a business strategy or working with the operation team. Strategy is a well-defined roadmap of our organization. And the objective of strategy is to maximize the organization strength and achieve better performance. And while you are doing this, you also have to build a long-term competitive advantage. Now, in this specific strategy definition, you have to work on three important things. You have to figure out the strength of our organization. For that, you can do things like swat analysis. Then you have to make sure that the organization is having a better performance. And we'll talk about performance in the coming video. And in the end, you have to build a competitive advantage. And we also have a dedicated section on competitive advantage. When we talk about business strategy, then you can implement the business strategy, a three-level. You have your corporate level strategy, then you have business level strategy, and finally the functional level strategy. So let's understand all these three different level of strategy. With the help of business strategy pyramid. The business strategy must plan, is implemented by the management team to secure a competitive position in the market. So the way you build a competitive position is by making sure that you have efficient operation. You are working on customer satisfaction and you have some desired business school. We have to work at three-level to make sure that we are implementing strategy at all the level of business. So we have to figure out a corporate level strategy of business level strategy and a functional level strategy. Now if you're starting your career as a MBA graduate, in that case, there is a very high chance that you'll be working at making functional level strategy as you progress in your career. Let's say if you have five to ten years of experience, then you'll be making business level strategy. And if you are someone who have 15, 20, or let say, 30 years of experience, in that case, there's a very high chance that you will work in making corporate level strategy. So a business have these three different level and we will discuss about these three concept with the help of this strategy pyramid. So as the name suggest, business level strategy is related to a business or a vertical, or a product. So in business level strategy, you will answer questions like, how do we compete as a brand and how do we gain competitive advantage as a company? So in business level strategy, you normally figure out different ways by which you can improve your product. You can adjust your pricing based on the competitor. All of this is a part of business level strategy. Then you have your functional level strategy. And remember, business level strategy is a work that is done by people who have almost three to five years of experience. So those people exactly understand all the competitors, how those guys are selling the product into the market. Then we have the functional level strategy. So if you're starting out your career than there is a very high chance that you will be working at making these functional level strategy. So these functional level strategy often aim to improve the effectiveness of a company's operation. And these are normally developed by the first-line manager or the supervisor to solve the functional areas like marketing, production, human resources, and development. So if you're working in a warehouse or let's say in a production line. In that case, you will try to solve the day-to-day problem that you and your other colleagues are facing. And these are all functional level strategy that our individual employ a first-line manager or supervisor make. In the end we have corporate level strategy. And the strategy are normally made by people who have at least 20 to 25 years of experience and who are at the CXO position of the company. So corporate-level strategy is normally made by the top management. So in corporate level strategy, they will think about mergers and acquisition. How they can build a new business unit? Or do they need to shift their manufacturing facility to a different location? These are all your corporate level strategy, which are made by people in the top management or maybe CXO level people. 19. Sumsung's Corporate, funtional and busines level strategy: Now I know that some of you may have a question in mind that now the power we're going to cover this section and these concepts related to business level strategy in this course. So I have a small diagram for you so that you can understand how we will be covering these different concept related to business level strategy. So for internal analysis, we have two framework. We have a value chain analysis and a Brio model. And as the dome suggest an internal analysis, we'll be analyzing everything about that company. Then we have some external analysis framework. And in that, we will talk about Porter's five forces model and PESTEL analysis. Now once you have this framework ready for your company, then the top management can move on the strategy formulation by using this framework. So once you are done with all these framework, then we will discuss about value discipline, Blue Ocean Strategy, and Porter's generic strategies. And we'll talk about all of this and the coming video. Let's understand the business strategy pyramid. And the first part of that is your functional level strategy. If you're planning to work in corporate, then you will start your career making these functional level strategy. So functional level strategy can be defined as the day-to-day strategy, which is formulated to assist in the execution of corporate and business level strategy. So if you are starting out your job working in a company, in that case, you will work as a manager in one out of these five different department in a company. So let's say you might work as a marketing manager in the marketing department, or a finance manager or a financial analyst in the finance department, or maybe a human resource manager, a production supervisor, or research and development manager. In that case, you might be working in any of these five different department. Let's say you decided to work in the marketing department and you have a good understanding about marketing. And in that case, you'll be making the functional level strategies like, how do you figure out the right marketing mix for your product? And you will be doing swot analysis and concentrated marketing. On the other side, if you are working as a financial analyst or as a finance manager, in that case, you will be dealing with day-to-day operations related to closing the books, creating budget for these different department. How can you allocate funds to different departments AND operation? On the other side, if you get a chance to work as a human resource manager, in that case, your responsibilities are recruitment, development, motivation, retention of employees, and industrial relations. In production, you'll be working as a supervisor whose main responsibility is to make sure that you enhance the quality of the product by making sure that the high-quality material is being used in the manufacturing facility. And all of your colleagues or employees are working. In the end you have your research and development. And it's very obvious that in this department will be working on developing new product and how exactly can you innovate things faster? Now I know that the story is boring and that's why I'm not going to discuss more about business level or corporate level strategy in this way. Rather, I'll be using an example so that you can understand a functional level of business level and a corporate level strategy. So let's take an example and let's understand about business strategy pyramid. If you look at a company like Samsung, Samsung is a conglomerate consisting of multiple SBUs. Sbus stands for strategic business unit. And these SBUs have diverse product portfolio, ranging from smartphone to cameras to tv, microwave, and refrigerator. So each of these product or SBUs need a business strategy in order to compete successfully within its own industry. So at the top, Samsung have a corporate headquarters that is in South Korea. And from that corporate headquarters, they are managing all of these different strategic business unit, also known as SBU's. They have a different strategic business unit for silicon component. They have a different strategic business unit for smartphone. And they have a different strategic business unit for home appliances. If you look at their strategic business unit that is related to the manufacturing of the silicon component. Now, just to give you a perspective, almost 70 per cent of all smartphone or LED display are manufactured by Samsung. So whether you use a smart phone from brands like Apple, 1plus, apo. All of these, all at display are made by Samsung and they are the market leader in the OLED displays peas. Similarly, the manufacturer, majority of the RAM memory chip of all the smartphone brand. And Apple have our strategic alliance with a company like Samsung. Now, this may sound a little counter-intuitive because these companies fight a lot on the market share of their smartphone division. But they have a mutual benefit on the component division. At the top, you have your corporate level strategy. Then you have your business level strategy related to managing these strategic business unit. And then you have your functional level strategy in order to manage things like manufacturing, finance, marketing, and research and development. In the next video, we will discuss about how exactly as a brand you can figure out a strategic management process and how you can start implementing the strategy. That's all about this video. In the next video, we will discuss about strategic management process. 20. Components of business strategy: Hey everyone. In this video, we'll talk about the different components of our strategy. And to be specific, in this video, we'll talk about the components of strategies treatment. Now, there is a famous book written by Stephen Covey and the name of that book is Seven Habits of Highly Effective People. In that book, Stephen Covey said that begin with the end in mind. Now, this statement reminds me of the different components of a business strategy. Because as a company, you need to focus on achieving its long-term goal and aspiration. Before even attempting to accomplish anything. You need to think in terms of decades and not in terms of years. Now, if you look at a strategy statement, it has four different component. You have your strategic intent, then you have a mission statement, then you have a vision statement. And finally, goals and objective. So let's start with strategic intent. Now, strategic intent will help management to emphasize and concentrate on priorities. So you'll be inspiring people with your vision and your mission statement. And we'll talk about your vision and mission statement in a minute. Then you will be encouraging individuals and team participation. And then you will be utilizing your intent to allocate resources to all of the Strategic Business Unit you have in your company. So when we talk about the component of strategies treatment, the three most important concept you need to understand is your mission statement, your vision statement, and your goals or objectives. So let's understand about the mission statement first. So a mission statement describes what exactly your organization does, home it serves, and what makes an organization unique. Now when I'm saying what exactly organization does, these are all the present capabilities the company have. Now, what I'm talking about home, does it serve? That means all these stakeholders, like your investor or your customer and all the channel partners you have. And finally, what makes an organization unique means? What is the reason for existence? So if I'll give you a simple example to understand this, if you look at the mission statement of a company like Walmart, their mission statement is to give ordinary people the chance to buy the same thing as rich people. So basically they wanted to democratize the retail for everyone. And that's their mission statement. Now let's talk about the vision statement. A vision statement will help you understand where exactly as a company you want it to go. So our vision is a potential to view things ahead of themselves. Now, when you look at the vision statement of a company like Walmart, their vision is to become a world leader in retailing. Now, let's discuss about goals. Now, goals are more prominent in concrete, and goals are all the desired future state the organization wanted to achieve. So these are all the different components of a strategy statement. Now I understand that you might be a little confused between a vision statement and mission statement, what exactly goals and objective means. But in the next video, we'll go deep into vision statement. How exactly does accompany set a vision and mission statement? And how do they use goals and objective to go close to their mission and vision they have. So let's discuss about all these individual component of a strategy statement. In the next video. 21. Vision Statement of a company: Hey everyone, my name is not deep. And in this video, we will discuss about mission statement, vision, goals, and objective. But before we discuss about these things, Let's understand why do we need these things at the first place. So in the last video, we were discussing a lot about the corporate level, the business level, and the functional level strategy. So your entire corporate level and functional level strategy will work towards achieving this vision, mission, goals, and objectives. And that's why we need to understand this topic. Because if you look at all the senior executive of your company, whether he is a CFO or CEO of the company. They are inspired by these vision and mission statement. And that's why we have to understand about these vision and mission statement so that you can inspire all of your team members and employees so that they can choose a common goal instead of going into random direction. So to understand mission, vision, and goals, let's look at this specific diagram. So our vision statement will paint the future in your company for next five to ten years. Anytime you join a company, you may find them having a vision statement in their office or on the wall. And if you look at that specific vision statement, you will find that those people are painting the future of the company in next five to ten years. And that's their vision statement. I mean, as a company, they wanted to go to that specific point because that's their vision. Then you have your mission statement. And mission statement will help you understand fundamentally, the reason why the company exists today. So what they do exactly at this point of time and why they are doing that. And that's your mission statement. Now, in order to go from this mission to that specific vision, you have to have some goals and objective. These goals are the future state of an organization. That means you normally set the smaller goal and you will achieve these goals so that you can reach to your vision. So you start with your mission. That's your fundamental reason why you exist today as a business. And you have to achieve your vision in the next five to ten years. And you do that by setting smaller goals and achieve those goals by having these objectives. So let's discuss about vision statement first, and then we will discuss about mission, goals and objective. So our vision statement paints a picture of where you're going and why you want to go there as a company. And as you know, our vision statement normally paints the future of your company for next five to ten years. Let's say if you are someone who wanted to understand about the vision statement, or let's say if you are working with the executive in order to meet the vision statement for a company. In that case, you have to understand two core pillars of a vision statement. Where, and why. So you have to ask yourself, where are you going as a company? And what does success looks like in the future for your company? Now let's understand these two points with the help of one example, because I love giving example. Let's say you're starting out your career working in EV company, then your vision statement will look something like this, that your company, XYZ, wanted to be a leader in the manufacturing of electric vehicle. That's your vision statement. That means your main aim is to become a leader in the manufacturing of a electric vehicle. If you look at this vision statement, you will realize that our vision statement is normally short and brief. It is written in simple language. That means it doesn't have any jargons or any complicated word. There is hard for a normal person to understand. It is crystal clear and it should complete all the aspect of your business. If you look at this vision statement, they want to be a market leader in the manufacturing of a electric vehicle. So they have specifically highlighted their position by doing something in a specific segment, leader in the manufacturing of electric vehicles. Then it is also non-ambiguous and non-conflicting and it should motivate the employees. So that's your vision statement. Now this vision statement fits really well for a company like Tesla, but will not pick a company specifically for this video. But let's talk about the mission statement now. 22. Mission statement of a company: So a mission statement talks about the fundamental reason why you came into this business. It will help you understand about the purpose for existing into the business. Now if I give you a small example than, let's say a company like XYZ wants to make the most compelling car of 21st century. That means, as a company you wanted to create most compelling car. That's your mission. I mean, that's the fundamental reason why you exist today. You wanted to revolutionize the car manufacturing industry. If I combine the mission statement and divisions treatment, then the mission statement explains the company's reason for being present. While the vision statement gives its purpose for the future. These two mission and the vision statement will defines the overall future, or I would say the overall growth strategy of any company. Now, let's talk about the features of a mission. A mission must be feasible and attainable. And it is possible for a company to achieve it. Also, it should be clear enough so that action can be taken by the people who are working in the company or the senior executive team can work toward achieving this specific mission. Also, it should be inspiring, which is obvious because if it is not inspiring management staff and people, then there's no purpose of having a mission statement than it should be precise enough. Which means it should neither be too broad for a person to understand or know to narrow. It should be unique and distinct so that it can leave an impact in everybody's mind. Also, it should be analytical and it should analyze the key components of your strategy. And in the end, it should be credible so that all stakeholders should be able to believe on omission that accompany wanted to achieve. These are all the futures of mission statement. You don't really have to remember these features, but I was just covering these features in case if you end up working with a senior management team and writing about the mission, the vision, and the goals of a company, or let's say a startup if you are creating. Now let's pick a company because I love giving example. And let's look at the vision and the mission statement of a company. And after that, we will discuss about the goals and the objective. So let's pick Tesla Motors. If you look at the vision statement of Tesla, than their vision is to accelerate the world's transition into sustainable energy. If you look at their mission statement than they want to create the most compelling car of 21st century by driving the volts transition to electric vehicle. Nephew closely absorbed their vision statement, you will realize that this is their five to ten year goal that they have in mind. But if you look at their mission statement, then you will realize that this is the fundamental reason why the company exists today. Because they wanted to create the most compelling car company of 21st century. And the do it by driving the bolts transition into electric vehicle. That's their mission statement. So vision is long-term and mission is what exactly they want it to do as a company. Let's look at Google as an example. Let's look at their vision statement. The vision statement of Google is to provide access to the world's information in one-click. If we look at their mission statement while they want to organize the world's information and making it universally available and useful. And that's their mission statement. That means why the company exists today. That's the mission and what they wanted to do in the next five to ten years. So let's say in 20 years, that's their vision. Let's look at Amazon as an example. The vision statement of Amazon is two, be the world's most customer-centric company. And they wanted to build a place where people can visit and find and discover anything they might want it to buy online. If you look at their mission statement, well, they strive to offer the lowest price possible to customer with the best available selection and at the utmost convenience, That's their mission statement. So I hope after watching these many videos, you'll be able to understand about the vision statement and mission statement. Now let's discuss about the goals and objectives in the next video. 23. What are Goals and objective?: So from the last two videos, were discussing a lot about the vision statement and mission statement. In this video, we're going to talk about goals and objectives. Goal is the future state of your company and objective or the specific action that you will take as a company in order to achieve these goals. And I know this definition may sound a little confusing. So let me try to simplify this by giving you some example. Let's say as a company, your goal is to increase the revenue by ten per cent. And you want it to decrease the waste reduction by five per cent. That's a very realistic goal as a company. Now let's talk about objective. Objective or all these specific action that you will take in order to achieve these goals. So your first objective, in order to achieve the goal, that's your ten per cent increase in revenue and five per cent reduction in waste. That you can just add five new customer and you can retain the existing to customer. And you can simply increase your revenue by ten per cent. So if you're adding five new customers, let's say if you're selling a software product. So in that case, you can acquire five new customer and you can retain to customer. And that's how you'd be able to increase your revenue by ten per cent. That's your objective number one, by which you can achieve the first part of your goal. Now in order to achieve the second part of your goal, you have your objective number two. So our goal is to decrease down the waste reduction by five per cent. Now if you wanted to reduce down to faced by five per cent, you can follow two different approach. Either you can optimize the manufacturing process. So in a manufacturing process, waste can happen at three different stitch. Waste can happen at the time of manufacturing. So if the raw material is bad, then you may have a little more waste. Or a waste can also happen when you're packaging or product. So sometime people end up damaging those products. So you can optimize the manufacturing facility. And that's how you'd be able to decrease down the waste reduction by five per cent. Also, you can outsource the process to have byte label manufacturer or to a contract manufacturer. And that's how you can also reduce down the waist. So you have a goal in mind, that's the future state of your business. And you will take these specific actions called objective in order to achieve that goal. So in the end, goals are the desired future state of your business that will make your mission more prominent and concrete. And your goal should have the following features. You should have a precise and measurable goal. And one of the framework that you can use for a precise and a measurable goal is the smart framework. We'll talk about smart framework in the coming videos. But you have to have a precise call and you can also measure that specific goal over time. Your goal will always look after a critical problem, insignificant issues. That means achieving a specific target every single month or every single quarter should not be your goal. Your goal should be critical and significant for the business. Also, it should be realistic and challenging. And again, we have covered all of these things like precise, measurable, realistic, challenging in the smart approach. And it should include both financial and non-financial component. Now let's discuss about objective. So I assume that you already understand that objective are all the tasks that will help you achieve a specific goal. And objective will act as the foundation of planning. And these objectives have the following feature. You should not have a single objective. You can have multiple objective in order to achieve a goal. Your objective should be short-term and long-term and accompany. And your objective must respond and react to changes in the environment. So they must be flexible. And as usual, your objective should be feasible, realistic, and operational. Now personally, I'm not a big fan of theory, so you don't really have to memorize these concept or theory. I mean, if you have to work with a senior management or a senior executive in your company, then you can use these concept. Otherwise, you just need to know the basics of our mission, vision, and goals of the company. 24. Amazon's Mission, vision and goal: So now you have to complete a small assignment and you have to think about the vision and the mission statement of a company like Amazon. Please do not search this mission and vision statement of Amazon on Google. Just think about it. From an employee who works at Amazon. How does vision and a mission statement in Amazon looks like? I'm in what they wanted to become a next five to ten years. And what is their current state of business there next five to ten year. Ambition is their vision and their current state of businesses, their mission. So just think about it and maybe then you can continue playing this video. Because in this video, I will also help you understand about the mission and the vision statement off Amazon. So just spend some time and think about the mission and the vision statement of a company like Amazon. So I hope you have written a vision and a mission statement of Amazon. Now, your vision or mission statement don't really have to match with Amazon because the main purpose of assignment was to make sure that you are using your brain. You are writing about the vision and the mission of the company. So let's discuss about it. So the vision statement of Amazon is to be the Earth's most customer-centric company. And to build a place where people can discover and find anything they want, and they can also buy those products online. Their mission statement is to offer customer the lowest possible price and the best available selection and at the utmost convenience. Now if you look at this mission statement, you will find three important terms. Lowest price, best selection, and utmost convenience. And I think you have seen these three important keyword in the previous few videos that Amazon always wanted to focus on giving you a product at the best possible price, the different variety of a product. And they always focus on shipping the product in this shortest duration of time. And we had a discussion about this framework in the first video of this course, that Amazon wanted to be a faster, cheaper, and a better alternative to retailing. I'm in offline retailing. Now if you really want to understand how exactly amazon makes sure that they have a wider selection and they are selling their product at the most affordable price and they're shipping is super-fast. Then you have to understand about the flywheel concept. Now, if you don't really want to understand this, it's perfectly fine. You can skip this part, but this is a super interesting concept that you should know. So if you look at any company in this world, one of the reason a company can become successful and it is evil to target a mass market is when the company is having the lower cost structure. In a company you have to have a lower cost structure. Saw that you'd be able to hire enough people to do a specific operation. So Amazon has a lower cost structure and that's how they are able to maintain lower prices on their platform. So if you have lower cost structure and you're maintaining a lower prices on your platform, then customer will buy more number of product from your platform. That's very obvious, and the customer will have a good customer experience. So imagine if people are purchasing more number of product from your website or e-commerce app. In that case, the traffic is good. If the traffic is good, then these sellers will list more and more number of product. And that's very obvious because you have more number of people coming to your platform and they are buying a ton of product. In that case, more number of sellers, or we call these as third-party sellers, will be interested in selling their product. If more number of sellers are selling the product on an e-commerce platform, and that platform have more selection or variety of product. And if they have more selection or variety of product, then customer will become happy. This is a vicious cycle. And we call this vicious cycle as a flywheel. Because once the platform have a wide selection of products and these sellers are competing with each other because they wanted to listed on their product at the cheapest rate possible so that they can get maximum number of orders. These sellers are competing with each other. They are listing variety of product for people and they are selling at the most affordable treat. And that's why people are purchasing the product from the platform. And that's the flywheel concept. And this flywheel concept, we'll make any platform unstoppable. If more number of people are purchasing product from Amazon, in that case, more number of seller will list their product at the most affordable price because they are competing with each other and people are purchasing more quantity of product. So earlier, if a delivery guy was carrying, let's say 50 or 100 different boxes at once. Now he's carrying 300 or maybe 400 different boxes because more and more people are purchasing from Amazon. So the success of Amazon is a combination of lower per unit cost because of economies of scale, they have a flywheel running. And that's why Amazon is unstoppable. Now if you look at their goals that they have in their supply chain, is that they wanted to make sure that they have the right kind of product on the platform. And they're maintaining the right quantity of that specific product. And they are making sure that the product is being delivered at the right time. And they also have these quality check. The seller have to go through these quality check. And they are also sending to us product at an affordable price. And that's how Amazon is able to succeed in their business. 25. What is BCG Matrix?: Hey everyone. In this video, we'll talk about BCG matrix. And BCG stands for Boston Consulting Group. And before understanding about this specific metrics, let's understand what exactly is BCG matrix and why do we need it? If you go back to the first video where we were discussing a lot about the Samsung strategy example, where they were having multiple strategic business unit. So if you look at Samsung as a company, they have multiple strategic business unit. So they have a SBU for semiconductor. Then they have a strategic business unit for IT solution. They have one for LCD and visual display, and they have few for smartphone and then for home appliances. And you can see that from all of these different strategic business unit. So Samsung as a company, have multiple strategic business unit from semiconductor to IT solution to LCD and visually display, to smartphone, and finally the home appliances. And their most profitable one is this semiconductor. And I've told you in that specific video that almost 70 to 80 per cent of all, all at displaying your smartphone is made by Samsung. So whether you purchase a smartphone made by Apple or Samsung, or one plus or any other brand, there is a very high chance that they're all at display will is off Samsung. In order to understand in which strategic business unit or company should invest or divest, we have to understand about BCG matrix. So VCG is a tool that is used in the corporate strategy in order to analyze these strategic business unit or maybe product lines with the help of these two variable. The number one is relative market share, and the number two variable is market growth rate. Let's discuss about these with the help of a diagram. So as you know, BCG matrix is a tool that is used to assess the value of a product in terms of their growth and market share. Now, when we are discussing about growth, that means with the help of growth rate, you can understand how desirable the product is in the market. And with the help of market share, you can understand, do they have any competitive advantage as accompany or not? Now, BCG is a two-by-two matrix. And let's discuss about this two-by-two matrix. If you look at the first quadrant, then if a product of your company have a high market share and they have a high growth rate. It's a star product of your company. If you look at a company like Apple than iPhone is their star product, then if you have a product that has a low market share, but that specific product have a high growth rate, then that's a question mark product. And a really good example is MacBook. Macbook is having a low market share. If you look at the laptop, it as an industry in that case, you will find that almost 80, 85% people have windows as their primary machine or laptop. While less than 15% people have macOS or let say MacBook. Macbook is a really good example in this caution my quadrant, where they have a product that has a relative low market share, but the growth rate is quite high. Then you have a product that has a low growth rate, but a high market share. And a really good example is a product like iPad. If you look at iPad as a product, in that case, they have a high market share because iPad is the category leader in the tablet category. But they have a low growth rate because people are not that interested in buying iPad. I mean, they're more interested in a smartphone or a laptop. In the end, you have dogs. So if a product is having a low market share and low market growth rate, then that product will be categorized under the dog category. And you have to divest from these specific category. With the help of BCG model accompanying can easily prioritize in which product they wanted to invest their money, time, and effort. Now let's understand about all of these individual quadrant. Because if you look at a company like Apple, than they have multiple line of product and they wanted to put all those products into all of these different quadrant. So let's start off with star code red. So all the products that are there in this star quadrant, they are most profitable and they have a large market share. And therefore the company's advice to invest in these products so that they can generate massive amount of profit and they can prevent the star product becoming a cash cow product. Accompany need to heavily invest in these star product so that they can increase the market share. And a really good example of star product in reference to a company like Apple is iPhone. I'm an iPhone is the star product. The specific product has high market share and high market growth rate. So iPhone is the star product for Apple. Then similarly, you can look at the star product for different companies like Microsoft or maybe Adobe and all of these other companies. And I'm gonna give you assignment for that as well. Let's talk about the question mark quadrant. So all the products in the question mark quadrant hold a small market portion, but they have the potential to become a star product. And that's why company might be investing some amount of capital in these question mark product. A really good example in reference to Apple is MacBook. Currently, MacBook only holds around ten to 15% of market share, but they have the potential to become a star product. So if you look at MacBook as a product, MacBook holds just ten to 15% of market share, but that product have the potential to become a star product. You have to invest some capital into all the product that are there in the question mark quadrant. Then you have your cash cow quadrant. In cash cow quadrant you have all those products that are dominating a specific category or a specific domain. And they are generating good amount of profit. But that specific segment or category is not growing at the expected grade. As a company, you have to continue to invest in this specific product in order to melt the benefit from this specific category, then you have dogs. And in this quadrant you have all those product that neither dominate a specific market, nor they have the potential for a high growth rate product. And that's why it is in the organization's best interests to divest from these product in order to avoid misuse of companies. One. 26. Apple's BCG matrix: So now let's understand about BCG matrix by taking an example. And in this specific slide, I'll be taking an example of a company like Apple. And we'll be putting all the different products of Apple into all these different quadrant of a BCG matrix. Let's start with caution mock product. So caution my product have a lower market share and a high growth rate. So if you look at a product like apple Watch and MacBook, you will realize that these products have a small market share, but they are growing at a faster rate. And these are your question mark product. If you look at these star product like your iPhone or air pod or Final Cut Pro. These products have a high market share and they have a high growth rate. And the company have to put a ton of capital to make sure that they are innovating really fast in this product category. And these are all your star product. Then you have couple of products in the dark quadrant. And I guess I have some emotional feeling for this specific product for Apple. I mean, you can't really purchase it. Apple is not manufacturing this product anymore. But I hope you got the point. Then in the end you have some product in the cash cow quadrant. A product like iPad is a really good example of it, because this specific product have a high market share into the tablet category. But, but I guess this category is not growing at a good market growth rate. So that's why you can milk profit from this specific quadrant or all the product in this specific quadrant. Now I know some of you might be thinking that now we understand what exactly a BCG matrix is. But how do you exactly know where to invest or rare to divest? Because that was the core purpose of understanding about the BCG matrix. That in a company you have multiple strategic business unit. And you wanted to understand that in which strategic business unit you should invest or divest. So now that you understand everything about BCG matrix, you might be thinking, well, how do we exactly use this specific metrics? Well, in the starting of this video, we were discussing about strategic business unit and how as a company you will decide where exactly you should invest or divest. And that's why we are using a BCG matrix. So let's understand about the movement of your cashflow and desired movement of your focus as a company. So as you know that cash cow product have a low market growth rate, but they have a high market share. So all the profit that you are generating from your cash cow product, you have to invest that capital either into the product that are in the caution my quadrant or in the star quadrant. Also, you should try to move your question mark product. Because these products have a high growth rate and soon these product can become a market leader. So you have to focus on moving these product from your question mark quadrant two, this star quadrant. And you may have to make sure that these products have a high market growth rate as well as high market share. In the end, you have to divest yourself from all those product that has a low market share and low market growth rate. 27. Limitation of BCG matrix: So now that you understand everything about BCG matrix, let's talk about the limitations of BCG matrix. Bcg matrix is a framework for allocating resources among different business unit. And it makes it possible to compare multiple business units you have. But apart from that BCG matrix, also got couple of limitation and I think you should a bear about it. So BCG matrix classify your business as low and high. But generally if you look at businesses, they can be medium also. So BCG matrix may not reflect the true nature of your business. Also in BCG metrics are specific. Market is not clearly defined. I mean, when you are discussing about market share, are you talking about market share in a specific geography or in a specific category? So the market is not clearly defined in the BCG matrix. Also, if you have a high market share into a specific category, that doesn't mean that you can generate higher profit from that specific product. Because some of the product may have a high cost involved with a high market share. Maybe you might be boning a ton of capital into advertisement. In BCG matrix is not just limited to one single brand or one single product. You can put different product lines of different companies into BCG matrix. Also, if you look at growth rate and relative market share, They are not the only indicator of profitability. So this BCG model somehow ignores the overall indicator of profitability. Now if you look at the dogs quadrant, then many companies might think that we should divest from all of these strategic business unit or product line. But sometimes the product in this specific quadrant can help a business gain competitive advantage. And maybe they allow them to on even more cash than the cash cow quadrant. But in the end, the only reason why we are using BCG matrix is because it's very simple. It's a two-by-two matrix and it is super easy for a normal person to understand. Now if you want to get rid of a couple of limitations that BCG metrics have. In that case, you can try out the ADL metrics. I mean, let's say if as a business, if you have multiple product line and let's say as a business executive or as a business manager, if you are making a presentation where you want it to put all of your different product into a specific diagram or metrics. In that case, you can also use EDL metrics. So alien metrics will help you develop strategies that can be used to understand the competitive position of your product or service and the market share that you have. Or I would say, at which stage of your product lifecycle your product is right now. So an x-axis you have industries less market life cycle stage, and on y axis you have competitive influence or position. 28. Assignment - BCG matrix: Now I have a small assignment for you and I highly recommend you to complete this assignment by yourself. Because by completing these assignments, you can always test your knowledge and you can understand how much have you learned from this specific video. So you have to put multiple product of Adobe into these different quadrant in the BCG matrix. So you have to first figure out all the different products that Adobe have. And then you have to put all those different product into these multiple quadrant. 29. Understanding Marketing Analytics: So hey, everyone. Now we are starting the module number one of this course, and in this module, we'll build a strong foundation for marketing analytics. This module is all about revising a couple of marketing concepts so that you are up and running and we can start understanding about complex topic. Now in this module, we'll first understand about what exactly is marketing analytics? Why do we need it? How do you use it? And what exactly does it help you achieve? After that, we'll talk about four piece of marketing and STP framework. Four PD stands for your product pricing, place, and promotion and STP is your segmentation targeting and positioning. These are the two framework that can help you break down a bigger problem into smaller chunks. After that, we'll talk about the different types of marketing data. You have structured data unstructured data. And in the end, we will talk about a couple of key metric that you need to understand, especially in marketing. Let's start with our first section of the course. Before I start a module or a new section, I always try to create oversimplified videos so that everybody can understand about that module and why do we exactly need to learn all those concepts. Let's understand marketing analytics like I'm explaining it to a 5-year-old kid. Marketing Analytics is just like a detective who look at the glue and who understand what people like, what they buy, and what they ignore. So that businesses can sell better. Imagine you have a lemonade stand and you're selling lemonade to different people, and you're writing down what flavor people choose the most. Do people like sweet lemonade, sour lemonade, or minty lemonade? You also started noted down all the sales that you are doing, so you're checking on which day exactly are you selling the most number of laminade? Is it Monday or Sunday? Probably Sunday because more people are free on Sunday. So this is exactly your marketing analytics. You're looking for what worked in your business and what didn't. And based on that, you optimize your business, your idea or your process. Now the reason these data and these numbers matters is because these can help you sell more. You can save more money, and you exactly know what customer need. Marketing analytics can help a business make smarter decision by learning what already happened in the past. Now, whenever we talk about marketing analytics, you may have one question in mind. What kind of questions does marketing analytics help us answer? So typically marketing analytics answers questions like what worked in the past and what should we do next. And it can help you answer broadly these kind of questions. For example, if you're running some ad campaign on Facebook, on Google, on Tik Tok, or on any social media platform, with the help of data, you will understand which specific ad got most number of clicks, and based on that, you can optimize that ad. Also, you'll know what product you are selling the most during weekends, which customer will likely to come back, and it will also help us know how much did we earn for every single dollar that we spent in the ad. But the main conclusion is that marketing analytics can turn the guesswork into decision using real customer data. Well, after this, you might have one question in mind that hey, now I understand what marketing analytics is and what kind of question can it help me answer. But where does this data come from? How exactly do we get this data that we need to analyze as a analyst or as a marketing analytics manager? Well the data comes from everywhere where the customer touches your brand. It could be a website, email, offline retail shop, or maybe a mobile app. So you have different sorts of data that you usually collect using different tools. For example, if you have a website, most of the website use Google narratics and your website data can help you understand how many people are coming on your website, how long are they using your website, which paces they have visited, and where exactly they are bouncing or closing your website. Apart from website, you'll also have ad data. If you are running a Facebook campaign, Facebook, ad manager gives you a lot of data about impression, number of click, how much revenue it draw, and which all demographic people are coming on your ad. Then you can also get some Purchase data. If you're using Shopify or anything else for your content management system or CMS, you can get revenue purchase data as well. Obviously, whenever you send an email, you get a email data on how many people have opened the email, how many people have clicked on it, and a bunch of those things. The main conclusion is that this data is like footprint in the sand, and we follow these people wherever they went and we'll try to understand why. So to conclude, marketing data comes from many sources, both online and offline, and each tell a small part of a big story. Now, you might ask me, Okay, perfect. I'm able to understand what marketing analytics is, what it helps me achieve, how does this data come from? But how do I do a decision making out of this data? How do I tell a story or how do I find a exact goal that I'm achieving? Well, that's where your metrics and KPIs come into the picture, and obviously we'll understand more about it in the later part of this section. So metrics are the number that help marketer measures the success. So you have a dashboard, you look at the numbers and the metric, and then you are able to understand what's happening. Now, think of the metric as your report card. For example, one of the metric is CTR or click through rate. As the name suggests, it will simply help you understand how many people have clicked on a certain thing. For example, if you have a button and you really want to understand if 1,000 people has opened this page, how many people have clicked on a button, so you will look at the CTR rate of that click. Then you have CPC, which is cost per click. In fact, a better example to understand the whole story is ad campaign. Let's say you're browsing through Instagram and you suddenly saw an interesting ad where people are promoting T shirts. Now, thousands of people scroll through Instagram every single day. Most of them just ignore the ad. That's called impression. Even if they have not taken an action, it counted as an impression. But let's say five to four to five person people clicked on the ad. That's your CTR or click through rate, which are the number of people who have clicked on a specific ad. Then you have CPC, which is cost per click. That means out of 100 or 1,000 people, if four to five person people have clicked on an ad, how much are you paying for that? That's your cost per click. How much are you paying as a brand? You could be paying 50 cent or $1. Then you have conversion rate, which means how many people actually purchase that item. Let's say your ad was shown to 100 people, five people clicked on it, but nobody actually purchased it, so it didn't give you any revenue. That's where your conversion rate is also important, which will help you understand how many people has actually purchased the product. In the end, you have your customer lifetime value. Which will simply help you understand how much revenue a customer is giving. Let's say you Shawn your ads 200 people, five people clicked on it, one person purchased it. Now obviously that person, is that person coming back and purchasing again from your brand or how much that person has purchased? That's your customer lifetime value. How much a customer is worth over the full lifetime. In the end, you have ROAS or return on ad spend. That means if you spent $1,000 on an ad campaign, how much are you generating in return in terms of revenue? That's your RAS. A good ROAS is 2.5, which means if you're spending $100, your ad should generate at least $2,500 in revenue because obviously you need to justify your cost of the product, your shipping, your admin and marketing expense as well. We'll come back to each and every single metric once again. Don't worry about this. This is just an introductory video. But if I have to conclude everything in marketing analytics, you use a bunch of tools, consider these like superheroes or gadget to track, measure, and decide how to effectively run these marketing campaign, and we'll come back to these. But we use tools like Google Analytics that can track our website visit. We use meta ads manager to run campaign on Facebook and Instagram, we use email tools like Mail Cham. We use Google Sheet or Excel to do a simple analysis from a CSV export. We have dashboard. And these are all the bunch of tools that we can use as a marketing manager or as a marketer. So this is the oversimplified version of marketing analytics. In next video, we'll understand more about it. 30. The Four Types of Analytics: So great. In the first video, I hope you are able to understand what is marketing analytics and why do we need it? That was the oversimplification of this topic. In this video, I'll help you understand the type of marketing analytics. But before that, let's revise marketing analytics a bit. Marketing Analytics is the practice of using data to measure, predict, and optimize the marketing performance. Marketing analytics involves collection of data, then analysis, and then interpretation. So that you can take a decision. Now, the key goal here is that you first need to understand the customer behavior using the data, then you need to optimize the return on investment on your marketing campaign, and then obviously you improve how efficiently can you target and segment your customer. And obviously, if you have good amount of data, then you can forecast future trend, demand, or even sales. But marketing analytics is the backbone of informed measurable and scalable marketing strategy. And I'll help you understand these things more with the different types of marketing analytics technique that you can use. So broadly, when we talk about marketing analytics, you have four different techniques. You have descriptive analytics, diagnostic analytics, predictive analytics, and prescriptive analytics. And let's understand about each of them one by one. The first one is descriptive analytics. Now, descriptive analytics simply answers what happened. For example, if you run a marketing campaign, it will simply help you understand how much sales you got from that campaign, how much money that you spent, what were the number of clicks and impression. Then you have diagnostic analytics, which will simply help you understand why exactly certain things have happened. For example, you were looking at homepage data and you realize that people are bouncing a lot from homepage. Diagnostic analytics will help you understand why certain things have happened. There could be many reasons and we obviously understand each of these topic with example. Third one is predictive analytics. Which will predict in the future, what will likely to happen? So for example, you have 1,000 customer they are using your product. Predictive analytics will help you predict the action in the future or the probability of that action in the future. Like how many customer will churn or stop using your product, What will be our future sales? All of these questions could be answered using predictive analytics. In the end, you have prescriptive analytics. Which will simply prescribe you something that hey, based on all of this data and based on this campaign, you should do this. So things like optimizing your campaign could be one answer of prescriptive analytics. Now, these type of marketing analytics technique will help you from past reporting to future planning. That's the main idea of business using different types of marketing analytics technique to improve their decision making. Now, I wanted to talk about a few key application where you need marketing analytics and why this topic is so important. I know you hate PPTs and you don't want me to spend a lot more time on these PPTs, but I still want to explain some concepts so that your foundation is strong. So you use marketing analytics on a wide range of decision making across the different step of your customer life cycle, all the way from acquisition to retention to engagement to upsetting. The first place where we use marketing analytics is customer segmentation and targeting. Let's say you are a marketing manager in a grocery delivery app. You have thousands of customer. Now, if you start targeting every single customer persona with the same lens, you will have very bad conversion and engagement because the kind of messaging or marketing you need to do to an 18-year-old kid is very different from what you need to do to a 40 or a 50-year-old man. That's where you need to properly segment your customer and target them properly. There are techniques to do it like creating a customer persona or ICPs. We'll come back to that. Don't worry. Those are too advanced concepts for now. Second thing is digital campaign optimization. Whenever you run a campaign or ad on Facebook, Instagram, Tik Tok, or on any platform, you always choose your demography. Where exactly do you want to run this campaign? Do you want to show it to men, women, or to any gender? What kind of interest do they already have? You want to show a protein ad to someone who is active in fitness. Also, are these people have to be salaried or in college? What kind of income you expect from these people that we are targeting? You need to optimize these digital campaign and that's where you need marketing analytics. Third one is funnel and conversion analysis, which is very simple. You need marketing analytics to understand. Thousand people came to our website, 50 visited this specific page. Only ten people clicked on a certain product, but just two people bought it. What is wrong with our landing page? Can we improve it? So these kind of decision making can be done using funnel and conversion analysis. Then you have postalization and recommendation engine, which is a different topic in itself. But whenever you open Netflix or Amazon, you see the product that you actually wanted to buy on Amazon homepage, and that's all because of recommendation engine, that's there in Netflix as well. Based on your previous history, they recommend you the right kind of movies. In the end, you can calculate your customer lifetime value, you can predict the churn, we have dedicated videos on every single topic that you can see on your screen all the way from segmenting a customer to doing a funnel and conversion analysis to solving complicated assignments and case studies on customer lifetime value and churn and obviously marketing mix modeling and budget allocation. We have videos on every single topic that you can see on your screen. Don't worry. I don't want to rush through all of these PPTs, but I'm still covering them because I feel they are important for you. Let me give you some real world application of marketing analytics so that you are able to understand and relate with these things better. Companies use data and analytics to drive growth. They optimize spending and personalization at scale. A very simple example is that whenever you open Amazon or grocery app, just on the homepage, just wait for some time and look at the homepage. You will see everything that you actually wanted to order. The first real world example of marketing analytics is Netflix. Whenever you open Netflix, it shows you a recommendation of all the movies and web CDs and TV shows that you actually wanted to watch and that's all because of all the things that you have watched in the past. Looking at your data in the past, it personalizes the content and the thumbnail and suggests you what you exactly wanted to watch next. Second example is Amazon. Based on your last purchase and behavior and interaction with the app, it shows you all the recommended product that you actually wanted to order, and you can save some time as well. Third one is Coca Cola. Coca Cola has a very good placement technique where they place their product in the right channel to the right kind of audience. For example, they might be selling normal Coca Cola to in bigger events. But when it comes to some sports events or some events where your youth and fitness people are more involved, they try to market different kind of Coca cola, like Coca Cola diet or zero sugar. So they have a really good placement strategy on which product to sell to which audience? In the end, you have a food delivery app like Zomato. They create a RFM segmentation, also known as your recency, frequency and monitory and they do AB testing in sending push notification and discount. I'll come back to this topic. Some of these are little advanced concepts like RFM analysis, AB testing, what will come back to these things. But it's all about the types of marketing analytics and what it can help you achieve. From the next video, we'll understand about one of the most important foundational concepts every single marketer need to understand, that's your SDP and four piece of marketing. 31. How STP & the 4Ps Shape Marketing Analytics: So here, everyone. Now we will understand about four piece of marketing and STP framework. Now, these two are super simple concept, but they are important. Anybody who is new in marketing need to understand about these concept. In fact, they are very old. I'm sure if you have studied about marketing even a little bit, I'm sure you already know about this concept. But still, I want to revise the concept for someone who is new to marketing analytics. Why do we need four Ps and STP framework at the first place? So in today's world, you have a lot of data, you have a lot of brand, and you need to structure all of this information to take smarter decision because data is everywhere and without direction, this data can cause some confusion and teams often act reactively, not strategically when it comes to decision making. The solution is STP will help you decide who to focus and how to stand out as a brand. Now, STP stands for segmentation, targeting, and positioning. That means looking at your product, what segment you actually wanted to sell to, how do you target them and how do you position them? Targeting means what kind of channels do you use and positioning means? Are you selling affordable or premium or a standard pricing? Now four piece stands for product place, price, and promotion, and it can help you plan what to offer to your audience and how do you deliver them. So in simple term, STP is your destination and four P is your travel plan. You need to use your travel plan to reach to your destination. Let's understand about four P, our travel plan first. Four Ps is also known as your marketing mix. It stands for product, price, place, and promotion. These are the core building blocks of any marketing strategy. Let's talk about product. Whenever you try to sell a product to people, you don't highlight their features, you explain the benefits. So that's your product. You highlight benefits, maybe features, design, packaging branding. All these things comes under product. Then price, which means, how do you sell this product to customer? Are you going to sell it at an affordable price? Are you going to make it premium? And how does this price change because of other product or other brands? So obviously, you need to do a price sensitive analysis, competitor benchmarking. These are all the analytics technique where you can use and optimize pricing. But let's not touch that. Third one is please. We will you exactly sell the product? What are your distribution channels? Are you going to sell it online or offline? If offline, which partner are you going to choose? And how do you going to optimize the inventory and the logistic? Remember, at each step of the user journey, the cost will increase because when you sell offline, you need to find distributors, sockist, subtckist, retailer, and then you need to do marketing. In the end, you have promotion. Where exactly will you promote the product? TV ads are very expensive, even social media ads are very expensive. Are you going to find some influences, some PR agency? How exactly are you going to promote it? And how does your return on investment looks like by each channel? Now, these are difficult question to answer. Obviously, you need to work a lot on your product, you need to find a perfect pricing. You need to find a better place where you can sell the product and you need to obviously promote it. I can give you some real world example in some videos to understand each part of the process. So that's your four piece of marketing. Which are the four key pillars that are very important in marketing. The second step is SDP which stands for segmentation, targeting, and positioning. Now, STP will tell us who do we market and how do we differentiate as a brand? For example, if you're trying to build a soy milk brand, one of your target segment are vegan people who actually don't consume dairy. In that case, soy milk, almond milk is a good alternative to dairy, and that's one of your segment. Now obviously, you can have super complicated way to categorize the segment and we'll come back to things like clustering RFM. But the more simple segment is that, hey, from a broader group of audience, you need to figure out which is your core segment that can buy your product. There could be other segment like normal people can also have soy milk. Normal people can also have almond milk. But vegan is one of your segment where you can sell your product. There could be five or six different segments as well. Targeting. Targeting is that, how do you select a segment and align your product towards them? Obviously, when you target this specific segment, you have to highlight benefit that, hey, soy milk contains Omega three, Omega six, Omega nine, and good amount of protein so that you can have one glass of soy milk every single day. Then positioning. Are you positioning your product as a flavored soya milk or a healthy soya milk? It depends on your target origins. For example, if you're just targeting vegan and general audience for soy milk. You have to make it more delicious. On the flip side, if you're just targeting fitness people or people who are actively in sports or fitness, in that case, they don't need anything with sugar. They want a sugar free soy milk. In that case, you sell them a sugar free soy milk instead of flavored soy milk. And that's your positioning. How do you craft a unique value proposition for the audience you have selected? And that's where if you look at a soya milk brand, you will see a unflavored one specifically tailored towards people who are active in sports and fitness and flavored one for general audience so that it tastes better. Otherwise, soymilk doesn't taste that well. The main idea is that you're combining four Ps and STP analysis. So SDP defines your target audience, and FP defines how do you solve them? So you first segment your audience, then you target them and try to position your brand. And then obviously, you try to optimize for product, what kind of packaging features, quality of product you have. How do you price it? How do you distribute and sell it to the end customer, which is your placement, and in the end promotion? Do you need a celebrity? Do you want to run some discount campaigns? All of that. These are the two crucial important topic in marketing analytics. That's your four piece and SDP. Now I can give you a real world example as well. Look at AirBnB. Let's understand the way Airbnb use four piece and STP side by side. Let's start with segment. Their main segment are travelers and host who actually have a rented property and they want to rent it out. Now, they have multiple segment. They have budget segment, luxury segment, people with usual travel as a hobby, people who are traveling for a business trip. There could be multiple kind of people. Then you have targeting. How do you target young urban traveler who have flexible itineraries? So they need to make sure that they are promoting these product, they are promoting their property with all the channels possible. And how do you position them live like a local in a unique way that Airbnb has position? So you first need to find a segment, then you need to target them. For example, when you look at a corporate travel, it requires you to have a lot of corporate tie ups with different companies. And then obviously, you need to position different kind of properties as places where you can have a Wi Fi, a stable connection and do a bunch of those things. So perfect. That's about your FO Piece and STP. In the next video, we'll understand about the different types of data that is produced in a marketing analysis. We'll talk about structured data, unstructured data, and a bunch of those concept. 32. Exploring Different Types of Marketing Data: So great. Now you have a really good foundation about marketing. In this video, let's talk about the types of marketing data and why is this so important in a marketing analytics course. As a marketer, you obviously will get different kind of data while you are running a campaign or doing some analysis. That's why understanding about these three different types of data is so important. When it comes to data, you have structured data, semi structured data, and unstructured data. Let's understand about each one of them one by one. We'll first start with structured data. Structure data is clean numeric numbers that you can easily organize structure data is ideal for dashboard models and tracking performance of different marketing campaigns or even team members. This data is usually stored in rows and column, maybe in an excel sheet and a SQL database. It depends. And some really good example of structured data is maybe the pricing data, the customer relationship management or CRM data, your financial transaction, your customer information. All of this comes under structured data. For example, let's say you are running a campaign on Facebook or on Google, and whenever you do that, you obviously use a bunch of data. Let's say you wanted to run campaign on a specific kind of customer persona. So all your customer data like name, age, gender, all of this is your structured data. In fact, whenever you run a email campaign, the email open rate, the click through rate, the number of emails that are sent to a specific customer, all of that is a structured data. In fact, the conversion data that Google Analytics give is also structured, and even your sales transaction, your storekeeping unit, all of this is in the form of rows in column, and that's why it is a structured data. So to simply conclude anything that you see and store in Microsoft Exl is kind of a structured data. So data that you can store in rows in column is your structured data. Obviously we work with a lot of structured data. The second type is unstructured data. As the name suggests, this data is not very well structured. That means you cannot use it directly. You first have to parse this data and then you need to get some insight from the data. Unstructured data is qualitative, it's messy and it is rich in insight, but it's harder to analyze without any good tool. And the reason I say good tool is because you can still use large language model to parse images, videos, sensor data to make a sense out of it. But this data cannot be adjusted into raws and columns and usually company use vector databases to store this kind of data. That is a slightly advanced concept. We'll come back to this maybe in some other course. But the reality is that if you have thousands of images or let's say millions of images, millions of audios, you are not going to listen to every single one of them. So you cannot summarize, you cannot get really good insight out of that data, and that's why it is difficult to get useful information from unstructured data. Now obviously, you have Chat GPT, deepsk, a lot of these large language model, where you can pump in millions of images and audios, and it will start giving you insights, and you can store all of these things in your vector databases. But again, for a marketing person, this is still a very messy area to work with. Whenever a customer gives you review, they obviously write two, three lines of long review, and it is difficult for you to get actual insight from millions of reviews without using machine learning or AI tools. Second, is your social media post. You have thousands of posts where people are commenting different sorts of things, and it's really difficult for you to summarize or analyze that data. Then you have voice recording, images, and content. So you got the point. With structured data, these are simply in rows and column. That means you can calculate some average, you can multiply, you can find means, standard deviation, a bunch of those things. But with unstructured data, you have images, files, reviews. It's difficult to get a proper information. Now, obviously, in some sort of assignment, I can look at unstructured data, parse it in LLM, and show you how exactly do you get inside. But still, for a marketing manager or for a marketer, it is still a messy area to work with. Third one is your semi structured data where your data is in a structured format, but it's still difficult to get an actual information. Semi structured data blends the structure with flexibility, and this is very common in marketing system. These data usually have tags and hierarchy, but they do not have a standard schema. The reason I say that is because if you come from software development background, then you might came across JSN. JSN stands for your JavaScript object notation. Now, it's a industry standard format to store your data in key value pair. It sounds complicated for someone who doesn't come from software domain, but JSON is just like your rows in column that you see Exil but in a much more predictable fashion. So whenever you work with APIs or different tools, all of them follow a standard, which is JSON. They don't follow XML anymore, so that you can utilize that data. Maybe I can solve a couple of these problems and show you how exactly does it work. But with semi structured data, you have meta datas that are in key value pair, JSON and event logs from tools like mixed panel amplitude, you can still utilize this data, still analyze it, but it's difficult to store it in rows and columns, just like what you see in Axl. So that's your semi structured data. I'll come back to each different kind of data. Maybe I can show you or we can solve a small assignment and exercise. So perfect. Let's summarize all the things that we have learned in this video. You have three main different types of data. You have structured data that you can store in rows and columns in Microsoft Axel, like how many units you have sold, how many transaction happened, how many customers are there, and all of their information like customer ID, customer age, customer number, customer purchase. All of this can be stored in multiple databases or in simple terms, Excel sheet. And then you can calculate how many orders a customer has placed, what they have placed, and what is your average order value, what's your customer lifetime value, and a bunch of those metric. And that's super easy with structured data because you can store it in rows and column. With semi structured data, it's difficult to store directly in rows and columns, but you can parse your JSON and create a CSV out of it, and then you can summarize it. With unstructured data, you can still get some information, but it is difficult to put it in rows and columns, and you mostly need to store it in vector databases, and then you need to use AI tools to get some sort of information. So you can do sentiment analysis and a bunch of other things. So that's our semi structured structured and unstructured data. Maybe I can give you one example and maybe we can then solve one small assignment or case study. So let's talk about Spotify. Spotify is a streaming app where it gets a lot of data from its consumer who is using the app and obviously Spotify get all kind of data, including the structured data, semi structured data, and unstructured data. So if you are a data analytics or a marketer in Spotify, the structured data that you would have is user information, like, what's the phone number? What's the user ID? What's the location or the carrier the user is using? What's their subscription plan, what kind of music they are listening to? All of this is your structured data. Now, in an app like Spotify, whenever a customer click anywhere in the app, they usually trigger event log so that they exactly know how a user is moving into the app. Now, your event log data is usually in the form of event driven architecture, which looks something like a JSN data, key value pair. And in fact, not just that, all the music that they like, how much music they have listened to, all of this is also present in the form of semi structured format. Again, semi structured data can still be analyzed, but it is difficult to interpret and parse it. Or once it difficult, it is complicated. Unstructured data is what kind of music they are listening to, how much they have listened to, the name of the playlist, the transcription, all of this is unstructured data. Obviously Spotify have machine learning, AI engineers to analyze all of this data and make sure they are improving their recommendation engine to recommend the right songs to the user. But the main idea I was trying to convey is that as a marketer or as a marketing manager, you will mostly work with structured data or semi structured data. You will not work with unstructured data in most of the cases. 33. Customer Acquisition Cost (CAC) & Lifetime Value (LTV) – An Introduction: So perfect. In the last video, we discussed about the different types of data you may expect as a marketer. In this video, we'll talk about marketing metric and KPIs and why they are so important in analyzing the performance of your campaign or effort you're putting in a product. But let's start by asking one basic question why metric and KPIs are the central of modern marketing. The answer is, without the right metric and KPIs, you will lack direction and accountability and you cannot prioritize one thing or the other. Now the problem with a lot of marketer is that they focus largely on vanity metric to how many people their campaign is reaching, how many people are looking at the campaign, the impression, and they don't really care about ROI or return on investment. And they mostly make decisions without the clarity on the performance side, how exactly do you define the success of the campaign? And that's where these metric and KPIs will ensure that you have the right accountability and you are doing the right optimization. So things like customer acquisition cost, customer lifetime value written on ADSpnd these are KPIs and metric, we'll be discussing in this specific video. But before that, we'll first clear the doubt of the difference between marketing metric and KPIs and how they are different. Metric are the raw performance data, while KPIs are the strategic goal that you build for your team. So KPI stands for key performance indicator, and they measure how well you are progressing towards your business goal. While metric provide you data on the progress, the status, and your activities and progress. So when we talk about marketing metric, we're talking about customer acquisition costs, click through rate, written on Adspend. And KPIs are more goal oriented. So a metric could be a simple chon rate, 15%, but your KBI has to be that you need to reduce down the churn to 10% in 90 days using a specific action, let's say, a win back campaign. So you need to think of metric as ingredients and KPIs are the recipe that actually drive results. So let's cut the clutter and let's understand about our first metric, customer acquisition cost. As the name suggests, customer acquisition cost reveals how much are you spending in acquiring one paying customer? Could be free as well. And to calculate your customer acquisition cost, you simply divide your total cost by total number of customer. So let's say if you're spending $1,000 and acquiring 100 customer, your customer acquisition cost is $10. Now, when we talk about total cost, it obviously includes your sales cost and marketing cost. And obviously, you have to divide that by number of customer to calculate your customer acquisition cost. Now the reason customer acquisition cost is so important is because it measures your acquisition efficiency. As a marketer, you might be acquiring customer using your blog post, using your website, using your Facebook campaign, using your email marketing tool, and maybe due to word of mouth. You need to find out how exactly or how much exactly are you spending in acquiring a customer. You obviously have two different types of customer acquisition cost, blended CAC and normal CAC, but let's not talk about it and let's not make it complicated. But the reality is that you might need to spend a lot of money in acquiring customer from the ad campaign, and you can also get it for free, but you need to account for each and every small cost so that you know exactly how much are you spending and how much budget you need to allocate. Just as an example, if you're spending $20,000, acquiring 400 customer, your customer acquisition cost is 50. Now the reason this metric is important is because you will be measuring your customer acquisition cost by every single campaign that you're running, whether you're doing it on social media, TV, press, or newspaper, and by every single channel. To find the best performing campaign and how can you scale it up further? That's your customer acquisition cost in simple term. There are so many advanced concept inside customer acquisition cost. In fact, we can solve some real world case study assignment to understand more about this topic. Let's move towards our next metric, which is your customer lifetime value. Customer lifetime value estimate your total revenue. A customer will bring in their whole life cycle. This means that if you're acquiring a customer today and if the customer is using your product for the next three to four year, how much of total revenue that customer will bring to you? That's your customer lifetime value. And the formula is pretty simple. You have to multiply your average purchase value or average order value by the frequency of purchase in the complete lifetime. Let's say if I'm buying a subscription, let's say, a Netflix subscription for $10 a month and I'm buying it for the next five years, you have to multiply 120, which is your ten into 12 into five. Close to $600. That's my customer lifetime value where I as a customer will not use the product after five year. I'm assuming that. Your average purchase value multiply by average frequency rate. Now your average purchase value or AOV can be further broken down into the total revenue that you are generating in a time frame divided by the number of purchase. So for example, let's say you have hundreds of customer and they are spending different money and they are having a different frequency of purchase. Let's say some of these customers are on an advanced plan, some of these customer are on a basic plan, some of them are on a premium plan, and some of them have add ons, some of them I don't have any add on, and they are doing a purchase at different frequency. In that case, the easiest way to find a middle point of what is the average value of these customer buying from us is to simply calculate average purchase value or average order value. Obviously these customers are purchasing at different time, a different plan, a different frequency. The best way to normalize it is by calculating average frequency rate. That's where I mentioned that to calculate the customer value, you have to multiply your average purchase value with average frequency rate. Now, a customer lifetime value can be calculated using your average order value, purchase frequency, and how long are they using your brand. So if I'm paying $10, paying every single month and using a product for five years, that will be my customer lifetime value. The use case is that as a marketer, you need to know that if you're spending $50 acquiring a customer, how much of the revenue you can generate from them over the whole lifetime. Now, it is obviously difficult to calculate it because let's say you acquired a customer, they use your product for four or five months and then they churned off and you were not able to justify your customer acquisition cost. But they came back after six months, and that's why calculating customer lifetime value is slightly complicated. Now, obviously, customer lifetime value will also influence retention and upsell. Like, let's say a brand is spending $100 acquiring customer, and their first basic plan is just $5 a month. It will take two years just to get your money back. But in reality, brands are really good at upselling and cross selling, and if they're able to upsell or cross sell some other plan at 50 or $20, they are able to get their money back faster. Okay. As a marketer, you need to identify a high value customer segment, a low value customer segment. In fact, brand also personalize their pricing by geography as well. Some of those brands are and I can show you a couple of those plug ins as well that can let you customize or personalize your price by different geography. To conclude, customer lifetime value will help you shift your focus from one time revenue to long term revenue that you can generate from the customer. 34. Mastering the CAC to LTV Ratio: Now you have a really good understanding about customer acquisition cost and customer lifetime value. But if you carefully observe, these two things are actually connected to each other, and we call them as LTV to CAC ratio. Let's say you're able to generate $300 from your customer in the whole lifetime. This could be one year, this could be ten year. You don't know yet. But let's say your average customer lifetime value is coming close to around $300 and your average customer acquisition cost is $100. If you divide your LTV by CAC, your ratio is three is to one. $300 divided by $100 is three to one. Now that ratio is considered good in the industry. The reason being is that you spent $100 acquiring a customer. The customer is generating $300. That means you have $200 left where you obviously need to subtract your administration cost, operational cost, a bunch of other cost, and you will still have 50 to $60 left as your profit. So three is to one as your LTV to CAC ratio is considered good. That means unique you can only spend one third of the total customer lifetime value in the acquisition cost. And that's a thumb rule. Obviously, you can violate it and you can still build a good brand. And it also depend on different industry. Like, usually, SAS has high customer acquisition cost, but they also have really good customer lifetime value. But when you look at other categories where people just use the product once, maybe, let's say, shampoo, uh, a t shirt or anything that people usually purchase just once there, the reoccurring purchase is not there and your LTV to CAC ratio could be lower. If your LTV to CAC ratio is less than one, that means your lifetime value is closer to your customer acquisition cost, which is bad because let's say you spend $100 just in acquiring a customer and obviously, you need to incur some cost to build the product, to sell it, to manage. All of that cost is not involved. That's why you're losing money customer, if your LTV to CAC ratio is less than one. If this is 1-3, you're still at the risky stage, but it's fine. More than three, it's a healthy unit economics. So your customer acquisition cost has to be lower than one third of your customer lifetime value to justify that your business is good. Then you have your return on ad spend, also known as ROAS. Now, return on ad spend, as the name suggests, shows how much revenue your campaign is generating on every single dollar you spent. Imagine you spending $1,000 for something that you sold for $1,000. You have not accounted for product cost, administration cost, packaging, marketing, shipping cost. You are in loss. If this is 2-5, it's kind of in a healthy range because you can obviously subtract your product cost, shipping costs, packaging cost, administration cost. Just to make sure that you have enough money left on the table. If this is 8-10, that's excellent. Very less brand actually have eight to ten return on ad spend. But the main idea is that whatever you are spending in ad campaign, you need to ensure that at least you are generating three times more revenue than that budget that you're spending, that's your return on ad spend. This will simply help Marketer RoAS will simply help Marketer double down on winning channels and cut down on the underperforming spend. 35. Selecting the Right KPIs for Your Business: Now, another important concept you need to understand in marketing is funnel, and I'm sure most of you already know about the funnel. Whenever a customer visit your website, whether they came from Instagram, Facebook, organically from Google, they are at the top of the funnel. They just lend it on your website. They might read about your website, and that's your awarenesstge. Once they have read your website, they may or may not sign up and that's your interest stage, this one. Let's say they end up signing for your product, signing up for your product. Well, then they will browse through different features, different functionality, and different things you have. That's your desire. They have gone through everything, but they still may or may not purchase. Let's say if they end up purchasing your product, then that's your action. And they might come back again and again use your product. But if you carefully observe this funnel was broader at the top and it narrows it down at the bottom. That means a lot of people might visit your website or platform or app, but very less will show interest. But all those people who have shown their interest, they may not even explore your product. Even if they have explored your product, very less of them will take action like buying a subscription or purchasing an item, and very few of them will again come back and do the action. So when you look at a funnel, you have top of the funnel, also known as tofu, middle of the funnel, morfu and bottom of the funnel. Your top is usually broad and bottom is usually narrow. So when we talk about awareness stage, you have so many key performance indicator you have to measure. Things like impression, how many people are landed on your website, reach an ad recall lift. These are your awareness KPIs. Then you have consideration or interest. Things like how many people clicked on signup button, click through rate at two card, signup. These are your consideration. Where people are actually showing interest. Then you have your conversion where people are actually converting, they're signing up, they're adding their details, and that's where your actual conversion happens. Then you have retention and advocacy. Retention is when people come back and use your product and advocacy is when they raf for your brand. That's a typical marketing funnel that I'm sure most of you are already aware of. In the end, let's summarize everything that we have learned in this specific video. How do you choose the right KPI for your business? Obviously your KPI depends on the type of company or product you have at what stage in the life cycle of the journey of the business and what is your current goal. You first need to define your business goal that if you are starting a new business, is your business goal acquiring more number of customer? Is your business goal sell more number of product to the existing customer or you want to increase the attention. But in most of the cases, your business has all the goals. And when you have multiple goals, you have to set multiple success metric as well. Then obviously, you need to measure metric, things like retention rate, customer lifetime value, average order value, basket size, ticket size, customer lift. These metrics sounds complicated, but they are not. And then you need to set a KPI target on what do you want to achieve as a marketer? Do you want to increase revenue, reduce customer acquisition costs, increase activation in a specific period of time, reduce CAC by 15% in three months. This could be a KPI target. So that's your example. Let's say if you want to increase retention, your KPI would be to improve monthly repeat purchase from 30% to 40% by quarter three. Now, let's summarize everything that we have learned in this video, and let's create marketing KPIs for Spotify. When you look at Spotify, it's a music streaming app, and this app is used by millions of people around the world in different geography. They use metric and KPIs to scale and personalize their marketing campaigns or pricing across different reason. Let's talk about three most important KPIs, your customer acquisition cost, your customer lifetime value, and return on ad spend. Let's start with customer acquisition cost. Now Spotify sell their premium subscription in different countries at a different price. They adjust it by purchasing power parity. A simple example is that a person in India will obviously pay less than a person in United States. The simple reason is disposablele income. The person in India might be earning less than the person in the US, and that's why they will pay less for this company or this brand or this product. But you still need to calculate how much revenue are they bringing to your brand to how much are they paying? So customer acquisition cost will help you understand how much are you spending in acquiring a paid Spotify customer. Most of them will obviously going to be free. Then you have customer lifetime value where you need to know how long are they using your product and how long will they stick by. And you can use churn prediction model which will simply help you understand if somebody is not using your app, how likely they will churn. I have a dedicated case study assignment on this topic on how do you predict Jon. Third one is written on ADSPend. Obviously, for a customer to install Spotify, you need to advertise your app on Appstore Playstore, on Google, on Facebook and a bunch of those channels. So the KPIs you will be measuring in case of Spotify is that, how do you increase repeat weekly listeners by 10%? This is a retention KPI where you need to improve the retention of users in the app. Then you have the second one. How do you improve premium subscriber NPS by five points this quarter? Now, NPS stands for net promoter score. So anytime if you use a app, you might see a pop up. Hey, how likely will you refer this app to your friend? That's your NPS pop up or scale. If you rate five star, that means your NPS score for that specific platform is increasing. One product manager or marketing manager might have this growth KPI that hey, you need to improve the NPS by five point this quarter with all the premium subscriber. Third one could be just grow the artist discovery section by 25% for Zenz cohort. If you see this KPI is very specific. It is targeting a specific segment which is obviously Gen Z cohort by 25%, this main idea is to simply let them discover new artist. 36. Introduction to Normalization: So hey, everyone. In this course, I will be creating a lot of case study assignment, and one concept that you'll be using the most in solving these case study assignment is normalization. Before I talk about a case study assignment, so before I give you a case study assignment, let's understand about this specific concept first so that you can solve all the case studies and we will come across this concept over and over again whenever we are solving a case study assignment. Let's understand the difference between direct normalization and inverse normalization and what exactly it is and why do we need it at the first place? You will be using this concept a lot in all the case study assignment that I'll be giving you in the course. So let's first understand normalization with the definition, and then I'm going to give you some real world example to understand it better. Normalization is a way of converting values on a different scale into a common comparable scale 0-1. And the main purpose of normalization is to make sure that a variable, which is, let's say in rupees or percentage or in a number holds the same weight when you combine them together. Let me help you understand this with the help of a very simple example before I go into the formula. So let's say in your dataset, you have three important variable. You have delivery time in number of days. You have a customer rating out of five and you have shipping cost. Now, if you look at these three dataset and you need to compare the values in three different variables, in that case, you cannot do that. The problem is that all these three variable has a different scale. For example, your delivery time might be from one day to let's say ten days or 15 days. But your customer rating is 1-5 and your shipping cost could be from $2 to $20. If you're calculating in dollars, if you're calculating in rupees or Yan, then it can varies a lot. So if you have to carefully compare a specific data set, let's say you have one line item where you need to compare where exactly the delivery time stands with the customer rating and the shipping cost that you cannot do that. And that's where you need normalization. If you have, let's say, four dataset and you have delivery time in number of days and rating 1-5 and cost, and in all of these four dataset, if you want to compare the dataset, you cannot do that because remember, the scales are different. The purpose of normalization is to convert these values from a different scale into a common comparable scale 0-1. That means we will simply look at this value and then just try to create a new scale 0-1, and the formula is this X is equal to X minus minimum of X divided by maximum of X minus minimum of X value. And we'll come back to this formula, but let's understand a bit more about it. So if you look at the normalization table, I have four different dataset with me. Let's say the dataset of company A, B, C, and D. I have these four rows of data with me, and I have their delivery time in number of days. I have their rating and their cost, and I need to normalize these value from a scale of zero to one. So I will use this formula and I will normalize the value and the value comes out to be like this. My normalized time is 0.750 0.001 and 0.50. If you carefully look at the data, one means that all the values 0-1 normalization scale that are closer to one are high, while all the values that are closer to zero are low. So if you carefully look at this specific normalized data set 0-1 for time variable, you will observe that it's taking five number of days, which is maximum, and that's why the normalized value is closer to zero. In fact, it is zero. But if you look at this dataset is just taking one day and the value is closer to one or exactly one. Three is closer to three is in between and two is at 0.75. That's your normalized time. Similarly, you have normalized rating and normalized cost. So any value is closer to one is considered on the higher side, any value closer to zero is considered on the lower side. So if you look at the normalization score, the company C is the fastest because the normalization value is one, which is this one over here. Similarly, company B is the cheapest because their normalization score is one over here, and company A has the highest rating because their normalization score is one or closer to one. Now you might be thinking that if you carefully look at the dataset, you will realize one important difference. You always choose a company that has a higher rating but lower cost and lower delivery time. How exactly do you solve this problem? To solve the problem, you need to understand about inverse normalization. So we know the formula for normalization. It's X, which is the new normalization score. We want you to calculate for a dataset, which is X. So you simply subtract the minimum value of X from X, and then you divide that by maximum value of X and the minimum value of X, which is within your dataset. So if you have a long dataset over here, you can see that I have so many values. The maximum is this, the minimum is this, and I want it to calculate normalization for a specific data. So now, we know that we can calculate a direct normalization where the higher value is always better. Look at variables like rating. If you have a higher rating, it's always preferable. I mean, it's better to have higher rating. But when you want to prefer a lower dataset of anything may be time and cost. If you have low time or low cost, that means that value is better. Then you have to invert the direct normalization. The name for this is inverse normalization. If you want to invert anything, let's say your value is 0.20 and you want to invert this. Well, you can simply subtract this from one and your inverted value is 0.80. So when you have to invert anything, you simply subtract it from one. And that's why in inverse normalization, we are subtracting it from one. That's where you see that your normalization of rating was direct normalization, but your normalization of time and normalization of cost is your inverse normalization where we first normalized a dataset and then we subtracted it from one. So where exactly does this concept of normalization can help us? Well, this concept is quite useful when you are using a multi factor scoring model. Let's say you wanted to solve a problem that has multiple variables into the dataset, and you need to use those variables to create a scoring model. For example, when you want to choose a supplier, for yourself, you will look at their rating, their delivery time, and their cost. Now you need a higher rating, a lower delivery time, and a lower cost. You will do a direct normalization of rating and inverse normalization of delivery cost and delivery time. Similarly, if you have to calculate a banking credit scoring model, in that case, you will look at their income, debt to income ratio, and credit history length. You need a higher income where the higher is better. So you will simply normalize the value 0-1 where higher is better. So any value closer to one is better. But if you normalize debt to income ratio on a scale of zero to one, in that case, lower is better, you need to invert the normalization score. The best way is subtract it from one. Similarly, you need a higher credit history length. You need a credit history of three months, six months, one year, sometime even five years. Higher credit history length is better, so you don't really have to invert the normalization. Perfect. This concept is also useful for e commerce seller as well. In a ecommerce seller, you will look at their delivery time, sorry, delivery speed, customer rating, return rate, and sales volume. If you guess it, the lower is better in case of delivery speed. The higher is better in case of customer rating, lower is better in case of return rate, and higher is better in case of sales volume. Wherever the higher is better, you simply normalize wherever you prefer the lower value, you invert. Same goes with marketing campaign. If you're calculating how good your marketing campaign is performing by a specific channel or in a specific event, you look at the conversion rate, where if you have a higher conversion rate, then it's better. You look at the customer acquisition cost. Lower customer acquisionOst is always better. So you will invert the normalization. Same goes with retention, where higher is better and click through rate, where higher is better. So I know this sounds a little confusing, but don't worry. Whenever we are solving these case study assignment, I'm going to use this normalization concept a lot and I'll be using this formula to calculate the normalization. So if I come back to this formula now, let's say you have a lot of values. Um, if you go back to the dataset, over here, you can see that I have four rows of data and I have these values two, five, one, three, where the minimum is one, the maximum is five and I'm normalizing a specific value. Let's say, I might be normalizing two or three or one or five, or in fact, you have in fact, if I have 100 more data into this dataset, I can normalize that. You need to take the X value that you want it to normalize, you need to figure out what's the maximum value, what is the minimum value in that column, and then you normalize that X value. Over here, the X that I wanted to normalize is 11.69. My maximum value is 22.78, my minimum value is 3.65. If I have to normalize this value, then I'll simply write X, which is 11.69 minus minimum value divided by X of maximum and simply subtract the minimum value from it. So B three minus B five divided by B four minus B five. That way, I can calculate my normalization of this X. Similarly, if you have, let's say X one, X two, X three, and hundreds of dataset, you can follow the same formula to calculate the normalization. And if you prefer a lower value to be a better, you simply invert the normalization by subtracting it from one. Please don't worry if you're not able to understand the concept just from a theory perspective. Whenever I'm solving the key study assignment, I'm going to explain this concept once again. But I just wanted to create a dedicated video so that you can come back and watch this concept again anytime I use a complex normalization formula while I'm solving all these case study assignment that you can see at the top. 37. Assignment: Campaign Evaluation – Retention and Revenue: So here, everyone. Now we are solving one small case study assignment where we will be evaluating the campaigns that we are running on Facebook, Tik Tok, Instagram, and we'll be looking at the revenue of the campaign, the customer acquisition cost, and what's the retention of the users that we acquired. But first, let me help you go through the problem statement, the dataset. What do we need to solve and why this case study assignment is super important and how it will help you solve a real world problem. So I'll be first taking you through this document, and then I'm going to show you all of this dataset and then we're going to solve it step by step. I'm also going to explain you all of these variables that are there in the dataset. Let's go back to the dogs file. So great. Let's start with the problem sttment. So let's say you are working in a performance marketing team at a ecommerce company and they are giving you the campaigns that they were running in social media with influences, and they also have some referral campaign. In this specific data set, you have all of this data, things like campaign ID channels, how many users were acquired, how much money they spent in acquiring those users. What was the revenue and retention? What is the business challenge or the problem that we are solving? You have all of this data of this campaign, things like how much they spent, how many clicks we got. Did we converted a user into a paid user or not? And you also have some retention data. But the company doesn't really have insight about this because they were not able to realize that if they are making money or if the campaign is actually performing good or not. Let's look at the dataset that we have. And I have explained everything in the talks, but let me show you the actual dataset first. So you have campaign ID, which is a unique ID for every single campaign. You have channels in which you are running that campaign. A channel could be a referral campaign, an influencer campaign, an email campaign, or a paid social media campaign. This is the sponsored ad that you see on social media. Then you have number of users that were acquired from that campaign, how much money we spent, how many orders we got from that campaign. What was the revenue from these orders? And how many of these users actually retained after 30 days? So if you look carefully, you have these many users that were there after 30 days. Perfect. Now in this specific case study assignment, you first have to calculate the customer acquisition cost, which is super simple. We will then be calculating the return on ad spend, which is a metric to understand if you're making enough money from the amount of money you are spending. Return of amount spent shows you if you're making enough money so that you can spend more money on the upcoming campaign. Then you have retention for 30 days, and now you see a unique term normalization roast normalized retention, and inverse normalization cap. So if you don't understand the meaning of normalization, I have created a dedicated video on what is normalization and how does it work? And in that specific video, I've explained about normalization and inverse normalization. I highly recommend that you go back and watch that video, but don't worry, I'm going to explain it in this video as well. So let's go back to the data set. So in case if you don't know about the technical concept or the variables like CAC RAS, let me explain you using this specific dogs. So CAC is your customer acquisition cost. This is the money that you are spending in acquiring a customer. So if you're spending $100 and acquiring ten customer, then your CAC is $10. Then you have ROAS or return on ad spent. This is the amount of money you are spending to earn every dollar. If your company is making $1,000 and you're spending $100 in the campaign, then your ROAS is ten. Then you have 30 days retention, which is self explanatory. 30 days retention means that how many users are staking to your platform even after 30 days. Then you have a concept of normalization and inverse normalization. Let me explain you with a very simple example. Let's say there are so many students in a class and you have their data, things like the marks they got in their math exam, what's their height, and what's their body weight. If you look at this dataset, the exam marks are between zero, 200, height is 5-6 ft or even more than six and weight is between 50 to hundred Kg. If you look at these three different dataset, they are on a different scale. Weight is in Kg between 50 to hundred, height is in feet 5-6 and mark is between zero to hundred. These three scale are very different. You need to normalize these three scale on a uniform scale 0-1, where you will look at the height weight and marks of one student. Compare that with all the student in the class and give it a score 0-1, where zero means that that specific student lies in the lower or the higher band. Zero means it is closer to the lower band, one means it is closer to the higher band. Whenever we talk about preference, what is your preference? Will you choose somebody who is scoring less in math exam or will you choose somebody who is scoring more? Well, you go towards the higher normalization score, anybody who is scoring more. How about weight? Will you prefer somebody having lower weight or higher weight? I would say lower weight, and how about height? Well, I will choose maybe somebody who is very tall. You are preferring a student that is scoring really good in exam is really tall and has less weight. In this case, if you look carefully, the weight should be less, and that's why you need to inverse the normalization so that you can counterbalance it. So if you subtract anything from one, that is your way to normalize the normal score. Please go back and watch the normalization video if you're still confused. And then we'll be calculating the campaign score by looking at variables like K RAS and retention, and then we will optimize the budget so that the team can run the marketing campaign. Now let me go back to the sheet and we can start solving the assignment. So perfect. I first need to calculate the customer acquisition cost. The formula to calculate the customer acquisition cost is very simple. You simply need to divide your number of users that you have acquired by total cost. If you're spending $100 acquiring ten customer, then ten is your customer acquisition cost. The formula is very simple. Then we will calculate ROAS or return on ad spent. The formula for this is also very simple. You're simply looking at the total revenue that you are generating from a specific campaign and the money you spent. If you divide the total revenue by the total money you spent, you have your oAS. Over here, if you look carefully, the Roas is 1.31, and that's because you generated close to $31,000 or this value could be in millions as well. But let's say you spent 30 you generated $31,000 in revenue and you only spent 24 and your Roas is 1.31. Generally, 1.3 Roas is not great if you are early stage starter. You at least need a Roas of three to four times. Then you because generally, if you're spending $100 in the campaign, you need 300 or $400 in revenue because you are selling a product at a specific price. You also need to adjust the cost of that product. For example, let's say you're selling a wallet at $20 and you spent $50 in a campaign. So if you generate $400, that means you sold 20 wallet if your par wallet price is $20. Now, if you're selling a wallet at $20, obviously there is some cost associated with the wallet. Let's say it takes you $10 to procure a wallet from a supplier. Your profit margin is only 50%. So out of $400 you generated as revenue, you only need to account $200 as your gross profit margin. You still need to subtract your salary, your shipping cost, your admin cost, and you might end up with five to 10% of net profit. That's why Roas of three to four times is really good. More than that, it's a win win. Let's look at retention of last 30 days. The retention data is already given, which is simply the users that were returning after 30 days and the number of users or unique users you acquired in a specific campaign. Over here, 3655 users were returning back and you acquired 2162 in a specific campaign. Your retention for 30 days is mostly positive or on the higher side. In some campaign, it is not that great, but mostly it's better. Now it's our high time to normalize the score, and this is very important. I highly recommend you that you search a normalization video in this specific course. You watch that first, and then you come back to the assignment. Because if you don't understand the normalization concept, you may not be able to solve it. Let me first build some context before I explain the normalization. So how about normalization of RAS? When you have to normalize the ROAS for, let's say campaign number CMP 001, you are normalizing this value, 1.31 with all the data set you have. And you'll be putting 1.31 on a scale of 021 after looking at all of this data. Similarly, if you have to normalize retention and if you're normalizing the value of campaign CMP 001, then you're normalizing 1.69 out of this complete dataset or the full column, and you are placing this data on a scale of 021. Similarly, why are we inverting a normalization of Kak? The simple idea is lower Kak is always better. That's why you will normalize CAC and then you invert it. Over here, higher ravas is better, higher retention is better, but you need a lower Kak because lower Kak is better and that's why it's inverse normalization of so perfect. Now I can quickly use the formula and calculate the normalization. The formula is pretty simple. In case if you don't know about the formula for normalization, let me open a box for you and show you the normalization concept. In fact, you can go back and read about normalization. Perfect. This is the normalization formula. I highly recommend that you go back and read this document. So if you have to normalize X value over here, this is your X value. 1.31 out of all of this dataset, you figure out the maximum in the column, the minimum in the column, and then you apply the formula. So if you're normalizing a value X, then you simply use the original value, and then you subtract the minimum value from it, and then you divide that by maximum minus minimum of that specific column. Over here, you first need to figure out what's the maximum, what's the minimum, and then you use this specific dataset which we are normalizing. Similarly, you can normalize retention as well by using the same formula. So we are normalizing Z J two, which is this value from the complete dataset, and we are looking at the minimum and the maximum, and then we are dividing maximum minus minimum. Perfect. And we obviously have to invert the normalization value of CAC. You can do it in two steps or you can just do it directly. You can go back to this document in case if you're confused about normalization. In the end, we need to create in the end, we need to calculate the campaign score. Remember, the effectiveness of any campaign is a factor of how much money are you able to generate? What is the retention of those user and are you spending less while generating that money? If you go back to this document, I'll show you the formula. So while creating a campaign score to find out the effectiveness of campaign, we are giving a 40% weight or contribution margin to ROAS, a 40% weight to retention, and a 20% to CAC. Now, you might ask why are we giving 40, 40, 20? Well, I don't really know the answer. This is mostly a dynamic value that we create using iteration, but I'm assuming that return on ADSpnd is the very important factor. That's why it's all 40% of the contribution. Retention is very important. That's why hold 40% and CAC is somewhat least important, and that's why I'm giving it a 20%. You can flip these values as well. Generally, as a data science student, you try to adjust these weight and come up to a value which really gives you an ideal campaign score. But we are not a machine learning and data science student, so I have simply hard coded these values. Perfect. K, K two is my normalized ros. I'm giving a 40% weight to it. Then I have my L two. I'm also giving 40% weight to retention, and I'm only giving 20% weight to my customer acquisition cost. The reason I'm using normalized value is because I wanted to put all of this value on a scale of zero to one. Perfect. I hope now you are able to understand everything. Now I have given it a campaign score 0-1. So anybody so you simply look at the campaign score and you sort it by ascending order. In fact, I have one small take home assignment for you. Give me top ten campaign by this campaign score. So try to apply pivotable and calculate the campaign score and sort it by higher to lower value and give me top ten campaigns. That's one thing that you can take complete or solve use as a take home assignment. So perfect. This was our campaign evaluation to understand which campaign is generating more revenue retention or return on amount spent. Perfect. 38. Exercise 1: Marketing Campaign Performance Analysis: So, hey, everyone. Welcome to the awareness and acquisition section. This is the first exercise of this section. And in this exercise, I'm going to give you a problem statement, the dataset, and I'm going to give you an exercise objective on what all we need to calculate. Now, I'm going to attach the link for this Google Docs and this Google Sheet, and you can just open this Google Dogs and this Google Sheet, and you can start solving all of these assignments by yourself. And not just this assignment. With every single section, I have a take home assignment that you have to complete. And in between all of these exercises, maybe I'm going to skip one or two, and you have to solve that exercise all by yourself. But let me solve some exercises for you so that you have a clear understanding of what all will be learning in the course. So let me solve all the exercises. Maybe I'm going to give you a take home assignment by the end of this specific section. So let's solve exercise number one, where we have to analyze the marketing campaign performance. Now, whenever you're selling product on Internet, you would be running a lot more marketing campaign on Facebook, on Google, Tik Tok, or any other platform. The main purpose of these campaign is to drive more users on your platform. I mean, mobile app or a software product. So you run a lot of paid marketing campaign on Google, on Facebook, Tik Tok, or any other platform. You pay money on the platform, and they will show their ads, I mean, your ads to the end user. Now, whenever we run the campaign, we can also start targeting the audience by their interest. We can also start bidding keywords, and we can upload the media and creatives the way we want. So I hope you have a basic foundation on how do you build and create a ad campaign. Let's solve this exercise. So we have to analyze the performance of various marketing campaigns for a grocery app, and the main goal is to evaluate which campaign provides the best return on investment. AKA ROI. Now, this ROI is also known as ROAS, return on ad spend as well. So we have to look at impressions, clicks, sign up and purchase. In case if you don't know, impression is how many people are looking at the ad. Clicks are how many people are clicking at the ad. Sign up is when they sign up, purchase is when a transaction happens on that specific app or a platform. So let's first look at the dataset, and then I'm going to explain you each and every parameter of the data set and what's our exercise objective. So I'll open this excel sheet and let's look at the dataset. Let me zoom in a bit more. Yeah, I hope you can see the numbers clearly. So we have the dataset for campaigns that we ran on Gmail. So this is an email campaign. You have a social media campaign data, paper click data, some affiliate campaign data, and a campaign we run with influencers. And if I have to explain you each and every type of campaign. If you open Gmail, you have a promotional tab. If you want to send email campaign to a large number of people based on their interest, you can run a email campaign using Google Ads. Similarly, you can run a social media campaign using Facebook, Tik Tok, or any other popular social media platform like Snapchat. Then you have pay per click where you're paying as the number of clicks. I mean, it depends on the kind of platform you're running it on. Then you have affiliate, where you pay a affiliate commission to the people who are driving traffic or purchase on your platform. In the end, you have influence or campaign where you pay some amount to famous social media influencers on Instagram or Snapchat, so that they can talk about your product and they can give a link of your brand or product, and their audience can purchase something. So this is the amount of money you are spending on each of these campaign. I guess, this is a data for a specific time duration. Now, this are the number of impressions we got from these campaigns. For example, after spending $100, we got 50,000 impression. We got 2000 clicks and 300 sign up and 100 people purchased from the platform, and the total revenue generated was 5,000. So in our dataset, we have different types of campaign where we are spending some money. We are getting some impression, which means the number of people who are looking at the we have some number of clicks that we are getting in the ad, some sign up, some purchase, and the revenue data. So irrespective of which tool you use, mostly you get all of this data. For example, if you use Facebook Ad Manager, it gives you the total money that you have spent in a specific time duration, how many impressions, clicks and sign up you got, how many people purchased a specific product from your website or app, and what's the total revenue you generated? You can get all of this data, whether you use Facebook ad manager, Google Ad Manager, TikTok or Snapchat. Almost all ad platform gives you all of this data. In fact, they give you more data, but I've just cleaned up the dataset and make things more clear for you. So let's talk about exercise objective, and let's understand what we need to calculate in this assignment. The first thing is ROI or return on investment. Now let's say if you're spending $1,000 in an ad campaign and you're generating $5,000 in revenue. To calculate our return on investment, you're investing $1,000 and generating $5,000 in revenue. So you simply need to subtract your spending from revenue. You need to divide that by spending, and that's your ROI percentage. Then you need to calculate your click through rate, which means if 1,000 people are looking at your ads and just ten people clicking on it, so you simply need to divide ten by 1,000 and that's your click through rate. Similarly, you have to calculate sign up conversion rate and porches conversion rate. Sign up conversion rate means if 1,000 people are clicking on your ad, how many are signing up. For purchase conversion, if 1,000 people are signing up, how many are making the purchase? So in this exercise, you need to calculate ROI, CTR, sign up conversion, and purchase conversion rate. And let's try to solve the assignment. So when you closely look at the funnel, it always start with a big number at the top, and that number reduces down till the time you reach the bottom. So 50,000 people looked at your ad campaign. Out of these 50,000 people, only 2000 click on the ad, and out of these 2000 people, only 300 people signed up and only 100 people. Purchase the product. And the revenue you generated from these hundred people is just $5,000. So that's your funnel that you have from this specific campaign that you ran on email. Similarly, you have a social media, PPC, affiliate and influencer campaign. Now, the first thing you need to calculate is the ROI or return on investment. So our investment was $100 that we spent on this campaign. Let me change this into $1 value. As ROI is known as your return on investment, which means if you spent $1,000, how much revenue did you generated? So to calculate ROI, I simply need to subtract my revenue from the cost, and I need to divide this by cost, and I can multiply this by 100 to calculate ROI. And I can just prefill all of this data into other cells as well. It's perfect. So I have a 400% ROI from email, 350 from social media, 300 from a PPC campaign, and 300% from influencer. Now, generating this ROI in 2025 is super difficult, and that's why you have to be much more realistic. So I guess you need to just reduce down the revenue number or maybe increase your spending. But this is just an exercise, and the main purpose is to teach you how exactly do you calculate these metric and solve the problem. The second thing you need to solve is the click through rate so that you can understand which channel is giving you a good click through rate. So to calculate click through rate, you simply need to divide your number of clicks from your impression and you can also multiply this by 100 just to calculate this in percentage. And your click through rate is 4%, 5%, 5%, and 5%. In fact, instead of multiplying this by 100, I can also start converting these values into percentage so that it looks much more realistic. I can do the same thing with this one as well. No need to multiply this by 100 and just add it, change the formatting into Percentage. Perfect. So your ROI is 400. You click through rate is 4%. Let's also look at the sign up conversion rate, which means if 2000 people are clicking on an ad, how many of them are actually signing up? So to calculate your simple sign up conversion rate, you divide your sign up with a number of click, and that's it. You can also convert this into percentage. So your sign up so your sign up conversion rate is 15%, 20%, 15 17, and 15%. So you can see that, 2000 people clicked on our email, a email ad campaign, and 300 people signed up. Now, this data is too now, I know this data sounds too good to be true, but this is just an exercise. The main purpose is to solve the problem. Similarly, we can also calculate our purchase conversion rate, but I made a small mistake. This purchase is not in terms of dollar. This is in terms of number of purchase order. So I'll convert it back to the default value. Perfect. So 100 people purchased out of what? 300 people who signed up. So my purchase conversion rate is this much. I can also convert this into percentage. So perfect. Now we have solved the assignment. You can see that for email campaign, the RI is 400%, the CTR is four, the sign up conversion rate is 15, and the purchase conversion is 33. So you might ask me now which one is a better campaign according to the data. So if I clearly look at the ROI, the email is a better campaign. If you look at the CTR, the CTR is much better in influencer because influences the people, which we might listen a little more than, you know, other type of campaigns where we don't have much information about the brand. When you look at the sign up conversion rate, it's best in case of social media. When you look at the purchase conversion rate, it's great in case of PPC campaign. But if you simply have to look at just one number, RI is the number you should ideally look at. But again, different people, you might have to use different types of campaign to reach different audience, and it depends more on you where exactly you spend more amount of money. So that's our first exercise on awareness and marketing campaign. 39. Exercise 2: Calculating Customer Acquisition Cost (CAC): So now let's look at our workbook and let's try to solve one more exercise on customer acquisition cost. Let's first look at the problem statement, then we will look at the dataset, and then we will try to solve this exercise. And then we'll look at exercise objective and what are problems we need to solve. So in this exercise, you need to analyze the impact of discount on user acquisition and purchase for a grocery app, and our main aim is to determine how discount level influences the acquisition cost, the purchase behavior, and overall return on investment. In the last exercise, we discussed about return on investment. So I hope you have a decent understanding about it. So let's look at the dataset and then we will look at the exercise objective. So if I show you the dataset, you have five different campaign type all the way from email to social media to PPC affiliate and influencer. So these campaign types are very similar to what we have in the marketing campaign exercise. Now you have different types of discount that you're giving. So let's say if somebody is coming up from an email campaign, you're giving them a 10% discount. If somebody is coming from social media, you're giving a 15% discount on their first purchase or on their first grocery delivery order. And similarly, you're giving type of different range of discount, and then you're spending this much of money in operating these campaign. So for example, let's say you're running a email campaign, you're spending $1,000 and giving a 10% discount to all the people who are coming from these campaigns. You have impression data. Impression means how many people are looking at your ad. Clicks means how many people are clicking on your ad or your campaign. Then you have signup, the people who are purchasing and what's the revenue that they are generating. Now, let's look at the exercise objective. At first, we need to calculate our cost per signup and cost per purchase. So how much are you spending on every sign up and how much are you spending on every purchase? Then you have to calculate the revenue p P cheese. The discount ROI means the discounted revenue minus spending divided by spending. Discount ROI and ROI is very much similar. In discount ROI, you take discounted revenue. In normal ROI, you take a normal revenue, and then you need to identify the acquisition efficiency and which discount level make more sense. So you have to calculate which discount has the maximum ROI. So you need to calculate cost per sign up. That means if you're spending $1,000 on a campaign, and the number of sign ups are 350, your cost per sign up, I need to change the formatting to a number. Your cost per sign up is 2.8. Let me decrease the decimal places. Perfect. That's your cost per sign up. Let's look at the cost per per cheese. So you're spending 1,000 and the number of purchases that you're getting after spending this much of money is 120. So your cost per purchase is $8 $5.05 sent and $4.2. Perfect. Let's look at revenue per per cheese. So how much revenue are you generating from all the Pochis? So you're generating $6,000 from your 120 Po chis. And similarly, I'll click on Autopil and it autofills all the details, and my revenue per purchase is $50, 30, 25, 30, and $22 across all these different types of campaign. Now I need to calculate discounted ROI or discounted return on investment. Now in the last exercise, we saw that the formula for ROI is simply your revenue that you are generating minus spending divideed by spending into 100. Now, in this case, we are generating a total revenue of 6,000. But remember, out of the 6,000, we also need to subtract the discount that we are giving. That means I need to multiply this 6,000 multiplied by my discount, which is 10%, and I also need to subtract this further from my spending, and then I need to divide this with my total spending. So what we are doing here is that we are simply subtracting our total revenue from the spending, and then we are dividing it by spending to calculate the ROI. So let's say as a brand, if you are generating $500 in revenue just by spending $100 in an ad campaign. In that case, to calculate your ROI, you simply subtract your hundred dollar of spending from 500 and you divide that by $100 of spending and multiply it by 100. So you simply have to divide 400 by 100 and 400% is your ROI. And that's what we are doing over here as well. Total revenue was 6,000 and our total cost was $1,000. Now, out of the $6,000 in revenue, we have to calculate discounted ROI. So we are subtracting the discount. So 600 minus the discount, which is 600 multiplied by 10%, which is $60. So we have to subtract the $60 from the 600, and then obviously to calculate ROI, we are subtracting the spending and dividing it by spending to calculate the ROI. Now, this is in terms of 4.4 X if you have to just simply calculate this in terms of percentage. I'll multiply this by 100. Or the other way around is just remove this 100 and convert it into percentage, which automatically adds up 100 to it. So I can convert this into percentage. So that's ROI. So 444, one, zero, and 300. So I'm getting my maximum ROI from email campaign. So that's your customer acquisition cost exercise. Now, don't worry, if you feel like I'm giving you all of these exercises in Excel, I have a proper dedicated section where I'm going to solve these complex problem and exercise in a product analytics tool like mixed panel or amplitude or Adobe Analytics. I'm going to solve these exercises into mixed panel, but you can use whatever tool that you like. The main idea of giving you these exercises in Excel or Google Sheet is because it clears a lot more doubt and it builds a strong foundation. And that's the main purpose. Let's move to exercise number three. 40. Exercise 3: Funnel Analysis - From Impressions to Signups: So let's look at exercise number three, which is on funnel analysis. So when you look at a funnel, it goes from impression to sign up, and there are many things that happens in between. So when you run an ad campaign, you get some impression on a specific ad. Impression is basically the number of views or people who looked at the ad. Then you get clicks. People click on the ad, but they don't do any action. And in the end, you have signup and in fact, sign up after sign up, you have multiple things like poaches, retention, activation, engagement, referral. But let's just focus on these three step. Impression, click and signup. But the problem statement is that we have to perform a funnel analysis to understand the user drop off just from impression to click and to sign up. Now, our goal is to calculate the conversion rate at each step of the funnel and see which step is most effective in terms of campaign. Let's first look at the dataset, and then we will look at, you know, the funnel and what's the drop off at each step of the user journey. So I have my campaign number one, two, three, four, five, and you can just name these campaigns whatever you want. Let's say you might be running one campaign on Facebook, one on Twitter, so called X, one on Snapchat, one on Google Ads, on whatever platform. Now let's look at the impression clicks and sign up. So in each campaign, each and every platform gives you impression data, some click data and some sign up data. So let's try to calculate the conversion rate. So calculating conversion rate is super simple here. You simply need to understand which campaign is giving you more signup when you look at the impression to sine up data. For example, this is the top of the funnel and this is the bottom of the funnel. And you simply need to calculate the conversion from top to bottom, and it is going to be super simple. You simply divide your sign up with a number of impression and multiply this by 100. That's it. That's your conversion rate. So 10,000 people look at your ad and only 500 sign up. I think this is not 10,001. This is 100,000. So if 100,000 people look at your ad and only 500 people sign up, your conversion rate is 0.50. And to calculate the conversion rate for other campaign, I'll simply drag this forward, and that's your conversion rate. I'll suggest you to remove 100. So this is your conversion rate. This exercise is super simple. I don't think it has anything complicated. Let's move to our exercise number four. 41. Exercise 4: Channel Attribution Analysis: So hey, everyone. Now in this exercise, we're going to understand about channel attribution analysis. So whenever I start up acquire a customer, usually a customer has more than one touchpoint. Now, a customer might have seen an ad campaign on a TV or on a billboard, and then they will see a ad on their social media platform. So one customer can have two or three touchpoint before they actually start using a product. So in channel attribution analysis, we simply help the user understand which source is contributing most to the revenue. So let's first look at the dataset, and then we will look at the exercise objective. In Column A, you have unique user ID, starting from user ID one to user ID 30. In Column B, you have acquisition source from email to influencer campaign. And similarly, in Column C, you have revenue and Column D has touchpoint order. For example, you might can send email campaign to this customer and you can have an influencer or any other type of campaign to the same user. So if they have seen the email first, then this would be Touch 0.1. Social media would be Touch 0.2, and influencer would be Touch 0.3. So I think it all depends. Let me scroll through a little and see if you have multiple Touch points. So yeah. Now, at first, we need to calculate what is the total revenue from affiliate campaign, what's the user count, and what's the average revenue per user. Now, there are two ways to calculate it. You can simply calculate all of this using pivot table, or you can simply apply AL's formulas. Let's first calculate this using a pivot table. Now, in case if you don't know about pivot table, maybe I'll give you a small introduction. I'll simply click on Insert, pivot table, maybe to a new sheet. I'm going to delete this sheet for sure. So what is pivot table and how it is super useful if you know how to use it right. I'm going to give you a small four to five minute introduction or tuitorial on pivot table, and after that, we'll go back to our channel attribution. The primary purpose of pivot table is to simply group and summarize the data. You can group your data by different categories. You can summarize it, calculate average, count. Standard deviation. You can also apply filters and you can do a bunch of things with pivot table. Now pivot tables simply have these four options, rows, columns, values, and filters. Now rows will simply arrange your data in a row, column will simply arrange your data in a column, values will simply help you summarize the data. So whatever data you have in rows and column, you can calculate the average, you can calculate the sum and the count of that data. And filter, as the name suggests, will simply help you filter down the data that you have in rows and columns. So what do we need to calculate first? At first, we need to calculate total revenue by acquisition source, which means in rows, I will simply choose my acquision source, affiliate email, influencer PPC and social media. And I simply need to summarize it by revenue, and that's it. That's how easy Pivot table is. What's the second thing I need to do? I need to also calculate the user count. User count means the unique ID or the total number of unique IDs. I simply need to summarize my unique IDs. And these are my unique IDs by acquisition source. That's it. That's what we need to calculate here. What else do we need to calculate? We need to calculate the average revenue per user. Now, the average revenue per user means the revenue. I'll simply add revenue once again and I'll change it to Average, that's it. That's how easy it is to play around with pivot table. It's super simple, nothing complicated. I guess pivot table is something that you all should practice. You can practice pivot table with any data set. In fact, you can practice pivot table with data set number one as well. But I would give you a couple of exercises for which you can practice pivotable. Let me delete this and I'll go back to our channel attribution. Now instead of pivot table, I'm going to use functions in Google Sheet, and functions are very similar to how you use formulas in math. So if you come from programming world, I'm sure you're familiar with functions. Now functions are single keywords that simply summarizes many things inside them. So let's try applying a function. In fact, if you're not aware of functions in Google Sheet, you can simply apply sum and average functions. So I hope you have a basic understanding about Google Sheet. And I'm going to show you two simple formulas that you can apply in Google Sheet. Now, in this case, we need to calculate total revenue for all these different acquisition source. Now to calculate total revenue for each acquisition source, I can simply apply a sum A formula. Now, as the name suggests, sum if will sum the value if the condition is satisfied. So let's simply use the sum if function or the formula. I'll simply select the Sumi formula. And you don't really have to use your brain. You just have to follow the suggestion and the guide that the Google Sheet gives. It says that, Hey, first, help me select the range in which you want it to satisfy the if condition. So going to say, I wanted to select a specific acquisition source in the acquisition column. Now, what is that criteria or the source? I'm looking for affiliate as a keyword inside all of these acquisition source. And what do I have to sum? What do I have to sum if the condition is satisfied? Well, I simply need to sum my revenue. This formula says that, Hey, go to column B, look for F three or this specific acquisition source, if this is satisfied, just calculate the sum or just add up the value. Yeah, that's it. Now it is suggesting me to auto fill this and I'm going to do that, that's how you calculate total revenue for all these different acquisition source. And you can exactly do the same thing for user count. For user count, you simply need to count the number of users. Instead of sum if you will use countif which simply counts the number of users. So I'll simply select Countif and I will select the range and the value. And that's it. And you can calculate the average revenue per user by simply dividing your total revenue from the total users. And yeah, that's it. That's my average revenue per user. So this is super simple. You can do it with pivot table or you can simply apply your Sumi and countif and average formula. Now the difficult part is this advanced touchpoint analysis. Now, one user can have three touchpoint at max. So for example, if you look at user with ID number 13, PPC campaign is the third touchpoint. It's not the first touchpoint. Now the reason I've written acquisition source three times is because one acquisition source have three order of touchpoint. So affiliate can have order one, order two, order three. Similarly, email, influencer, PPC, and social media can have order one, order two or order three. I'll simply write the order and then I will simply calculate the revenue and the user count. Let me type one, two, three, and I'm going to simply copy paste the values. There are better techniques to do it, but I just use the all boring way to calculate it. Now we need to calculate revenue and user count. Now, instead of using a simple sum if formula, I'm going to use some ifs. Now, sum ifs will check two if condition. So let's apply some Is formula. Sum if is a conditional sum across a range, and sum if is a sums of a range depending on multiple criteria. So I'm going to select sum ifs, and now you can see the recommendation from Google St on how exactly this specific function work. So what is the sum range? My sum range is my C column. And then I have to choose a criteria range. The criteria range is that I'm looking for my acquisition source inside the acquisition source column. So my criteria range is B. Then I have to choose a criteria. I'm looking for affiliate inside this column. And after that, I have to choose my criteria number two. I simply need to check my touchpoint order and now, this means that my acquison source affiliate with TouchPoint order number one contributes to $235 in revenue. My affiliate with Touch Point order two contributes to $80 in revenue. Similarly, your email, influencers, PPC, and social media contributes to this different revenue number. So if I simply repeat you this formula, some Is will simply calculate sum from multiple if condition if they satisfy. Now, over here, I'm checking affiliate in the acquisition source, and I'm also looking at the touchpoint order. Since simple term, it is simply checking if this specific value exists in column B and if that value has a specific touchpoint order. And after that, if both the conditions are satisfied, you simply create a sum. In this case, we simply adds up the revenue, and that's how your sum ifs work. Simply look out for this value in B, then look at the touch point in D and simply calculate the sum of the values if both these conditions are satisfied. Now, similarly, I can also calculate count or sorry, count if as well, and this is very similar to your sum ifs work. We'll first look at the criteria range which is B in this case. Then you look at the criteria. So I first need to check the criteria which is affiliate in this case, then I need to check if this specific acquisition source has a similar touch point that we are looking after. And yeah, that's it. That's how you calculate your user count. And both of these are super simple. And now you're able to understand the total revenue by each acquisition source and the touchpoint order. You can see that your affiliate with Touch Point Order one contributes to 235 and as expected, your Touch Point order three contributes to very less, only $80 because your user is also one. Similarly, you have email with Touch 0.1, the revenue of it, influencer with Touch 0.1, PPC with Touch 0.1, and social media with Touch 0.1. So now you can look at the revenue number, the touchpoint order, and the user count. And we have already calculated the total revenue, the total user, and the average revenue per user by the acquisition source. So this was a summary and this was a drill down of the summary in case if you want to know more about it. So that's your channel attribution. Our primary objective was look at the total revenue for each acquisition source, the number of users for each acquisition source, the average revenue per user, and we were looking at the touchpoint analysis where we were looking at the revenue contributed by each touchpoint order, which is first second and third and the user count contribution, sorry user count distribution by touchpoint for each source. So great. This was all about your channel attribution analysis. In the next exercise, we'll discuss about segmentation of a new user by a different acquisition source. 42. Exercise 5: Segmenting New Users by Acquisition Source: Now I'm going to show you one more exercise and how can you solve it. And after that, I'm going to give you three assignment that you have to complete by yourself. Now in this exercise, we will look at the segmentation of new users by the acquisition source. So we have to perform advanced segmentation analysis of the new users by their acquisition source, and we have to evaluate which acquisition source contributes to more signup and purchase. We will also look at the purchase behavior of these users by simply understanding their average order value and the time they took to place their first order. Because even after sign up, people may not place their first order and they have to wait for 24 hours or one day or two day to get the first order placed. We'll first look at the dataset, and then we will look at the exercise objective. Our main objective is very similar. We have to calculate the total signup, the total purchase, the conversion rate, the average order value, and the average time to first purchase in terms of days. So let's look at the dataset. In this dataset, I have my user ID. I have my acquisition source. I have the sign up data, the purchase data, and the order value, and the time to first purchase. So if I'll just give you a simple example. In case of our user ID one, the acquisition sources email, he signed up for the product, but he never made a purchase. He never placed the order, and time to first purchase doesn't exist. Now for user ID with five, this person cames from influencer campaign. This person did sign up. He did purchase a product. The order value was $50 and the time it took for him to purchase or time to first purchase was 60s. That means it took this person six days to place the first order. And same goes with other specific data set. So for customer with customer ID 15, you have one, one, one, 52, it took him two days to place their first order. So let's start the video by calculating the total sign up. So with total signup, we simply need to add all these ones by the acquisition source. And whenever you have to add a sum by different categories, you simply apply a sum a function. Si function suggests that if condition is satisfied, you simply add up the value. So I'll simply choose this function. My range is B, my criteria is that I'm looking out for H two inside the B column, and if that condition is satisfied, simply add up the value. And that's it. That's my total number of sign up by each acquisition source. Luckily, it's 15 for all the cases. Now let's do the same thing for our total purchase. For total purchase, I can simply repeat the same formula, but this time, I need to simply change it from sine up to Purches. I'll say it's going to be B where I'm looking out for this value, and this is going to be Etch two, and I'm looking out for Purches, which is D. Yeah, that's it. My total purchase is there only for people came from a influencer campaign. Now I need to calculate the average order value. Now I need to calculate the average order value. Now, average order value can be calculated by simply dividing your total order value by the number of orders. So you have your order value in column E, and you have your purchase in column D, and I simply need to calculate the average order value. This, I'm going to combine two simple formula. I'm first going to validate if there is a purchase for a specific acquisition source or not, and if there is, I'll simply apply a sum a formula. Let me first apply if formula to check the value if I have some value, and this is equal to zero, then simply type zero. But if this is not there, then I need to calculate the sum I. Now this formula sees that if your total purchase is zero, then your average order value will also be zero. But if that's not the case, then use a different formula or function like sum if and now I'm going to calculate sum of if if the condition is satisfied. In this case, my acquisition source, which is the range, I'm looking out for this specific value in the range, and if that's the case, just simply add up the value. Because I'm calculating average order value, I now need to divide this with my total purchase, and this is total number of purchase. And yeah, that's it. So only my influencer acquisition source or acquisition campaign has a total purchase count for 20, and because it has a total purchase count of 20 and a total sine of 15, this was having average order value of 50. Let me repeat this formula in case if you're confused. So I first validated the condition that if there is a value in total purchase, then only the function will apply. So if your J four is equal to zero, in that case, just type zero. But if that's not the case, then simply use this function to simply calculate the sum of my condition. Sum F will simply calculate the sum pair if condition is satisfied. So I'm looking out for at two as my accusation source in column B. And if that's the case, then simply just add up the value. Okay, so what's the formula here? So I simply validated this first, then I calculated the average if the function is applied. Now, in average time to purchase, I'm looking for how much time will it take for a user to place their first order by different acquisition source. So by influencer, it is taking almost four days to place their first order. So I simply calculated the average time to purchase. Now, similar way, I can also calculate the conversion rate as well. Now, conversion rate can simply be calculated with your total purchase by total synu. But before that, let's first validate the condition. So if your purchase is equal to zero, then type zero and don't validate the condition. Then simply divide your purchase by total sinu. So my total purchase is zero. I need to divide this by total number of signup, and I can also multiply this by 100 just to calculate the conversion rate in terms of percentage. And, yeah, that's it. Only influencer has 133 percentage or 130 3% as conversion rate. This doesn't sound correct, but let me see if I can fix this. So total sign up did 20 purchase, I think the number of purchases are way more than the sign up. So there are some customers who did multiple purchase, and that's why this is coming more than 100. So this data is in terms of percentage or 1.33 X. So that's how you calculate all of these different metric. I know this assignment is slightly more difficult, so I advise you to practice these formulas. So if, average if some ifs and average ifs. The next assignment is going to be super simple and we'll also discuss about that. 43. Exercise 7: A/B Testing for Landing Pages: Now let's look at exercise number seven on EB testing for a landing page. Let's say on your website, you have two different landing page, landing page A and landing page B, and you're running some EB test to check which landing page is having more number of clicks, signup or in fact purchase. Now, you can create different landing pages for signup for Purches, and even for post purchase as well. Just to see which is more effective. So you can change the color of the button, you can change the content structure, and you can change anything you want in a landing page. The primary purpose is to simply check which landing page is more effective in terms of conversion and purchase when you yourself are confused. AB testing works really great with large amounts of data. If you're having thousands or even millions of people on a landing page, then AB testing works great. Now there are multiple tools that can help you AB test these different landing pages. Optimizely is one tool that is very popular. In fact, there is a tool from Google Analytics or Google as well that simply changes the HTML content or it simply allows you to route the traffic into different landing page. Long story short, Landing page will simply help you test which variant is driving more conversion and more revenue to your brand. There are plethora of tools that can help you do that. Now, in this case, let's first look at the problem statement, then we will look at the dataset and the exercise objective. The problem statement is that we are conducting AB test analysis to compare the performance of two landing page, Group A and Group B, and the goal is to identify which version of the landing page has the higher engagement and conversion. With engagement, I mean clicks, with conversion, I mean sign up. In fact, they can also add revenue as a part of the problem statement as well. In the dataset, I have my user ID. These are unique user ID that are tied to a user. I have my group A and B. I have the number of clicks, the number of sign up purchase, and time spent on a page. This is also known as your session duration in simple term. And in the end, you have your revenue. So for user ID one, group A, these people landed on a landing page. So the impression is always one, but they haven't clicked on it. They didn't sign up. They haven't made any purchase, and the revenue is zero. The time they spent on the page was 130 1 second, close to 2 minutes. Now for user ID four, they saw the landing page with group B or the variant B. They clicked on this, but they haven't made any purchase or sign up. And when you look at a data where they have made a purchase, so these people were already logged in. That's why they directly made a purchase. They didn't sign up. Or, in fact, there might be a possibility that these people never signed up and just made a purchase. Let me look at the dataset where everything has happened. Let me see if there is any short of that data. Yeah. User with ID 32 with variant A, this person clicked on it, signed up, made a purchase, and the revenue is zero. How can the revenue be zero if they have made a purchase? I think there is a fundamental flaw with the data, but let's simply still calculate all this data set. It should not happen, I think. I don't know why it's there because every time you made a purchase, there is always some revenue added to it. Maybe I think I miscalculated this data, but that's fine. Let's try to complete the assignment because the main goal is to understand the core problem statement that we are trying to solve. So at first, we need to calculate the total number of users. The simple way to do that or I think there are two ways to do it. You can use Pivot table. I recommend you to use Pivot table as well, or you can simply apply Sum A or average formulas. Let's simply apply account formula to calculate the total number of users. I need to simply check column group, and I need to check how many of A variant are there and how many of B variant are there. That's it. That's how you can calculate the total number of user with variant A and variant B or group A or group B. I think group is a better term than variant. So for group A, I have 144 user, for group B, I have 161 56 user. Now I need to calculate the number of clicks. Similarly, you can calculate the number of clicks, the number of sign up, the number of purchase, average diamond purchase, total revenue, average revenue per user, click through rate, sign up rate and purchase rate. I think I will still give this as a take home assignment because this exercise is very similar to how we have solved other assignments and exercises. So just try solving this by yourself. It will improve your foundational understanding of a concept, and you can also practice this. If you're not able to solve it, I'm going to attach the assignment solution as well. 44. Exercise 8: Evaluating the Impact of Discounts on Acquisition: Now let's look at exercise number eight. In this case, we first have to check the impact of discount on acquisition. So if you go to the data, you can see that you have different campaign source. You have the discount that is being offered. You have impression, click, sign up, purchase, revenue, and the money you spent in generating this much of revenue. So you have campaign source, the discount that you're offering for that campaign source, the impression or the people who have seen the ads, the number of people who have clicked on the ads, then sign up, then purchase, then the total revenue, and this is your acquisition cost. Now you need to calculate all of this data, and it's going to be super simple. I think we already did this. So first, you have to calculate the total spending, and you can simply calculate the total spending by the discount offered. You simply have to apply a sum formula. So you simply need to calculate the sum of this column etch by this category, or this is your criteria range. After that, you need to calculate the total revenue, which is simply your total revenue by your discount. Category, and your total impression, you simply have to create it like that in total click signer purchase, CTR, and signup. This is very much similar what we did in the last few videos. In fact, I would recommend you to use Pivot Table in this specific assignment because let me show you a glimpse of the pivot table. And how pivotable is going to make this job super simple. So I am looking for rose with the discount that is being offered. And now I need to calculate the sum by click and by revenue se Impression, it's perfect. I have my sum of click, some of sign up, and some of revenue. And you can also calculate other metric as well. In fact, you can simply calculate your CTR, which is your click through rate. You can calculate your sign up rate, your purchase rate, and ROI. So yeah, this is going to be super simple. I would really recommend you to solve this exercise on assignment all by yourself. I'm going to delete this pivot table from this. And yeah, in the end, you have one take home assignment, which clubbed all the small small assignment that we have solved so far. So in the ninth exercise, you have your take home assignment, which is a comprehensive analysis of this specific section on marketing and acquisition. So if you look at this, I have clubbed almost all the concepts that we have learned so far. So you have your tampin, let me zoom in a bit. So you have your campaign source. You have your acquisition source, you have discount offered, impression, click, sign up purchase, spending revenue, time on page and group, which is your group A and group B. These are your lending page variant. So we have combined your campaign, your discount offered, your time on page, which is time you spent on a page. In fact, one thing that I have missed is adding time to first purchase, and then I have also added your AB testing as well. So you can see that this dataset is quite good. It contains all the variables that you need, and the thing we need to solve in this is the total impression, total click, sign up, purchase, CTR, sign up rate, purchase rate, total revenue, total spending, and ROI. And yeah, this is going to be a super interesting assignment, and I'm sure you guys will learn a lot if you're able to complete this. Great. Just try completing it. I'm going to give you the solution for all of these assignment, but I still want you to complete it because if you are able to complete this take home assignment, all the concept in your acquisition and awareness will be cleared. If you're able to solve this take home assignment all by yourself, then you have clearly understood all the concept in awareness and acquisition stage, and then we can confidently move towards the next section. 45. Introduction to Market Segmentation: So Hey, everyone. Now we are starting a new module in the course and in this module, we will discuss about customer and market segmentation. Now, after completing the first section, I hope you have a really good foundation about marketing. So in this specific module, we'll first understand why segmentation is important in marketing and why it matters. After that, we'll talk about the different ways you can segment your customer like our demographic technique, our behavioral data, or psychographic data. After that, we'll talk about K mean clustering technique for segmenting your customer and then we'll talk about RFM and LTV based segmentation. Now, these are just different ways to effectively segment your customer so that you can target them better and make the marketing more personalized. After that, we'll talk about how do you create customer persona from the help of data? And in the end, I always give one case study assignment with every single module that I create just to make sure that you're not just learning the concept, but you are also solving some real world problem. So perfect. That's the overview of Module two. Let's start with our first video of this module. And whenever I start a new module, I always try to create a video Uh, like, explain me like I'm 5-year-old, just to explain all the concept and topic in the first video in oversimplified manner. 46. Segmentation Made Simple: So great. This is our first video of this section or module number two. Now, in this video, I'm going to oversimplify all the topics that we will be covering in this module. Now, the reason I create these oversimplified videos every time I start a new section is because I believe all of you are not expert in this domain, and many of you are new to marketing analytics, and that's why building a strong foundation is really important. Now, these oversimplified explanation of each and every concept make it easier for you to relate these concepts with these stories. And that's why I create these explain me like I'm five kind of videos every time I start a new section. So the primary purpose of this video is to simplify all the topic that we'll be covering in this module. So let's understand this concept with the help of an interesting story. Let's say you open a toy store where you have different kind of toys from robots to crayons to dolls, you have different variety of toys, and hundreds of kids walk every single day in your toy store. Some of them like puzzles, some like robots, while some just want these shiny dolls. But you don't exactly know which kid like what so sometimes you end up showing robot toys to a kid who just like coloring book. And that's a problem because you are showing a wrong toy to someone who doesn't really need it. So what should you do so that you do not treat every single kid the same way? And you show them the kind of toys that they like? The answer to this question is that, Hey, can we group our customer and show them the toys that they like? And that's where your first topic will come into the picture. It's called segmentation, where you group your customer the way we usually sort toys. So if you have used crayon as a kid, when we purchase them for the first time, all of them are arranged in a proper sequence. But once we start using them, obviously, the color will shuffle and we usually mix them together. And this is how your customers are. And very similar to that, this is the same feeling that store owner will have once random people start walking into the store. So that store owner has to put the customer in every single group. And this is very similar to how you arrange your crayon in a specific box. Like you put red first, then green, then blue, then white, then yellow. And just like that, you need to arrange your customer into some groups or segment as well. But the question is, how exactly do you do that? Well there are multiple techniques that a person can follow to group these different kids based on their interest. One single parameter could be simply their age. How old are they? A kid who is one to 3-year-old will obviously need different kind of toys versus a kid who is three to 5-years-old. And I'm sure you will find a good number where you can show those particular kind of toys based on their age. That's one way to group these customer. It's called as a demographic characteristic or demographic grouping. The second technique is what they buy. If you have some data on what these customer or kids have purchased in the past, you exactly know about their interest. Similarly, you can also recommend the right kind of toys. Third technique is psychographic, which is very difficult to understand, especially with kids. So instead of showing random toys to random kids, you are showing the toy that these kids actually want, and all of them are now happy. But the problem is, well, how exactly do you find these groups? Like, there are thousands of kids walking into your store every single day. How will you start grouping these kids and start showing them the toys that they actually need? I mean, it's practically impossible, right? Can we use some computer to sort them out? Maybe not for a retail store, but in a digital world where a lot of people are using your website or your app, that's where you can just personalize their experience, and you can sort them with the help of so many technique. One of them is K mean clustering. This sounds complicated, but it's not. The main idea here is that you are using a robot helper or a computer that looks at the data of all of your customer in this case, kids and look at the pattern on what they have looked in the past, what they have purchased, what kind of categories do they like, and then group them together into individual clusters. For example, there could be kids who just wanted to explore robot toys. They are picking up five, six robot toys every other week and they are purchasing one robot toy a month. That case, you can create a cluster of kids who exclusively like robots, but very few of them also like dolls and so many of them also like dolls, you can obviously create one cluster of dolls and these two clusters might overlap with each other. That is known as your key mean clustering. We'll come back to this topic. Obviously, this topic need a dedicated video. But right now I'm just oversimplifying the concept. So this robot or computer simply group these kids interest into clusters because these kids should have something in common. Consider this robot having a magic class. It can look through your customer and put them into a cluster so that it is easy for you to group them. But this one small doubt that we still have, how do we know which of these customer is actually important for us? Like which of these customer will drive revenue or will actually purchase? Let's understand how do we know which customer is the best among all the customer we have in different clusters. And to understand that, we will talk about RFM and LTV. And I'm sure you have studied about LTV as a concept in the first module, if I'm not wrong. So when you look at all of your customer, whether you run a ecommerce company, a retail shop, or anything, all of the customer are not the same. Some people buy very frequently from you in case of retail store, some people might walk into your store every single day to buy something, while some people might visit your store once in a month or once in a year. So not all customers are the same. So how do you exactly know which customer to prioritize, what to offer, and how can you increase your overall revenue? To understand that, you have a concept of RFM, recency, frequency, and monitory. You will analyze your customer on how recently have they purchased, how frequently have they purchased from you, and how much have they purchased? Do they spend a lot or do they just buy a very small thing? And if a customer have a high RFM score, that means they are purchasing more frequently from you, more recently from you, and they are spending a good amount of money. You already know about LTV. LTV stands for lifetime value, which simply means how long the customer is sticking by you and how much lifetime value they are bringing. It's just a revenue um metric that simply look at your average order value, frequency of purchase multiplied by retention duration. So great. After doing RFM analysis and LTV analysis, now we know one of the kid is our VIP. I mean, this kid is purchasing a lot. That means this kid is probably convincing his parents to walk in into the store every single day and just get him some toys. So we need to take special care of this kid because he's at the top of the leaderboard. Perfect. Now we know which customer is more important. But we still need to understand about this customer. What exactly does this customer want? We have now grouped our customer, segmented our customer know why it is so important. The next topic is profiling the customer, and that's where your customer persona come into the picture. I think this is something that you should do first, but I came to realize that when random people are walking in, you first have to look at their data and identify some clusters and then you can start profiling them based on their history. Now, in profiling, you already know the cluster in which your customer exist. You just personalize their experience more. For example, we have a cluster where kids purchase storybook, or they are just picking a storybook, looking at it, and then just dropping them off. We are looking at storybook. Now we want to know what kind of now the main purpose here is to personalize the recommendation for them so that they see the right kind of storybook that they might like. For example, Discount Disney is one marketing campaign that we can run to make sure that we are marketing the so now we have our cluster and we know which customer is most valuable for them. Next thing we'll do is create a cluster profiling, also known as creating customer persona on how do you personalize the experience for different people inside the same cluster who might have a different interest? That's where your cluster profiling and persona is very important. We'll obviously understand more about it. But once you have your customer profiling ready with you, then you have to run real campaign to make sure that you are promoting your brand, you're generating the revenue, and dropping a right messaging to people so that they convert. Because remember, the main job of a marketer is to make sure to personalize the campaign and drive the conversion. You give right discount coupons to people, right promotional images, pictures, advertisement to people based on their interest. Anything from email marketing to ad targeting to loyalty program, you show the people what they exactly need. And great. This is the oversimplified explanation of our first topic. I know the video is super long. It doesn't have to be, but I just wanted to build a strong foundation for everyone, even if you're new to marketing analytics, and that was the main purpose. Now, from the next video, let's start understanding each topic one by one. And obviously, we have so many case study assignment and exercise that you can do so that you are not just learning the concept, but you're also solving real world problem. And these kind of similar problem, you will also see once you start working working in corporate or in actual company. 47. The Four Major Types of Segmentation: So great. I hope after first video, you have a really strong foundation about customer and market segmentation. This video we understand about different ways to segment your customer and market and why they are so important in marketing. Let's understand why segmentation is important in marketing and then we'll understand about different ways to segment your customer. So imagine you start treating all of your customer the same. You are sending them same kind of email, you are sending them the similar kind of push notification, and you are just showing them the same kind of ad. The result is obvious. You will have a low return on investment. The simple reason is that if you don't segment your customer really well and personalize the experience, you don't see much engagement or even conversion. So the problem is that if you start sending same message to everyone, it will obviously leads to lower engagement and increase your acquisition cost. Same goes with product pricing, messaging. If you start showing the same kind of product to people across different geography, with a similar kind of pricing, you will not see much conversion and engagement. The simple reason is that different people across different geography, age group have different interests. So the solution is you need to divide the market into meaningful groups based on people's needs, traits, behavior, or characteristic. And then you need to personalize the messaging, target each segment with different campaign, and make sure you are able to achieve a product market fit. The main idea of segmentation is that it can help you sell better and smarter by aligning your values with the right kind of audience. So let's understand what is customer segmentation and why is it important? So customer segmentation is an idea of dividing your customer base into distinct group based on the shared characteristic or traits like demographic characteristic, behavioral characteristic, or psychographic. And we learned about the same concept in our first video as well. The main goal is that if you categorize these different customer based on their shared characteristic, you can target them better. Whatever campaigns that you're running, you will have much better return on investment. You can personalize these product and you can improve the retention of these customer on your app platform website, whatever you have because you have properly segmented your customer. The business outcome we can expect from this is that you have right persona to do marketing. You are sending different email or different kind of content in the same email for different segment, and similarly, you are pricing your product differently for different segment and geography. So let's talk about four core segmentation types. And if you come from a marketing background or marketing analytics background, I'm sure you have already studied this concept so many times. When we talk about market segmentation, you have four important key pillar demographic, psychographic, behavioral, and geographic. Let's understand about these one by one. The first one is demographic. You are simply segmenting your customer based on their age, gender, income, and marital status. The simple reason is that when you look at any platform, maybe a ecommerce company, a grocery delivery company, a food delivery company, whatever they purchase from your brand is largely dependent on their age, gender, income, and if they are married or not. And this way, obviously it's very difficult to categorize and find out this information, but there are simple patterns that you can look into, and we'll come back to that topic obviously in the coming videos. Second is geography. If you're running a brand that operate in multiple cities, even in multiple countries, in that case, you need to segment your customer based on their climate, reason, city, interest. For example, a person in San Francisco versus Miami will expect a different kind of fashion. Third one is psychographic, where you look into some lifestyle, interests, values of these customer, and then try to find out their segment or predictive behavior. In the end, you have your behavioral segmentation on what they have purchased in the past, how they have engaged with your app, and what kind of product they might end up purchasing. Now, if you carefully think about it, you can't really have a clear demographic geographic, psychographic and behavioral segmentation. Many many times a customer is often layered together across these different customer profile, and that's why understanding segmentation is important. So first, we'll talk about demographic and geography and then we'll talk about behavior and psychographic segmentation. And after that, we'll obviously study about other techniques as well. So as the name suggests, in demographic and geographic, you're simply looking for a customer profile in terms of their age, interest, value, gender, and where they stay. When you look at a insurance company, they usually plan their insurance by age group. If you're elder, obviously, chances of you getting a claim is really high, they will increase the principal amount and even the interest. Then credit card, if you have a really good job, the kind of money that you can get from your credit card will be very high, and they obviously have so many parameters to include apart from you having a good job like your age, your credit score and a bunch of other thing, but income is one of those parameter. Education is also one of those parameter. These are usually your demographic and geographic segmentation, where a credit card company, maybe a company who sells some fashion item or grocery item, look into these characteristic. Similarly, geographic. Geographic is much more relevant if let's say you have different temperature, different climate in different cities in a single country. In that case, you have to personalize your clothes, your branding, your marketing, everything for these two different customer. So that's your demographic and geographic segmentation. Then you have your behavioral segmentation, which is based on the behavior on what exactly a customer do, not just who they are. So in behavioral segmentation, you're looking at the purchase behavior of your customer in the past if they have done any purchase. You're also looking into how often are they opening your app or website? How frequently are they purchasing? How much are they browsing and which category and product are they browsing on your platform? The main idea is that you're trying to understand their purchase frequency, the categories and the product that they like. How recently have they engaged? And what device are they using? Are they using an expensive mobile phone or a cheaper device? Because that also can help you understand which kind of product can you offer them. A really good example of behavioral segmentation is cart abandonment. Anytime a lot of people who add a lot of product in their card, and if they don't check out, you usually send them an email saying that, Hey, we can see one item in your card. How about you checking out? We can give you 5% off. That's one example of behavioral segmentation. Another example is that you have a power user who purchase a lot of item from you and he's not engaging and you are launching a new sale. Then you should tell them, Hey, you have bought this much of items from us, we're giving you the best discount and the best product. Why don't you purchase? Then you have loyal customer one time discount buyers. When you look at the targeting for each customer, you can use their behavioral data and segment them really well and then you can personalize your card abandonment email or marketing email or launch email or anything in general. The next one is psychographic segmentation, where you want to understand their mindset, like their personality, their lifestyle, attitude, fitness, or social status. Now some of this can be understood with the help of their income data and the place they live. But still, psychographic segmentation is all about understanding their attitude, lifestyle, and personality trait, which is really difficult to understand, honestly. But there are still ways to understand it, if they are browsing through eco friendly product, that means they are very eco friendly. They need organic product more than in general product. Also, you can look at their trend, the kind of clothes they are purchasing, or the products that they are buying, just to understand what is their lifestyle and social status. And psychographic segmentation has very good application for some fashion brands or some ecommerce brand who is selling you clothing items. So can tailor the product, the categories, and the catalog for these people who are browsing through your platform. Perfect. That's your psychographic segmentation. Let's talk about segmentation by values. Now, apart from these, another way to segment, which is obvious is your RFM and LTV. So if a customer has recently purchased, has purchased many times and spending a lot of money, they will have a high RFM score. These are loyal customer. You should win back and you should try to cross sell the customer. On the flip side, if a customer has a higher customer lifetime value, if you don't know the formula, customer lifetime value is simply your average order value multiplied by your ticket size, multiplied by the number of month or EOs the person is with your brand. So if they have a high lifetime value or revenue contribution to you, in that case, you need to prioritize them. The main idea is that 80% of your revenue comes from 20% of your customer. That's the parato principle which applies everywhere. 80% of food delivery revenue in your country comes from 20% of your customer. 80% of fashion purchase is done by 20% of your customer. 80% of Income tax is filed by 20% of your people. Parto principle is applied almost everywhere you see. The main conclusion is that when it comes to segmentation, you need to segment your customer really well so that you can target them, drive engagement and retention, which will eventually increase your revenue. From the next video, we'll understand about RFM analysis. Also stands for recency, frequency, and monitory. 48. Introduction to RFM Analysis: So great. In the last video, we discussed on how you can segment your customer by their demographic, geographic and psychographic characteristic. In this video, we will understand how you can segment your customer by the value that they provide to your company or product. In this video, we'll talk about customer segmentation using RFM analysis. This is a really important topic when it comes to loyalty management or understanding about the value that each customer segment contributes to your brand. But before that, let's understand how does segmentation drives revenue and retention to your brand. So most of the businesses treat all their customer the same. They send same kind of generic messages to these customer, whether they are loyal customer or inactive user. And when you do that, you eventually waste your budget, and it will lead to poor results. And that's why you need to do a personalized segmentation messaging. Because without segmentation, marketing is less personal and less effective and you're simply relying on the guesswork. The solution to this problem is RFM analysis, which stands for recency, frequency, and monitory. It simply group your customer on how recently have they purchased any item from your brand, how often they are purchasing those item, and how much money they have spent. So recency means how recently have they purchased the item. Frequency means how often are they purchasing and monitoring means how much money have they spent? And this specific technique will let you treat your loyal customer at risk and inactive customer differently. And I'll give you examples so that you understand about it. But first, let's understand what exactly is RFM analysis. So RFM stands for recency frequency monitory. It's a simple behavioral base scoring method, which is used to segment your customer. Based on their purchase behavior. RFM is not just used for segmentation, it is also used for loyalty management as well. So all of your customers who have recently purchased the item, buying from you more frequently and are spending a lot of money, you want to give some loyalty program to these customer. You can make them gold member, platinum member, whatever you want. But the main idea is that based on their recency, frequency, and monetary score, you put these customer into different bucket. So recency score will simply tell you how recently a customer has made a purchase. Frequency score will tell you how frequently they have purchase, and monetary score will help you understand how much are they spending with your brand in total? Now we'll obviously understand how do you calculate a recency score, frequency score, and monetary score. But let's understand how exactly does RFM analysis work? Long story short, RFM is one of the most widely adopted effective segmentation technique which is used for customer relationship management, ecommerce, and SAS marketing. And let's understand how exactly does RFM scoring method work. So in RFM analysis, we assign a score to each customer on a scale of one, two, five based on their recency, frequency, and monitory. So we'll first sort the customer by their recency. That means all the customer who have recently purchased from their brand we'll look into all of the data, create five different bucket, and put customer equally across these five different bucket. Let's say if you have handed customer, you might find probably 2030 customer with a recency score of five people who have recently purchased from you. People who have purchased people whose last purchase was six month or one year, you might give them one recency score. Similarly, you rank your frequency or by number of purchases. All the people who have purchase more frequently or more number of times from a brand, they will get a high frequency score of five. People who have purchased just once might get a lower score, maybe one. Same goes with monetary. People who have spent a lot with a brand, they will get a higher score, people who have spent less, they will get a lower score. So that's your recency, frequency and monetary score, and it's a three digit score. When you look at a top tier customer, they will get a score of 555. That means they have recently purchased from your brand. They have more frequently purchased these item and they have spent a good amount of money. Sopla Catalis, this person is your top deal. Then you have Bob who have not recently purchased, probably purchase a month back, but is more frequently purchasing and is spending a decent amount of money. And that's why you gave Bob as a score of three, four and two, and the RFM score is 342. Same goes with Clara. Clara has not recently purchased anything. She purchased almost a year back. She's not very frequent, she rarely orders from the brand and monitor is one. You can see that these are three different customer where Alice is the top tier, most loyal customer, which is super important for the brand, and Clara is the least important. So least important customer have lower RFM score, most important customer have high RFM score. And obviously, we're going to solve one case study assignment to understand this. 49. Case Study Assignment – RFM Analysis for an E-Commerce App: So, hey, everyone. Now in this video, we're going to solve one more assignment exercise with some dataset. Now, the main purpose of this assignment is to help you understand how do you work with customer segmentation and RFM analysis? And these kind of problems are more popular in ecommerce or in retail management. So if you are looking for a product management job in ecommerce company or in a retail company, in that case, this is a really nice problem that you can solve. Now, when a company look out for product manager in a specific domain, in that case, they gives you specific problems as well to solve. So for example, the RFM and the customer segmentation is a big problem related to retention. So if a company is looking to hire a person who can work with different business team on retention and boosting the customer lifetime value, then they might give you these kind of problems to solve. So first, I'll read through the problem statement, look into different business challenges, then I'll also go through the data. And after that, I might spend five to 10 minutes explaining you the concept of this assignment, and then I'm going to solve it. And not just that, by the end of these couple of assignment solutions that I have given, I'm also going to give you a few assignment that you have to solve by yourself. So maybe by the end of these all assignment, you might find a link where you have to solve three or four assignment all by yourself, and that will really build a strong foundation. So first, I'm going to solve these four or five assignment in this section, and then I'm going to give you a 34 assignment that you can solve by yourself. So let's start with this assignment. Let's look into the problem statement first. So you are a product manager at a fast growing ecommerce company, and the marketing team has noticed that the customer retention rate is declining, and they struggled to identify which customer to target and which promotion and personalize campaigns to run. Now, as a product manager, your goal is to develop a customer segmentation strategy that helps you optimize the marketing spend and improves the customer retention. So our goals are pretty clear, and we need to identify high value customer that brings most of the revenue to the business. Now, the business challenges that the company currently sends generalized marketing campaigns to our customer and which obviously leads to poor engagement and inefficient ad spend. And the leadership wants you to implement a data driven approach to segment these customer and best way to allocate your marketing budget. So that's the core problem statement. Let's look into the data set that is given and maybe a couple of tasks that we have to complete as a part of this assignment. So in this dataset, because it's a ecommerce company, I have my sales data, so you can see you have your customer ID, which is unique. You have your transaction ID, which is also unique. Then you have your transaction date and time in the UTC format. You have your purchase amount and product category. I think this is the basic data almost every single company track anytime they have a sale on their platform. Let's look into one more data, which is your customer profile data. I will rename it as customer profile. Now in this dataset, you have your customer ID, your total order, the total spend the customer has done on the platform, and the last purchase date and the sign update. Now, customer profile is very important because it gives you a glimpse about the customer, like how many total orders a customer has placed, what's their last spin, the last purchase, and the customer sign up data. So that is important for sure. Now obviously, if you closely look at both of these dataset, there's one common value that you will find, and that's your customer ID. The customer ID in the sales data and profile data is common. And whenever you have a common dataset in both of these different data, so whenever you have a common column or a common thing in both the dataset, in that case, you can combine them as well by simply using Lookup. Now, we should do that ideally, but we can also directly refer to that specific data set inside though function itself. So you don't really have to use Vlookup, but it's good to just put all the data set in a single sheet so that it is easy for anyone to understand. So perfect. Et's go through the task first because I'm sure you fairly understand the data description really well. Now, our goal is to analyze customer behavior to build a structured segmentation model using recency, frequency, and monetary analysis. Now, this exercise is specially tailored towards RFM analysis, also known as your recency frequency and monetary analysis. So before we go into the task, let me help you understand what exactly RFM analysis is and why is it used by companies. And I have not prepared any fancy PPT, I'm just going to Google it out and show it to you. So first thing first, what is a RFM analysis? Now, RFM analysis is a concept that is used for customer segmentation and retention. Now, it's a very effective data driven method to segment customer, and it uses three important factors recency, frequency, and monetary. Recency will help you understand how recently a customer has made a purchase. So more recent buyers are generally more engaged and they are likely to make more purchase because your brain usually forget a brands or a purchase over time. Second factor is frequency, which shows how often a customer has made a purchase in a given time frame. Frequent buyers usually indicates a strong relationship with the brand. In the end, you have monetary, and this is the total amount a customer has spent with your brand overtime, and usually a high spender contributes more to your revenue, and they are also considered more valuable as a customer. That's your RFM analysis. So recency shows just for an example, in recency, you need the time since they last ordered a specific product. In frequency, you care about the total number of transaction done by a single customer, and in monetary, you care about the total or the average transaction value. And you can see that we have all the data set. We have total number of orders, which is your frequency, we have the total spend, which is your monetary, and you have your recency data as well. So now what do we need to analyze in this dataset? We need to first analyze the customer transaction behavior. Then we need to develop RFN segmentation model and we need to assign it a score, and we'll understand about the recency frequency and monetary score in a minute. So first, we need to analyze the customer transaction behavior. Then we have to develop RFM segmentation model by assigning a score. I'll help you understand what do you mean by a recency, frequency, and a monetary score. After that, we need to classify customer into meaningful segments like your high value customer, loyal customer at risk and lost. After that, we will simply extend the segmentation model and understand about product categorization, customer tenure, seasonality or periodic purchase trend. We'll also look into customer lifetime value and some actionable business recommendation. So these are the five, six tasks that we have to complete as a part of the assignment. But before that, let's understand about how do you create a RFM segmentation model and assign a score, and what exactly is it? So to understand how strong relationship we can have with our customer, each customer is scored based on these three factors, recency, frequency, and monetary, and these are typically on a scale of one to five, but you can also use a scale of one to ten as well. Where if a high value is indicated for a particular factor, it indicates a better engagement. For example, let's say a customer one has recently purchased from your brand, have very high frequency and has spent a large amount of money. This customer is way more valuable than, let's say, customer five, who has spent a very small amount of money, has very less frequent purchase, and it's been a long time since he have purchased a product. So we will give a recency score, a frequency score and a monetary score 1-5 for all of these customer, and then we will simply put them into different buckets. And these are those buckets. This one. So a customer who is recently active and has purchased a large amount of product from your brand and has done a frequent purchase. These are your loyal customer. On the flip side, you might have customer that are not recently active, have not done a lot of purchase, and are not frequent buyer as well. These could be the customer that might join off, and you need to prevent that from happening. Now, let's understand why RFM analysis is important and why are we solving the assignment at the first place? The first thing that RFM analysis does, it helps you enhance your customer segmentation. Be RFM analysis will simply allow businesses to classify customer into different buckets like VIPs customer at risk and one time buyer, and this enable your personalized marketing strategy tailored towards each customer segment. Also, once you have identified the customer segment, it is easier to focus on a segment and improve the customer retention. Not just that, you can optimize your marketing campaigns because now you understand which are your VIPs customer, which are your high risk customer and one time buyer, and you can focus all your ad spend on optimizing for these high value customer. For example, all the frequent shopper may receive an exclusive discount because they have already spent enough on your platform and they do want to get early access to your product. Also when you have a proper view on the segmentation, you can maximize revenue and your marketing campaign and increase the overall ROI. For example, you can see this red color segment. These are your frequent buyer with high monetary purchase and they are recently active. To these customer, you can give them an offer like 20% off on your next purchase. For these people who have been active, but it's been some time and has a lower frequently purchased, then you have these yellow segment or these are all customers who were active a few weeks back or a few months back, and they have spent a decently good amount of money and they are not super frequent on purchasing from your platform, you can give them some membership so that you can transition the customer to become a high value customer. Then you have your fresh leads or users that did one time purchase and they were they were active long back. That's how you can optimize each and every customer segment, and obviously, you can use RFM analysis to reach to these customer segment. That's how you can optimize your marketing and your ad campaign for all these different customer segment and the best way to identify the customer segment is by looking into RFM analysis. Let's try solving the assignment. So at first, we will calculate the recency frequency and monitory. Then we will give it a score. Then we will calculate the RFM score, and based on the RFM score, we'll start creating customer segmentation, we'll understand about average order value, the purchase frequency, and the customer lifetime value. Perfect. Let's start with recency. We want to know how recently this customer has purchased from our brand. The best way to do that is to simply calculate the time difference between their last purchase date and today's date. Now, this data is obviously old, so you can use February 2024 as your today's date. But you can even use today's data as well because anyway, we are calculating a percentile where we are allocating a specific number based on the overall dataset. So it doesn't really matter whether you take today's date or you take the date at which we downloaded or exported this dataset. So let's say you're taking today's date and you're simply subtracting D two from it. So it's been for 73 days since this customer has purchased it. Now, this data looks old, so maybe the better approach could be to simply subtract February 2024 instead of subtracting it from today's date because the data was downloaded in February 2024. So instead of today, I'm going to replace this with 24 Fab 2024. Right now we are in 24 Fab 2025. So this data is almost a year old. Perfect. Now you have your recency data. As you can see, this customer was active 107 days back. This customer was active 215 days back. I think most of this data is all by three months. Maybe if I look into the lowest difference I can find, then I will exactly know when did I exported the data. So maybe I can sort this from A to Z. Since this data is close to two months before even 2024. So maybe instead of 24, I can put this as maybe 20th of December 2023. Let's see. Okay. I think this is wrong because I'm getting a negative value, so maybe January 5, 2024. Perfect. I think this seems fine. Because I'm not able to see any negative value. Perfect. I think this data looks good now because you can see that the least number of Ds are five. Perfect. I think we can take 2024, 1 January as today's date because this data is old. Then we need to calculate frequency. I think frequency data is already given. Frequency is your total number of purchase, and I think you can directly just refer to the B two cell. Monetary is also given, so I think we don't really have to calculate. We can just directly refer to C column. Perfect. Now you need to calculate a score 1-5, both for recency, frequency, and monetary. And the best way to do that is by simply using a percentile method. Now in this dataset, we are simply calculating the percentile and giving them a score. For example, within 20%, you give them a five recency score. Within 40%, you give them a four recency score within 633, within 82, and after 81 recency score. That's a very easy way to distribute your whole dataset into five different bucket. The formula for this is also very simple. You simply need to apply this formula where it is taking up the recency value and simply looking into the percentile and giving them a score. 0.2 less than 0.2 will have a five score, less than 40 will have a score of 40, less than 60 will have a score of three and similarly, you can see that. I have now given it a recency score of one because this customer was recently active. Now obviously, 165 is not very recent, so that's why score of two. 241 is obviously very late, that's why we have given it a score of five. Now, there seems to be something off with this dataset. A high recency score means the customer was recently active, while a low recency score means it's been a long time since the customer was active. You can see that 57 means the customer was active two months back, I've given it a high recency score, while 245 or 433, you can see that I've given one recency score because 433 is a long time. Similarly, for 245, I've given two recency score. Obviously, you can further optimize it by using a scale of ten instead of using a scale of five because you can see that even a customer that was active 57 days back, even to that customer, I've given a high recency score. Maybe this is something that I can correct. Because five would be still 40, 60s, ideally five frecuency score should be given to the customer who was active within last one or two week. So that's the simple problem with the dataset if you want to optimize it further. But let's focus on the approach more on the perfection. Then you have zero frequency. And you can use almost the same formula to calculate the frequency score sewll. So if they have done more purchase, they will get a high frequency score. If they have done less purchase, they will get a low frequency score and will use the same formula to calculate the frequency score as well. So perfect. For this customer, they have done a many purchase like almost eight purchases. They will get a high frequency score. This customer just has done just one purchase, so they will get a low frequency score. Similarly, we can calculate a monetary score as well. So this customer has purchased a good amount purchased in a good. So this customer has overall spent a good amount of money with the brand. That's why you have high monetary score. This customer has spent very less amount of money. That's why a low monetary score. And the easiest way to calculate RFM score is to simply add up all the values, your recency, your frequency, and your monetary. That means we are now adding all the three important factor that is super helpful in optimizing our marketing spend or our campaigns so that we can increase the overall revenue. Perfect. Now, obviously, based on RFM score, you can dynamically create segmentation, but I'm not a Python engineer to dynamically fetch the values, figure out three equal split, and then give a customer segmentation. In this one for simplicity, I'm going to simply hardcore ideally you should create equal buckets and, you know, you should let the system decide in which bracket they want to move this customer. So if your RFM score is more than 13, this is a high value customer. That means we have to give more exclusive discount. We have to give them special access or beta access because these are the people that are recently active. They have purchased from our brand multiple times, and they have also contributed to a good amount of revenue. Then RFM score more than ten is our loyal customer. Less than seven is at risk. That means they might turn, and below seven are all those customers who are kind of already lost because they were active long back. They haven't made any purchase. It was just a one time purchase, and they contributed a very less amount to the overall revenue. So perfect. So this is your customer segmentation. You have four different segment of customer, high value, loyal, at risk, and lost. And obviously, there are multiple ways to make more money from this and at least it do one more purchase because you can give them a 50% discount and they will do one more purchase. To this customer, you have to push them to buy a membership, or you have to push them to buy more exclusive high end product. Perfect. Now we will calculate the average order value, which is easiest to calculate. You have your total value and your number of orders, and that's your average order value. I think purchase frequency are also very easy to calculate. You have to divide your frequency with the number of days in a month. Your frequency is what? G two, today's date minus E two, which is your sign up date divided by 30. I think instead of today's date, I will still take February, sorry, January 2024 as today's date because the dataset is Okay, so this is our percentage frequency. Perfect. So this is how frequently these customer are purchasing in a month. So 0.42 means they are purchasing once in two months. So this is your frequency roughly. I think you can also multiply this by 100 if you want to calculate the frequency in percentage, but that won't be a fair way to look at the frequency. Maybe you can calculate a quarterly frequency instead of month. But I think customer lifetime value is something that is more useful here. You want to calculate your customer lifetime value. You need to multiply your average auto value, multiply it by your frequency and the retention rate. I don't have the retention read data, but let's say the retention is close to 12 months, that's your customer lifetime value. Perfect. Now, now you have all of the data with you. You have your recency score, frequency score, monetary score, and you have customer segmentation. You can anyway do a pivot table and understand more about the data. That's something you can do it yourself. But I simply have to show you maybe I can see the multiple customer segment and how many customers are there in high value, low value, medium value. The column name is customer segment. So customer segment, I want to see the value in rows, I need to see. So there are 126 customer at high risk, 144 are loyal, 107 are lost, and 112 are high value customer. Maybe I can see their contribution to revenue as well. So now you can see that loyal customer, high value customer, and loyal customer. Oh, this is a very unique and sights. The loyal customer are more, but when it comes to total revenue, high value customer has contributed to more total revenue. Just like this, you can obviously see multiple things and, you know, complete the assignment. So we have developed a RFM model, created the customer segment, extended the model. I think we haven't done this yet. I think this is something that we can park for some other assignment because you need a few more variables and dataset in this. We predicted the customer lifetime value and obviously you can write the recommendation as well. Perfect. You can look at the solution as well if you get stuck anywhere else, maybe the formula could be slightly different. Yeah, that's your assignment to your RFM analysis. 50. Interpreting & Summarising RFM Results: So now let's understand about RFM and a real world use case in ecommerce company. So let's say you did a RFM score analysis and you create a different customer segment, and you have 10,000 customer that you're able to segment into five different RFM bands. And your main purpose is that you actually want to improve retention, increase revenue, or customer lifetime value. So let's look at a couple of tempin example that you can run across these different customer segment. So let's say you have a customer segment that you named as champions and the RFM score for this customer segment is 555. That means these are the customer who are recently use your product. They are most active frequently buying from your brand and are spending a good amount of money. So to these customer, you can create a VIP offer campaign and you can give them early access because these customers are most important for your brand. Then you have at risk RFM segment 200-300 range. Now, these are the customer who haven't purchased anything for quite some time and you really want to activate them. To these customer, you can send them, we Miss you notification or email campaign, or you can create some urgency offer like, Hey, limited time period offer, the deal expires or the offer expires in 23 hours. In fact, some brands also gives a reverse countdown in the email saying that 50% off on all popular brand and the deal expires in 2 hours. Third example is all the customer who are new to your product, are recently active, have purchased a lot, but haven't spent good amount of money. So to these customer, you can cross sell, upsell, tell them about new features that you have in your product. And in the end, you have your lost customer segment. Who have RFM score of 100-150 and these have mostly unsubscribed from your product. They don't use it anymore. They are dormant. You can use some revival campaign. Now, the main idea is that in ecommerce company, when you segment your 10,000 users into these five different RFM band, you will have really good result. Things like you have 28% higher click through rate across all these email campaign because you're sending different email campaign to different user segment. Also you might end up seeing a revenue increase of 12% with the email and push notification that you send. In fact, if you look at all the push notification sent by a food delivery company or by a grocery company, these are super personalized push notification. One more result could be that after doing a proper segmentation, you might see something like 2.1 X, higher retention rate, and you might able to revive some of these lost customer as well. But long story short, when you do a RFMB segmentation, you will see a higher conversion, a higher revenue, and much better engagement. That's the main purpose of segmenting your customer. 51. Clustering in Segmentation – The Basics: So great. In the last video, we discussed about RFM B segmentation, and that's a really good technique to create multiple customer segment based on the value that they contribute to our brand or to our product. In this video, we'll talk about K mean clustering. Or segmentation. Let's understand what it is and why exactly do we need it? Initially we start segmenting our customer by behavior, demographic and psychographic characteristic. That was enough. But then we suddenly realize that, we also have to create our customer segment by the value that they provide to our brand or to our product. When I say value, it means recency, frequency, and monitory. How recently are they purchasing? How frequently are they purchasing and how much money are they spending. Both of these three things, all these three things actually delivers value to us. Like it brings money on the table. Why do we need K mean clustering? Now, before we understand about K mean clustering, I wanted to revise a couple of concepts that you might be already aware of, or you just want to understand these concepts first before you go into K mean clustering. And this concept comes under machine learning. We'll obviously understand where exactly K mean clustering stand. This is over here, but let's understand from the top. So let me explain what machine learning is for a minute. Machine learning is when your system learns from the data, when the machine learn by itself. That's the oversimplified definition of machine learning where you feed a lot of data to a machine and the machine will start looking at the pattern and start learning it by itself. That's your machine learning. Now when we talk about machine learning, we have three broad categories. We have supervised learning, reinforcement learning, and unsupervised learning. As the name suggests, supervised learning means that the machine is learning from the labeled data. Let's say if you give machine data like this is the temperature, this is the weather condition, this is the country, this is the wind flow, and please help me predict whether it will rain tomorrow or not. That's your supervised learning. You're giving some labeled data to a machine and you're just asking a simple question, either tell me it will rain tomorrow or not, which is a classification, yes or no, or predict a weather for tomorrow. That's your regression. So classification means you're simply asking the machine yes or no or a binary value and regression means you're asking machine to predict the future outcome. That's your supervised learning. Then you have reinforcement learning when you are giving a reward and a penalty to a machine. Every time machine makes a mistake, a person manually come into the picture, give it a penalty. Anytime a machine take a right decision, you give it a reward and the machine learn it by reinforcing the algorithm. That's your reinforcement learning. Then you have unsupervised learning. Where you have not given any labeled dataset and just machine is trying to find a pattern from some random data point. Let's say you have given some thousand data point to machine, and you're just asking machine, Hey, these are 1,000 data point, these are the million customer. Just try to put these million customer in some random segmentation based on the thousand data point that you're collecting. That's your unsupervised learning. In this one, you have two different techniques Association and clustering. Association is finding relationship between different item, and clustering is grouping the customer or data point with a similar characteristic. And you obviously have different type of clustering. You have hierarchal clustering, overlapping, exclusive and KMI. Now, we'll obviously discuss about these three in some other video. Our main purpose is to create different customer segment using K mean clustering. And we'll obviously solve a case study assignment. I'll give you some dataset and we'll understand about this concept properly. But let me give you a bookish definition or knowledge of what is Kaman clustering and why exactly do we need it? Like, we have a RFM analysis. That's great. Why do we need mean clustering? 52. K-Means Clustering Explained: Now before we understand more about me in clustering, let's understand why do we need came in clustering when we already have RFM scoring segmentation. To understand this, I'm going to tell you a small story or maybe oversimplified example. Imagine you are running a shopping app. Where people can purchase product online and you have two user on your platform, User A and User B. Now, user A buys every few months and spend close to $500. Now, User B opens the app daily, browse for almost 20 minutes, add a couple of things into the card, but he hasn't bought anything yet. So these are the two customers that you have on your platform. Now, our first way to create segmentation is to simply look at their demographic or geographic characteristic. Now, both your user A and B have the same age of 30 year, their gender is women, they are from London. You can't really create different customer segment for user A and user B. Then you will say, perfect now we can move towards RFM based customer segmentation, where we will look at customer A and customer B. You customer A is opening up the app once in every few months and spending $500. The customer is not very active, but he's spending some money. But when you look at customer B, this customer is opening up the app every single day. That means the recency is high, and this customer is also interacting. That means he's spending close to 20 minutes, but he doesn't have a frequency and monitory. If you calculate the RFM score of A and B, you might feel that the RFM score is also close to equal. How do you create proper effective segmentation? Because something feels off to me. When you look at User B, User B is clearly engaging with the product, but the user B is not getting the right offer to check out. And the problem is that RFM score only sees the transaction, the recency, and the frequency. RFM based segmentation doesn't see the behavior pattern, and that's where clustering comes into the picture. Clustering goes beyond these limit. It looks at the real behavior, like how frequently the customer is browsing through the platform. How much time are they spending? Have they added something into the card? Have they abandoned from the card? Can we lift these users and make them purchase from the platform? Now there are multiple techniques to do clustering. One of the technique is known as your K mean clustering, and that's what we will learn in this video. K mean clustering is an unsupervised machine learning method that groups your customer into K clusters based on their similarity. Customer who are in the same group who behaves or interact like each other, they are into the same cluster. And after you have run Kaman clustering on your dataset, your clusters will look something like this. I mean, this is too good of a diagram. The actual clusters look pretty bad. But the main idea is that you will have all these different cluster and at the center of each cluster, you have a centroid and from centroid to your data, you have your distance. All these clusters have a similar data point and these clusters are created based on their natural pattern, such as maybe these customers are spending a similar amount of money, maybe they are having a same behavior where they are just adding a product to the card and dropping off. K mean clustering basically create clusters based on the user behavior, pattern, habit, and usage. In K mean clustering, you create exclusive cluster. That means one customer can belong to one cluster at a time, where K stands for the number of groups that you want to create or number of clusters you want to create. And the main idea is that K mean clustering, group your customer based on their similar behavior or trait. This behavior is related to maybe interaction with the app, the spending behavior, the purchase behavior, the duration between adding a product to the card to checking out, you can have N number of combination. Now, we'll obviously solve a real world case study assignment on K mean clustering. But if I have to give you a high level overview on how exactly does it work, you first input the number of clusters that you want to create then you calculate the distance and you try to group based on the minimum distance from the centroid. If your clusters are very stable, in that case, you stop the process. If they are not stable, you recalculate the distance and try to minimize the distance from centroid. This sounds complicated, but it's not. Once I run Kaman clustering, this thing will become super simple for you. But the variable that you use in K mean clustering are these one, and I'm sure you can have thousands of variable that you can use. The first one is RFM score. All the customers who have recently purchased, more frequently purchased, have spent a good amount of money will have a different kind of RFM score. So RFM score can act as a input data in the Kmean clustering. Then you have customer lifetime value, which is a metric, this can also act as a source. Customers who have a similar kind of CLTV or customer lifetime value can or cannot go into a same cluster. Then you have your average session duration, the number of pages a customer has visited, the click behavior, the sign up source, the device pattern. You can have thousands of different data points and came in the came in clustering, it will go through all those thousand data points, see which all customer comes into the same kind of pattern or grouping, and then create these clusters. And we'll obviously talk more about K mean clustering after solving some real world K study assignment or after solving some real world problem. Then you will understand more about this specific concept. 53. Introduction to Predictive Analytics: So hey, guys, so far we have covered three module in this course, and now we are starting our fourth module on predictive analytics and customer lifetime value modeling. Now, in the first three module or in the first three section, we discussed a lot about marketing Analytics foundation, customer segmentation and profiling. And in the last section, we discussed about funnel analytics and conversion optimization. Now in this section, we'll discuss about how do you predict a future sales or a future behavior on anything, and how do you make sense out of customer lifetime value? So in this section, we'll first discuss about predictive modeling. Then we will discuss about linear regression, logistic regression, which are the types of predictive modeling. After that, we'll understand about John prediction, customer lifetime, and then in the end, we'll discuss about retention and acquisition cost. So this section is going to be a lot about predicting the future and understanding the customer lifetime value. So let's start with our first video where I will try to oversimplify all the topics and concept we'll be covering in this specific section. 54. Predictive Analytics Made Simple: So this is the first video of this section and I want to make sure that all of you have a strong foundation before we start understanding about all these concepts. That's why in the first video of this section, I'm going to oversimplify and explain all the topics that we will be covering in a super simple language or in layman terms. This will help you build a strong foundation if you have no idea about predictive analytics or about customer lifetime value modeling. That's why I usually tell you a small story and explain all of these concepts, and then I will actually create proper videos for each and every single topic. In this video, I'm going to oversimplify all the concepts to you. Let's first understand what is predictive analytics. I'm going to tell you a small story to explain predictive analytics and a bunch of other concepts. So let's say Anika has a lot of customer buying t shirts from her online. She is having an online store. Now, she's perfectly running the online store where she normally sell t shirt to different people on different events, occasions during the year. But Anika usually wonder, how do I know which user will come back and buy the t shirt again and which customer might stop buying in the future. She has little to no clarity or visibility on any of these questions. So she started talking to people, and some of those people said that, Hey, you need to use some tool or some product to understand predictive analytics, which will simply tell you what will happen in the future for every single customer. So she started talking to multiple people, and those people suggested her that you should use some predictive analytics tool that will help you understand insight about each customer, whether they will buy again from you or not, and how many of them may not purchase t shirts from you and might churn. This tool works by using clues or past data to guess what will happen in the future. And predictive analytics works exactly like prediction of weather where it is almost accurate, but it can go wrong as well. But in this case, it works for customer. So these kind of tool can help Anika plan better, keep customer longer, and sell smarter. Now Anika exactly know what her future sale is going to look like for the next month or for the next quarter. But she want to know about every single customer, whether someone will buy the t shirt again from her or not. To understand that, we'll talk about regression model and how these regression models can help her make better predictions. Anka started learning about two regression model or two regression tool. The first tool is linear regression and the second tool is logistic regression. Now, linear regression will simply help her understand how much a customer will spend in the next purchase or in the whole lifetime. While the logistic regression will help her predict if the customer will return back and purchase are going from the platform or from the website or not. So how likely they will purchase from her ecommerce store. Now to make sure these two model or regression model works really well, she have to feed a lot of data to the regression model so that they can predict the result better. So she started feeding the past purchase data, all the website analytics data, things like how many users came on the website, how much time they have spent, things like session duration, session time, all of that data. And then she's expecting some results from these tools or these models. And after using these tools, he saw that, Hey, linear regression is able to tell me that a specific person will spend $60 next month, and this person might not return. Again, this is just a prediction. It will turn out to be true or false. We don't know yet. But you might wonder, well, how exactly these tools work? And obviously, I'll explain that to you in a while. But she is more concerned about the second part. This person may not return. I'm sure there are more people who might join in the future. So how does Anika prevent this from happening? And you want to know how many of these users might leave or might not come back again. So to solve the problem, Anika started learning about churn model. And these prediction model will also help you understand which user will churn and which one will not. Now, all of this started when Anika started noticing that many people who purchase from the platform are not purchasing anymore, like many people stop buying from her. In simple term, this is known as your churn. So now she's focusing on building a churn prediction model, which will simply tell her which customer may not purchase in the upcoming month or year. And she's feeding a lot of data to this one model like your past sales data, your website analytics data, and if the customer has opened any promotional emails or not, and how many visit and how many of those customers have visited her platform. After feeding all of this data, I'm sure this data is very less. I'm going to give you an exhaustive list of data that you have to feed to all these models for them to work. After feeding all of that data, she is able to figure out that, hey, 20 customer might leave the platform soon. The advantage of this is that once you know that these customer may not come back, you can literally give them some discount or maybe send them a thank you note, just to bring them back or just to engage with the customer. Now, there is one more problem here. Apart from Chun, she also need to make sure whatever customer she has, they also need to buy back from her. So she's also curious in knowing which customer are worth keeping the most for my brand. To understand that, she started learning about customer lifetime value. We have covered customer lifetime value in the last section as well. Now, the main idea here is that Anika doesn't want it to spend too much money trying to keep every single customer. So she started using customer lifetime value model to help her predict how much each person will spend in the whole lifetime. So as soon as she started acquiring one customer and the customer has done some purchase, the model will tell her that this customer might spend close to $300 in the complete lifetime, while another customer will might purchase once. Now obviously, once you understand customer lifetime value, then you also need to understand how much money can you spend to keep these kind of customer. Now to do that, she has two choice either to retain existing customer and acquire new customer, and she can do both of these things side by side. So Anika wants to compare the retention and acquisition. She wants to compare how much cost she has to incur to retain the existing customer and what will be their customer lifetime value in the future. For acquisition, she wants to know how much will it take for her to acquire a new customer. And obviously, she knows that finding a new customer is much more difficult than retaining existing customer because existing customer can bring more revenue, you can easily upsell and cross sell to that person. So obviously so technically, she should spend. So she started spending more time retaining the existing customer, but very less effort in acquiring new customer. In the end, she has to put all of these data together things like linear regression where she needs to predict how much sales she will have in the next season, next quarter or next month, or logistic regression where she wants to know if a particular customer will churn or not or come back or not. Then she also has customer lifetime value data, retention data, churn data. So she want to make a super simple model that can give her a holistic view of all these different data point and she can run the business properly. So obviously we'll create a spreadsheet for her where she can look at all of this data and make more sense out of it. We obviously need to feed all of this data into the model so that it can start predicting the right things for her so that she can run the business on autopilot, which obviously means we have to interrupt or she has to come in and just check a few things if they are going in the right direction or not. Great. This is our first video. From the next video, we'll go deeper into every single concept that we have covered. Things like predictive analytics, logistic regression, linear regression, customer lifetime value churn model. We'll cover all of these concept one by one. 55. Applications of Predictive Analytics in Marketing: So great. In this video, we'll talk about predictive modeling and why is it important in marketing. Now in modern marketing, you want to know what will happen in the future the understanding the past. Anika who is running online ecommerce store, she wants to know who might leave, who is ready to buy again, and what should she focus on. So in predictive analytics, we simply use past data like your purchase data, your website analytics data to forecast the future behavior. This forecasting could be right or wrong. It depends how much data do you feed in and what's the noise, and we'll understand what exactly this noise is. Main idea is that all these predictive model will help Anika understand so that she can send the right offer to the right people. She can stop wasting money on those things which might not bring her the revenue, and she can keep her best performing customer so that they can give her more money over the lifetime. And not just online store, a real example of predictive modeling could be a grocery brand that is selling groceries to different customer every single day, and this grocery brand wants to know the jon rate, the retention rate, and how much revenue this customer will bring to her. So what is a predictive modeling? In predictive modeling, you simply use datas and algorithm to predict what your customer are likely to do next. And this is not just about customer, whether you want to predict whether, sales, marketing spent, revenue, anything. Predictive modeling is used in almost every single domain that you see. And it works on the top of statistical and machine learning techniques. And if you come from a machine learning or a data science background, I'm sure you have worked a lot in predictive modeling. Now, in this case, we are talking about marketing. So if I give you a couple of use cases that you can solve with predictive modeling in marketing, well churn analysis is one where predictive analytics will simply help you understand which customer will likely churn or not. You will know how many leads you will get in the future based on the past data, how much money a customer will spend based on the behavior of the customer because you do get the browsing data, the session time, the session duration in the app or website, and what product you should recommend next to these customer. The technique that is used is known as your lift or market basket analysis. Not go there, but predictive modeling can help you predict or guess a lot of thing. Not guess, maybe predict a lot of things. Now to run a predictive modeling, you have a small workflow, which is not very important to learn. Obviously, we're going to use the workflow whenever we solve a real world case study assignment that we obviously do by the end of almost all the videos. So every predictive modeling workflow, follow a series of steps that you have to follow to make sure it works really well. The first step is defining the outcome that you want to predict. Do you want to predict churn? Do you want to predict revenue or customer lifetime value? You first have to define your goal or the outcome you want from a model. The second thing is that you need to collect and prepare the data where you might need to clean some data, you might need to structure and label this data. Now when you want to predict something like churn or revenue or lifetime value, in that case, you need a lot of data. For example, you obviously need past sales data of all your product. You need a website analytics data. Things like how many people landed on your homepage, how many of those people bounced. What was their session time, session duration, click through rate. You need all of the data from your website and from ad campaign as well. Let's say if you're running ad campaign on Google, Facebook, YouTube, whatever. Third thing is that you need to choose the model. In regression, you have linear regression and logistic regression. Linear regression will simply help you understand the future sales or future forecast, while logistic regression will give you a probability, whether a certain event will happen or not. We'll come back to both of these topic, but you need to choose the model. Do you want to forecast the future or do you want a yes and no in the future? So you need to decide whether you need a number to forecast in the future, or you need state, yes and no. Then you need to train the model. You need to fit your training dataset so that the model start predicting better. And then you need to evaluate the performance, run it on small dataset and see if it works really well or not. Now, as a marketer, you'd never have to do any of these things because most of the model or software product work out of the box. But I'm still explaining you all of these things because they are important for you to understand on how things are working under the hood. 56. Understanding Linear Regression: Now Anika, who runs online ecommerce store, she is quite sold on the idea of predictive modeling, but she's still confused. On which model to use. Now she understand that using predictive modeling, she will be able to figure out which customer will churn and how much a customer will purchase in the future. But she's still confused. How do I use this model and which model should I use? To help her and to simplify everything, she has two options and obviously we'll be covering both of these two topic in the next few videos, linear regression and logistic regression. Linear regression, which is a predictive model will help her answer how much a customer will spend in the future, and logistic regression will give her a yes and no on if a customer will purchase in the future or if the customer will churn. Let's understand about linear regression first. Let me give you oversimplified example for linear regression and then we'll come back to this story. So let's understand about linear regression using a coffee shop. So let's say you're running a small coffee shop. Every day you try different things. You post on Instagram, you offer some discount to these customers, and you also change the music. Now, all of these things obviously drive your sales and you see increase in your sales, but you don't exactly know what is the contributing factor of the seals. Now techniques like regression will help you answer which variable is driving my sales? If a specific user will come back or not, and what makes the biggest difference in the user's coffee experience? This is like turning a messy data decision into a very statistical one or data driven one. Now, in regression, you have two method, linear regression and logistic regression. This video is going to be just about linear regression. Let's talk about linear regression first. The simple formula for linear regression is this one. Y is equal to Theta one plus Theta two X. You can also write this formula as Y, which is a variable that you want to predict. The simple formula for regression is Y is equal to one X one plus two x two plus B. Now, one X one and X two are there because you have multiple variables. So when you look at your sales data, it is dependent on ad spend. It is dependent on the price of your product. And then there are some random variable like season, timing, context. We call it as noise. In this case, you want to predict sales from the ad spend data. This is a simple linear regression. If you have multiple variable, it will become a multiple linear regression. Let's stick with simple linear regression which only have a single variable. So in this case, we want to predict sales from add spend data, and we assume that other factors don't really matters. So in this case, is your slop. Which shows you how strongly your ad spend will affect the sales. B is your baseline. That means even if you spend $0 on paid marketing, what is your minimum sales? So even if your ad spend becomes zero, so this part will obviously become zero. In that case, what will be your ad spend? So if this becomes zero, obviously you still have this. So if you're not spending anything in marketing, what is the minimum sales that you will still do as a brand? A linear regression model will simply fit a line by minimizing the error called Lee square and think of this as a trend extraction from your data soup. If you look at this specific data, you can see there is a intercept or your baseline spend. Even if you spend $0 in the ad spend, you will still have this much amount of revenue coming. So these old red data point are my past sales based on the ad spend. So this was my sale if you spend this much of money. Similarly, in this data point, this was my sale if you spend this much of money. And based on the past data, I am simply creating a simple line that passes through all of the past data, and the distance between this data and this line is minimum. So if you create a squared sum of all the distance that is least in this line, and that is my best fit line. That's why I call it as. It fits a line by minimizing error called e square. So this line, when you calculate this error from all of these data point to this, this line has a cumulative minimum error, and that's why this line is the most fit. And that is the main idea of regression. You're simply drawing a line and obviously this line will stretch dot dot dot, which means we are predicting something and we will be able to predict if you spend this much of money, obviously you have this much of revenue. Now you can have this line maybe in this way as well, in this way as well if most of these data points are present near to this line. I hope now you're able to understand what is regression and how does it work? We'll obviously solve a real world case study to understand this topic better. Perfect. Now let's look into a couple of real world use cases of regression model. Couple of those use cases are campaign forecasting. You want to predict how much revenue will you able to generate from a Black Friday campaign. And obviously, this will use the past data of the last few Black Friday campaign that you ran in the past. Then price optimization. You will find a sweet spot where you can increase the price of a product and the sales doesn't drop. Third one is ROI modeling. You will understand how each journal is performing and how will they perform in the past if you pump in more money. Obviously, you will max out there are a couple of ceiling and thresholds that you need to consider. But you can use your linear regression model across all these use cases, from inventory planning to forecast, how much seal you will have in the future to how much money can you spend, how much revenue can you generate? In how much price can you sell a specific product? Now there are cases where linear regression may not work. When you start plotting data, if you see these random patterns, in these cases, regression doesn't work. Like, if you look at this line, you will see that this data is present randomly at so many places. That means this variable and this variable is not correlated. They are not moving in the same direction. When I say same direction, it should not be positive. It could be this way as well. Similarly, when you look at this data, you will see there are a couple of outliers. And in this case, see, there is something wrong. And in this case, regression won't work. You have to use another techniques. If you have a nonlinear behavior, in that case, regression will not work. If you have so many outliers, let's say suddenly your brand is picking up, you're getting so much of random seals and you don't have some variables to predict that random sales. You have too many outliers. In this case, if something goes viral, then you can't really predict things. Also, if you have high correlation between two variable, in that case, you can't really make much sense out of it.