AI for MARKETING AND BUSINESS: AI PRODUCT MANAGER Skills for Agile & AI Product Management

1. Introduction: Guys, welcome to this section. AI for Agile product managers and product owners. If you've ever wanted to use AI to make better product decisions, refine backlogs faster, analyze customer feedback properly, or confidently decide what to build MX, you're in exactly the right place because AI in product management isn't about generating random ideas. It's about generating products your users love and value faster than ever before. In this course, we're going to take you step by step through how to integrate AI into real product work responsibly, strategically, and practically. The end, you'll know how to use tools like Chat GPT to support discovery, prioritization, sprint planning and roadmap decisions without losing human ownership of the product. You can go on to use these skills in any AI tool like Chat GPT, Claude or Gemini. Let's start with a real world scenario. Imagine you're a product owner or product manager on a fast paced, demanding project. Let's take an app for sports or ecommerce company as an example. Reviewing app store feedback, analyzing user data, preparing for backlog refinement, writing user stories, aligning stakeholders around a roadmap, deciding whether a feature is worth building, and setting sprint goals for the upcoming sprints. It feels constant and high stakes. Now imagine having an AI partner that can analyze thousands of feedback comments in minutes. Surface patterns and sentiment you might miss generate structured sprint ready user stories. Propose sprint goals aligned to your release. Stress test your roadmap decisions. Highlight risk before you commit to build summarized meetings and extract action items. Help you evaluate whether an idea should move forward at all. That's exactly what modern AI can do, but only if you use it correctly. This course isn't about hype, it's about structured AI collaboration inside the product owner role. Throughout this course, you'll learn how to clearly define where AI belongs and where human judgment must lead. Turn market research into a confident product vision and strategy. Use AI to generate and refine product requirements, create databack user personas, translate insights into prioritized user stories. Use AI for backgrog refinement and ambiguity detection, support estimation discussions with structured analysis. Generate strong sprint goals that ladder up to your release goal. Integrate backlogs into realistic adaptive roadmaps. Use AI responsibly in meetings and team processes. You'll follow a real life end to end product case study, creating a worldwide sports app. So you can see exactly how AI supports each stage of the product life cycle. You won't just learn prompts. You'll learn workflows. You'll learn decision frameworks. You'll learn how to think with AI while staying firmly in control. The end of this course, you'll be able to reduce manual analysis time dramatically, make clearer, more structured product decisions, improve alignment between vision, backlog and sprint execution, avoid common AI pitfalls like hallucinations and over reliance. Operate at the higher strategic level than most product managers. AI will not replace product managers. But product managers who know how to use AI effectively will outperform those who don't course gives you that edge. So follow along with the demos, refine the thinking. And by the end of this course, you won't just be using AI. You'll be leading products with it confidently, strategically and responsibly. So let's get started. 2. The Birth of AI: So to understand why we even need prompt engineering, we need to understand where it came from and how it relates to AI, the birth of chat, G, PT, and LLMs. From the simple rule based programs of the 1980s to today's smart creative chatbots, one thing has stayed the same. Our goal to make computers understand and respond similar to humans. The story of chat, GPT, and large language models, LLMs is really the story of how that dream became real. How we went from basic machines to powerful tools that can think and write using everyday language. Although it feels new, artificial intelligence AI has actually been around for decades. To understand where Chat GPT came from, let's take a quick look at how AI evolved over time. It all began in the 1950s when computer scientist Alan Turing asked a famous question. Can machines think? That one question started the entire field of AI. In the 1960s, a simple program called Eliza was created at MIT. Would hold short conversations by matching patterns in text. It wasn't truly intelligent, but it was the first step toward computers that could use language to communicate. In the 1980s, AI was used mostly for what were called expert systems, programs that followed rules written by humans. These systems could give medical advice, approved loans, or help design products. They were useful, but they had one big problem. Couldn't learn or adapt. If something changed, you had to rewrite the rules yourself. The 1990s and 2000 brought a big shift, the rise of machine learning. Instead of being told exactly what to do, computers started to learn from examples and data. This approach quietly powered many products we all use today. Spam filters learned which emails were junk. Google search got smarter at finding the right pages. Amazon began recommending products you might like while Google Maps learned to predict the fastest routes based on live traffic. Came alexa and other voice assistants, which could recognize speech and answer questions out loud, something that felt almost magical at the time. Computers were now learning from experience, not just following rules, but even then AI couldn't really create things. It could predict and categorize, but not write, imagine, or explain. That changed in 2017 when researchers at Google developed a new system called the transformer. It helped computers understand how words relate to each other in a sentence, not just one word at a time, but in full context. This was a huge breakthrough and laid the foundation for the next big step in AI, large language models. A large language model, LLM is an AI trained to understand and generate human language. It learns from massive amounts of text, books, articles, and online content by spotting patterns in how words and ideas connect. After the transformer architecture was introduced in 2017, models could finally understand context and meaning. This led to powerful systems like GPT, capable of writing, summarizing, reasoning, and chatting in natural human like ways. That next step came from Open AI, the company behind Chat GPT. They built on Google's work and created something called the GPT series, generative pre trained transformers. The first version GPT one showed that a computer could learn to write readable text by studying huge amounts of online data. Then came GPT two in 2019, which could write essays, stories, and even news articles that sounded human. A few years later, GPT three made an even bigger leap. With 175 billion parameters or neurons, it could write, translate, answer questions, and even code. There was just one thing missing, natural conversation. GPT three could give answers, but it couldn't chat as fluently. So Open AI improved it using feedback from real people, teaching it how to respond more naturally and politely. The result was chat GPT, a version that could hold a conversation. Remember what you said and reply in a way that felt personal. That was when AI truly became something everyone could use, and this is where prompting came in. The word prompt originally came from early computers. It was the line on the screen where you typed a command. The computer was waiting for your input. Over time, the meaning changed. Now, a prompt means the message or question. You give an AI a system like chat GPT to tell it what you want. At first, prompts were very simple. Things like write an email, summarize this text or explain this to me like I'm five, but people soon notice something interesting. The way you wrote your prompt completely changed the answer. A detailed prompt gave a detailed result, a clear question, got a clearer answer. The better your prompt, the better the AI's performance. This turned the act of prompting into both an art and a science. Today we call this art and science prompt engineer. 3. Why Product Managers need Prompt Engineering?: Okay, guys, after learning how chat GPT, and large language bbles came to life, the next step is understanding how to talk to them, and that's where prompt engineering comes in. If LLMs are the engines of AI, prompts are the steering wheel. They allow you to control the direction, quality, and creativity of what AI produces. At its core, prompt engineering means writing clear, structured instructions that guide an AI model like hat GPT, Claude, Gemini, or others. To produce useful, high quality results. A prompt could be as simple as write an email or as detailed as act as a marketing expert and design a campaign for my new product. The better you describe what you want, the better the AI's output will be. But how did this idea begin? Prompt engineering emerged as people started using AI tools like hat GPT, Mid journey, and Dali more creatively. Users quickly discovered that two people could ask the same question and get completely different answers. And the difference came down to how they phrase the prompt. This led researchers, creators, and educators to study the patterns behind effective prompting. Early thought leaders like Ethan Malik, Andre Carpathi and Seth Dobrin along with OpenAI's research teams began to share techniques that worked consistently. These evolved into the core prompt patterns used today, frameworks that help users think and write more strategically. Among them are the instruction pattern, giving clear direct demands, chain of thought pattern, guiding the model to reason step by step, persona pattern, assigning the AIA specific role or perspective, template pattern, creating reusable prompt structures and the iterative refinement pattern collaboratively improving the output. These patterns became the backbone of prompt engineering, a way to get predictable, powerful results from any AI system. Prompt engineering has since become the bridge between human creativity and machine intelligence. It's powerful because it gives anyone not just programmers, the ability to direct AI systems to perform complex or creative tasks. In short, it turns you into a kind of AI conductor, guiding the output you want, and it's not just for chat GPT. The same skill applies to a wide range of AI tools. Used across different industries, Canva, create marketing designs and social posts from simple text prompt, photoshop firefly, generate or edit high quality images using natural language, Notion, AI, draft reports, summarize notes, and automate documentation. Runway ML, turn written scene descriptions into professional videos. Microsoft 365 copilot, write emails, analyze spreadsheets, and prepare presentations from prompts. Prompting is now a universal AI skill. Whether you're a marketer, teacher, designer, developer, entrepreneur or project manager, learning to prompt effectively will help you work faster, automate tasks, and unlock new opportunities. It's also a practical way to earn more. At work, prompt engineering can save hours, automating emails, reports, and customer interactions. Entrepreneurs use it to scale content, analyze data, and generate business ideas without hiring large teens. Freelancers now sell AI based services from writing and design to strategy and automation. The people who know how to communicate with AI are already leading the way in productivity and creativity. Here's what I want you to do first. Imagine you've just been asked to use AI to make your current role faster, smarter, or more effective, or maybe even to help you land a new role entirely. Think about your day to day work. Are you managing projects, designing presentations, writing reports, planning lessons, selling products, or building marketing campaigns? We'll be using Chat EPT throughout this training to show you exactly how it's done. But remember, what you'll learn applies to all large language models and many of the other AI tools we mentioned earlier. Like Canva, Notion and Microsoft Cope either. The goal is for you to think in terms of how AI can assist you no matter your role or industry. If you're a project manager, imagine AI helping you summarize project updates, create meeting agendas, and identify risks in seconds. If you're a marketer, think about how it could draft, ad copy, analyze competitors, and plan social media posts, a teacher could use it to generate quizzes and lesson plans. A business analyst could interpret data and find patterns. A freelancer could create proposals, automate admin work and deliver projects faster to make it practical, is an example of someone running the day to day for an e commerce business because it applies to so many roles. So take a moment to think about how an AI, I can help you in your role because by the end of this course, you'll know exactly how to make that happen. So let's get started. 4. Which AI Tools To Use?: Now you understand why AI is valuable for agile project managers and scrum masters, and you've seen the types of challenges it can solve, backlog chaos, meeting admins, sprint planning complexity, and documentation gaps. Knowing why AI is valuable is only half the battle. The next step is knowing which tools to use because there are dozens of AI options out there, each with different strengths, weaknesses and ideal use cases. In this lesson, we'll cover how to decide which tools to use in your workflow. Why large language models, LLMs are central to many AI solutions. Why we chose Chat GPT for this course. Why Jira and Confluence with ovo are the right fit for backlog and project management. Why fireflies.ai is ideal for Agile and Scrum teams. By the end of this lesson, you'll have a clear framework for selecting the right AI tools for every part of your scrum workflow. Deciding which tools to use, consider three main criteria purpose, integration, and capabilities. Purpose. What problem are you trying to solve? Backlog grooming, documentation, sprint planning, meeting summaries, integration. Does the tool work seamlessly with your existing systems? For example, Jira and Confluence integration ensures AI outputs directly into workflows your team already uses? Capabilities. Does it understand natural language? Can it generate structured outputs like tasks, summaries, or story points? Some tools are conversational, others are data analysis focused. Rule of thumb, match the tools strength to your specific use case. Don't just pick the coolest AI. Why we use large language models, LLMs, like hat GPT are at the core of many AI solutions because they can understand context, generate text, summarize information, and answer questions in natural language. Scrum workflows, LLMs are used to generate or refine backlog items. Summarize meetings and discussions, draft documentation or confluence pages. Answer questions about processes or project status. LLMs are flexible, general purpose, and capable of reasoning over unstructured data, unlike a single purpose tool that can only do one thing like transcribe a meeting or suggest story points. They act as a copilot that can generate context aware content, summarize insights, and adapt to new prompts in real time. Choosing Chat GPT for this course, we selected Chat GPT because it is by far the most popular LLM at this time, and we've tested it extensively for Agile Scrum projects. It is reliable, widely integrated and works exceptionally well for generating backlog items, documentation, summaries, and other scrum specific tasks. Why project management and backlog workflows need more than just an LLM? While LLMs powerful, they cannot enforce project structure or integrate deeply with your team's workflows by themselves. A raw LLM can suggest a backlog item, but it doesn't know your boards, epics, or sprints. LLMs don't automatically push tasks into project management tools. They cannot track dependencies, assign tasks, or calculate team velocity. This is why we combine LLMs with Jira, confluence and ovo. Jira manages Scrum projects, supports sprints, epics, story points, and boards, Confluence centralized knowledge management, linking documentation with tasks. Ovo AI bridges AI with Atlassian products, automates tagging grouping prioritization and ensures outputs from LLMs fit directly into Jira and confluence together LLMs. Plus project management tools provide intelligent content plus structured workflow, allowing you to generate, organize, and execute work seamlessly. Why fireflies.ai is ideal for Agile and Scrum teams. Meetings are one of the biggest drains on a scrum masters time, and in agile, meetings are frequent stand ups, sprint planning, retrospective, stakeholder reviews, backlog refinement sessions. Fireflies.ai is our tool of choice because it directly supports the rhythm of Scrum. Automatically records and transcribes every meeting. So you never miss key decisions, generates summaries and actionable items instantly. Perfect for sprint planning, backlog grooming or retrospectives, pushes tasks directly into Jira and links them to backlog items, epics or stories, captures retrospective insights and decisions, helping teams continuously improve without manual note taking, integrates with confluence to update documentation and centralized knowledge. Compared to transcription only tools like otter.ai, fireflies is purpose built for agile workflows, turning conversations into structured actionable outputs that directly tie into your scrum process. This means less Admin, better alignment, and faster decision making, exactly what agile and scrum teams need. Summary takeaways. Here's what to remember. Choose AI tools based on purpose, integration, and capabilities. LLMs like Chat GPT are flexible engines for generating text, summaries, and insights, but they don't manage workflows alone. So we've chosen Jira and Confluence powered by Rovo AI to manage our workflows and backlogs, fireflies.ai to transcribe meeting notes and generate action items, and hat GPT as our LLM for general summaries and insights into our project. So now we've decided on our toolset, let's get started. 5. Getting Started With ChatGPT: Okay, everyone. In this lesson, we're going to learn the simple but powerful skill that's really going to make chat GPT instantly more useful at work, and that's prompt structure. The reason we're teaching that is because most people type a quick question or a quick instruction to chat GPT and hope that they're going to get the response that they want. However, that's why most people get vague answers and they get different formats, different responses that don't usually match what they actually need. So prompt structure fixes that issue. Instead of simply asking a question, you're going to learn how to brief Chat GPT properly, so it gives you a clear, usable answer on your first try. And that's how to be most effective with hatGPT. And we're going to use our real day to day scenario of running the daily operations for an ecommerce company, and we're launching a new track suit line. So, so here we're just getting started launching our new tracksuit. We're completely clueless about how we're going to do that. We're a new company. So we've come in, and we've said to ChatHPT, I want to launch a new track suit. What should I do? Just an open question. Let's see what ChaiBT tells us. So here we see that ChachiBTs come back with a lot of stuff. It says launching a track suit is a mix of product brand and execution. Here's a clear step by step roadmap you can actually follow from idea to first sales. Now, that's good. However, what you'll notice is, we've got a lot of stuff back here. We've got eight different steps here from defining the track suit, designing the samples, finding the right manufacturer. Branding it, pricing and margins, building hype before selling, launching, collecting feedback, and then it gives us some optional things that some optional things that it can do to help us even more. Now, while that's great, it's not very focused because in a company, sometimes we know exactly what we want to do, or sometimes these steps might change based on the goal, based on what we want to do in the end. So now let's add a little bit of structure to our question and see what happens there. So if we zoom in here and have a look, we've given it a goal to create a clear product goal for a new premium Traxit launch. The role, you are a product lead. So we're telling ChachPT it's the product lead so it knows what mindset to get into to do this, the context. The brand is a premium, modern athletic brand. We're targeting customers 25 to 40-years-old, and then for the style, it's going to be more of a performance based style. The key features are slim breathable, minimal branding, and the channel where we're going to be promoting this is on the online store and social ads. So we've been quite minimal in how we've worded this, but it's enough for Chat CHiPT to get the direction and the context within which we're marketing and promoting this product and the constraints. So we want the feedback practical and specific, no fluff and no longer explanations. The output format should be a product goal in one sentence and a success metrics using three bullets only and then some key decisions needed next, which should be five bullets only. So we've been really specific about what it should give back. And the reason if we're running a business, we know exactly what feedback we want. We may not need all of the steps that Chat GPT gave us before, but this will give us enough just to get the starting point for our product and we can go from there. So if we run that, let's see the difference between the output we got, which was an eight step plan and what we get back now. So here we can see we've got much more concise output. It's given us exactly what we've asked for. So a product and the product goal is to launch a premium slim fit, breathable track suit for style driven, performance focused adults age 25 to 40, achieving strong direct to consumer validation through online sales and paid social. So it's a lot more targeted via this goal, and it's something that we could explain to anyone in our business, the success metrics, and then the key decisions that we need next. So here we've got a product we know what success looks like, and we know what key decisions we need to make next. Now, if had we wanted a step by step plan, this would have been great what Chat GPT gave us. You can see already simply by being specific about what we wanted, we got a much more specific answer, and that allows us by being specific about what we wanted to get the starting point we want and avoid wasting time. The key thing to remember here is that the format you give it, the goal, the role, the context, any constraints, and the specific output format can completely change what you get back and be a much better use of your time when using Chat GPT, allow you to use it much more effectively. But in sum here's what we've covered. So we know that vague prompts produce vague answers. Prompt structure turns Chat chiefeet into a reliable work assistant, and that's what we've just showed. The core structure is the role, the goal, the contexts, any constraints and the output format, and we used it to generate a clear product goal for our track suit launch. So now it's your turn. So what I want you to do is I want you to pick something you need to do this week in your projects, in your business, in your career, your company, or in your personal life and write a structured prompt using the template. Tell Chat GPT you are a role, the goal, the exact outcome you want, the context, and only talk about what matters in that situation. So you can talk about the situation that you're in. You can talk about the environment that you're going to be in and anything that matters specifically to achieve this goal, the constraints, so the length of what you want back, the tone, and the rules around it, and the output format. Do you want numbered sections? Do you want bullets or a table? And examples of things you could do you could create a product goal. You could write a product description. You could draft a launch plan for something you're doing in different steps, or you could summarize some research, some competitive research into decisions. The key thing is once you get this right, Chat GPT stops being just a chat bot that gives you random information, and it becomes a tool that you could work through. So that's the end of the lesson. I hope that was helpful, and I'll see you in the next one. 6. How to Capture User Data (Comments & Reviews) - Part 1: So in this lesson, we're going to cover market research and how to analyze user data such as comments and reviews. So why are we doing this? Well, as a product manager or product owner, one of the hardest parts of the job is deciding what is worth building in the first place and having some facts to tell us how to do that. So we might love sports. We might have a strong idea in our head, something that we're really excited about, excited about building something new. But before we build anything, we need to answer a few questions. We need to answer the questions. Do users actually want this? What problems are they trying to solve? What frustrates them today as things stand? Passionate about sports, hence the name passion sports. And so, ideally, we'd like to build an app, what we're calling the Passion sports app. However, we need to understand what people want in the market. Is it even worth building this? So let's talk about what we're going to use in order to tell us whether this is worth doing or not. So the feature we're going to use is called deep research with Chat GPT. Why use it? We use it to get structured to do structured multi source analysis, so it's going to go to many different sources and compare the results and give us some research based on that. It solves, it helps us to analyze large volumes of user feedback in an organized and unbiased way. We definitely want a lot of data to make sure that we've done good enough research. So we want large numbers of data or a great number, a large quantity of data, and we want it organized, and we want it unbiased. So we don't want to be told what we want to hear. We want to know that this is the truth. So what does deep research do? Deep research allows Chat GPT to look at patterns across many sources, to organize its findings clearly, to separate the facts from the interpretation, and it also gives us a structured output that we can use in decision making. So it's not just a quick search that you might get in a web search, and it's not just quick feedback that you get from Chat GPT. This is really deep research, the kind that a researcher would do, and what are we going to analyze as part of this research? We're going to be analyzing things such as user comments at reviews and user feedback from other apps. So we've got a document which I'll show you, which is full of all that kind of research, and we can use that in our deep research, and we're also going to look at search trends, search signals. What are people searching for that relates to the Sports app? So first of all, we're going to concentrate on the feedback from users, and the goal is simple. We want to understand what users actually care about before we decide what to build. So here we are at Chat GPT, and as I said before, we really want to do deep research with this. So the first thing we need to do is to go to the plus sign, and here we click the plus sign, and we choose deep research. And that means that we're going to, as I said before, go a lot deeper and use multiple sources to get feedback on what we are asking Chat GPT. The next thing that we want to do is paste in our prompt, and here it is. So if we scroll up, we're using the format goal role, context, action, format, and then some dos and don'ts. And I'm going to explain that as we go. So the first is the goal. So we're telling Chat TPT what the goal is. Now, the goal is to understand what users value, struggle with, and request most in sports apps that provide news and stats. The role, so we're telling ChatGPT what its role is here. The role is to act as a product research analyst, supporting an agile product manager. That's us at the early discovery stage. Um, the context is we are exploring whether to build a passion sports app. The app may include sports news and AI enhanced statistics. We have access to user comments and reviews from existing sports apps. Search trends show interest in sports news, live scores, and sports statistics. We are still validating whether this idea is worth pursuing. The action, so this is what we want ChatPT to do. We want it to analyze user comments and reviews to identify reoccurring frustrations, frequently requested features, positive themes, users consistently praise and gaps in clarity, usability, or insights. Want it to focus on patterns across many comments, not isolated opinions. So this is really key. This is where we're saying really show us from this research that it's worth doing, based on what users are saying out there, and this is making it the best possible app by focusing on what users really want. The format we want this back in is we want ChachPT to present findings under these headings. One, most common user frustrations, two, most requested improvements, three, what users value most and four clearer opportunity areas. We want it to use structured bullet points, separate user signals from interpretation. Now, the dos and don'ts. So these are the dos and don'ts for Chat GPT when generating this feedback. We do want it to focus on repeated patterns, highlight high frequency pain points, be objective, separate facts from suggestions. What we don't want it to do is to jump to any product features yet. So we don't want it to suggest AI unless user signals support it. And I'm going to be a bit more specific here. So I've updated this to say, I don't want it to suggest AI as a feature in our app unless user signal support it, and we don't want it to introduce competitive analysis at this stage. That's something we'll be doing later. We also don't want to over engineer solutions. So if it gives us some solutions, we want it to be as simple as possible. And that's that, really. So with deep research switched on, we can start the process. It may take a while, but it's worth it. So let's go. We'll go there, press the button, give it a chance to grab the research for us. So here within seconds, Crack GBT has given us the kind of overview that we've asked it to. So it supports that review analysis. So it said, first of all, collect user reviews from the top sports news and statistics apps across app store. And it said extract and aggregate recurring keywords, sentiments and feature requests. Then it said, identify high frequency frustrations, praised features, and clarity gaps, and it said Cross check findings with search trend signals, search trend signals, timestamp and summarize patterns under the four required headings with evidence. Now, I've just realized I haven't uploaded the CSV, so I'll show you the CSV. This may time out, but we'll come back to it if we need to. So here we are inside the CSV I created earlier. What this is is an example of the kind of user feedback that we can get from some of the bigger sports apps out there. What we would do is we would use another app or we could just go manually to the Apple App Store and just gather all of the feedback from various users. So, for example, if we zoom in a little bit here so you can see what we've got, J Henderson on this day gave the Apple rating of two and said, still getting the maximum number of devices reached error after logging out everywhere. Support is non existent, so they're obviously getting some kind of error here and they're not happy. And we've go get 2026 who gave the rating of four and said, Love the new layout for the Premier League scores. This is obviously for a football app, and it said, but why does the app take so long to load the video highlights on five G? So they're obviously on a mobile network and they don't like how long it's taking to load video highlights. And these kind of things can be seen for many different features in the app. Now we've got over 100 or about 100 reviews in here, but we could make this up to 500 or as many reviews as we want. Grab as many reviews as we want for any app and put them all together in this spreadsheet. And then what we can do is send that into Chat GPT and get it to analyze this and give us some feedback. So we're going to include this in our research, and this is going to allow us to get a good overview of what users want and what users don't want and how we can improve in our app. So let's go over to Chat GPT and input data in our research. Here we are at Track GPT, and I'm going to include this in our big research. So I'm also going to include this by clicking on the Plus button down into files, and I'm going to choose the file. And there it is. So that's loaded now. It's in there. We're still in deep research mode. So I'm going to go back, copy what I did before, our prompt. So now we're going to run our prompt along with the files. Let's go. So now we get our summary. Summary says sports user reviews analysis, collect recent user reviews from top sports news and SAT apps across our app stores. Extract recurring keywords, sentiments and explicit feature requests from reviews. Aggregate counts and timestamps to identify high frequency patterns, cross check patterns with search trends, signals, and review dates for recency. Summarize findings under the four required headings. So now the deep research is running. You can see here it's telling us what it's doing, and we're basically going to wait for this line to get to the end, and then we'll have the results of our deep research. So we'll come back and look at that once it's concluded. 7. How to Analyze User Data (Comments & Reviews) - Part 2: So here we are back at Chat CPT, and we can see the deep research has stopped running. Research completed in 5 minutes, and there were 11 citations, zero searches. But I think by zero searches, it just didn't necessarily have to do anything more than what we asked it to. And it says most common user frustrations. So let's take a quick look at what it says. We don't have to go into detail, but it will give you a flavor of the way deep research works. It says here that the most common user frustrations were intrusive ads and promotions. A large number of reviews mentioned excessive ads and often gambling ads interrupting the experience. Users say things like ads also take up too much space, and video ads in the live coverage are intolerable. It speaks of frequent crashes and performance by lag. Many users report stability issues, poor navigation, and confusing UI. So let's dive in a little bit more. Poor navigation and confusing UI. Reviewers repeatedly described the apps layout as a maze or clunky to navigate. We've got irrelevant or missing content. Users often complain about force or missing news. Comments include forced to see news. I don't follow and lament that certain leagues, for example, smaller college conferences, you have seen niche sports aren't covered. We've got account login and device issues, notification overload, so too many notifications and the most requested improvements are pretty clear here. It says, better personalization and customization. Pretty good for us thinking about AI because AI helps with that. Don't want to get too bogged down with AI yet. This is just a clue. We've also got streamlined navigation and search. Multiple users say navigation needs a redesign, add missing features and content. There are many explicit feature requests, reduce intrusive ads, improve apps stability, so so too many crashes. They want it faster. Users value most comprehensive up to the minute coverage, customization and personalization, innovative content and features, quality and reliability, broad inclusive coverage, and then opportunity areas. We've got AD and UX optimization, personalization and feed control, and streamlined navigation and search. Content breadth and depth, many users want complete coverage or more complete coverage. We've got robust performance and offline capabilities. Given the frequency of crash and load complaints, delivering a fast stable experience is a clear need. Got advanced statistics and insights. While users didn't explicitly demand AI, they do appreciate rich data, wind probabilities, and climb stats where present. And then it says here, by addressing these high frequency signals, streamlining adds a new I, enhancing customization, bolstering content, and maintaining performance, passion sports can target the precise areas where existing sports apps earn praise and fall through. Each point above is grounded in multiple user comments with citations. So as you can see, we've got a really good overview of many different points. What users didn't like all comes primarily from what we uploaded and deep research that's been done on those points. Gives us a really clear path for moving forward and deciding, is there a need for this app anyway? We've got more research to do, but this is a great start. So what we can do with this is actually summarize it so that we've got a nice, concise summary of what it's telling us. So what we'll do, first of all, go back to the bottom, and we don't want deep research at present, and let's paste in a new prompt. Here we're saying, based on this research, where is user frustration highest? Where is user appreciation strongest? What appears undeserved? Is there enough demand to justify building a new sports app, and I've added to be concise and summarize. So let's see what happens. Let's go for it. So we've got quite a lot back with that. I could have probably said summarizing ten lines or five to seven bullets or something. Reading through it, it's repeated a lot of the same stuff. The most important things here are that it's again, message that there's poor personalization that's really relevant to the ideas we had. So they're frustrated with too much repetitive or low quality news. So that's interesting. And the important question is there enough demand to justify building a new sports app? The evidence of demand is shown here, but the conclusion is, yes, there's demand for sports news, live scores, and detailed stats, and that's good for us. However, it does warn that the space is crowded and expectations are high, so we need to deliver a faster, cleaner, more reliable information with clearer insight, not just more events. So that's again, good for us because we really do want to give clearer insights that plays towards what our idea was too we seem to be in the right direction so far. Let's continue with the research, and we'll see, so are we learning as leaders? This step helps us to avoid building features no one asked for. To avoid copying competitors blindly, avoid falling in love with our own ideas, and also it helps us to make decisions based on evidence, and AI is helping us to speed up the research process so that we can decide exactly what to build. It doesn't decide for us, but it does give us faster clarity. So I think this has been really worthwhile exercise before we do any more research. And more importantly, before we build anything, we've got real certainty using AI to help us to know what to build for our users. So now it's your turn. What I want you to do is use deep research and chat DPT a product area that you care about or that you've gotten an idea for, something you want to validate and repeat the process that you saw in this lesson. For example, you might think about fitness or a health app, finance or budgeting app, productivity app, and education or learning platform. And then what I want you to do is ask the question. Based on the user comments and search trends, you want to be sure what problems appear the most common. Where does frustration cluster? What seems undeserved? Is this strong enough to justify building something new? So the goal isn't to find the perfect idea. The goal is to practice making product decisions based on real signals. So have fun doing that, and I'll see you in the next lesson. 8. How to prepare Competitive Analysis: Okay, so in this lesson, you're going to learn how to conduct competitive analysis from market research with AI. So we've already analyzed user comments and feedback to understand the real users needs, and now we're going to move to the next important question that we ask as product managers or product owners, which is, if we build this passion sports app, how will it stand out from the competition? We're still in the research stage. We're not designing any features yet. We're at the very beginning. We're trying to understand what competitors do well, where they struggle, where users are dissatisfied, and where gaps may exist. And competitive analysis isn't about copying the competitors. It's about identifying opportunity and differentiating ourselves from the competition. Let's get into it. So the feature we're going to use is deep research. Again, as we've used in the past, and while we use it, we use it for structured competitor analysis from multiple different sources. It helps us to go a lot deeper into the web, into the Internet and get the information that we need. What it solves, it solves the challenge where we need something to help us to evaluate our competitors objectively. And we don't want to rely on assumptions. So deep research allows Chat GPT to analyze multiple competitor sources, identify recurring strengths and weaknesses, and highlight differentiation gaps. So things that are going to differentiate us from our competitors, and then we can structure what we find really clearly. Here we are at Chat GPT, and you can see that previously we did some deep research, and it's here. It's the user insights. Now we're going to do the competitive analysis. So let's start ourselves a new chat, and let's paste into the prompt that we're about to use. If we zoom in a little bit, our goal is to understand the strengths, weaknesses, differentiation gaps among leading sports apps that provide news and statistics. The role is to act as a product strategy analyst supporting an agile product manager in the early stage of product discovery, and that's the stage we're at. Context is that we are evaluating whether to build a passion sports app, which may include sports news and AI and heart statistics. We want to understand the competitive landscape before defining our product vision. We are focused on popular sports apps that provide live scores, news and statistics. We are not yet designing features. We're identifying gaps and positioning opportunities. So it's important to let the AI know that we don't want to go too deep into what features to build at this stage. First of all, we just want to see where the opportunities are and if this is even a good move. So the action, what we want hATTPT to do is to analyze leading sports apps then from that, we want to identify core strengths consistently highlighted, recurring weaknesses or user frustrations. Feature similarities across competitors where differentiation appears limited. So our opt opportunities to be different from the market from the rest of the competitors, and then clear opportunity gaps in the market. We want to focus on patterns across multiple different sources. In the format, how we want this presented, we want ChachiPT to present the findings under the following headings. One common competitive strengths, two, common competitive weaknesses. Three, feature parity areas. So where everyone looks similar because we want to be different. Clear differentiation gaps and strategic observations. We want ChachiPT to show us the information is structured bullet points, and we want to separate observed evidence from interpretation. That is to say, if there's something we can clearly see is the case, that's different from if we see something, and then we make our own interpretation about it. Want to know the difference between those two things and the dos and don'ts for doing this research. What we do want CHATGPT to do is focus on reoccurring patterns across competitors, highlight user reported strengths and frustrations, be objective, and identify real gaps rather than hypothetical ones. What we don't want CHATGPT to do is design our product yet. Suggest AI features unless evidence supports it, or assume competitors are weak without evidence. We don't want it to jump into solution mode. In other words, we don't want to move forward into working out what we should do yet. We want to do the research completely in isolation of that. Although I mentioned CachPTobviously, you can do this in any LLM, any of your favorite LLMs, as long as you're confident that it will do the job for you based on your experience of the LLM or what you've learned in this course. So let's get into that. What we want to do is we want to as usual, go down here and go to more in deep research, and you'll see this pop up here, meaning we're going to be in deep research mode. So if we zoom out again, we can now hit the button and let the research commence. As usual, we see this pop up and it tells us that it's going through the process of deep research. It's about to start, but before it does, it tells us that it's going to compile a list of leading sports apps and official product pages, collect user reviews and ratings from app stores and forums, extract common features, strengths, and complaints across sources, identify the feature parity and clear market differentiation gaps, and synthesize patterns into structured findings. So once you've read through this, if there's anything to edit, you can click Edit. Apart from that, just get started. And there it goes. You can see it's researching. This will take a while as usual, so we'll come back when it's complete. 9. How to conduct deep Competitive Analysis: So here we are back at Chat GPT, and our deep research for our common competitive strengths and weaknesses, all of the things that we asked for in our report are now finished. So let's have a look. We can look at it here or we can just expand it slightly. We completed it in about 4 minutes, so let's dive in. So at the top here, if we look at each of these headings under common competitor strengths, we've got real time updates and wide coverage. The all leading apps emphasis emphasize instantaneous live scores and extensive league coverage. For example, Yahoo sports advertises real time scores and stats across NFL, NBA, MLB, NHL, et cetera. So real time updates and wide coverage is a strength. And then it's got the others. Without going into too much detail, personalization is a strength of competitors. Most apps allow users to follow specific teams and leagues and receive customer notifications, rich content, news, highlights, and analysis. Leading apps bundle scores with news articles, and it's got some other stats, some other things here, like in depth statistics. So under common competitive strengths, we've got a good idea of where our competitors are, whether they're strong, and it's got also some names of competitors in here, giving examples of names like Yahoo sports. Common competitive weaknesses are things like advertising intrusiveness, rencurring frustrations ads, disrupting the experience. And we've got points below and some paragraphs under these points such as performance and stability issues, poor navigation or cluttered UI, excessive or mistargeted notifications, data accuracy or content gaps. Then for feature parity areas where everyone looks similar, core live scores and basic stats. It seems like basic stats. It seems like every app provides the live scores and basic box score stats for major sports. Favorite teams and alert settings seems to be the same in most apps. New news feeds and highlights. A news section with written articles and video highlights is standard. Multi sports league coverage covering the league for many different sports. Video and live streaming. Now, where are the areas of differentiation? Social community features, few apps have built in fan community tools, exclusive content or gameification. Yahoo sports includes proprietary content and games, so it gives that as an example, advanced analytics or AI driven insights. So apps like SFA Score and 365 scores emphasize deep analytics, heat maps, pre game predictions. So that's an opportunity there, and they explicitly mentioned AI driven insights, ad free or less intrusive experience, niche customization, and in some strategic observations, ads are a major pain point. Ability and speed are decisive content versus interface trade offs, users want customization and focus. Convergence implies differentiation like niche or innovation. And it says, Implication, passion could position itself around advanced analytics, AI driven insights or as a specialist. So it's clearly telling us that there's an opportunity there. This report is detailed enough. We've got some mention of our competitors in there. So what does this teach us about act as product leaders? So the competitive analysis should reduce the strategic risk, prevent feature cloning, where we just do the exact same things as our competitors. We might then be able we might there may be some of that, but we can also differentiate ourselves and do something different, and it should expose weak assumptions. Things that we thought were true that actually weren't true at all. Highlight realistic positioning opportunities, how we can position ourselves in the best way to serve our customer. So AI has been used here to speed up the comparison process. So now it's your turn. What I want you to do is use deep research to conduct a competitive analysis in the product space that you care about. For example, it could be fitness tracking apps, personal finance apps, productivity tools, learning platforms. And once you've done that, then you can ask yourself a question based on the competitive landscape. Where are your competitors strongest? Where are they weakest? Where is a true differentiation or opportunity to stand out from the crowd? 10. How to Generate the Product Vision: Okay, so in this lesson, we're going to learn how to generate a product vision. So far, we've worked out whether the passion sports app is worth building. And we've said, Yes, even if we start with a minimum viable product to show whether it's worth moving forward and it's definitely worth building, we've worked out what problem we're solving to some extent, and we've also worked out where AI might create meaningful value. Now we need to turn this research into a clear direction. As a product manager or product owner, your job isn't just to collect insights. It's turning this insight into a clear product vision, and a strong product vision explains who the products for. It describes the problem that it solves. It clarifies how it creates value. It guides the backlog, the roadmap, the sprint decisions, and everything that comes afterwards. But we'll get into that. And this isn't about features of the product. This is about which direction to guide the product in. The feature that we're using is just standard chat chPT chat reasoning. Why we're using it is to create a structured synthesis of research insights. In other words, to structure our research and make sure that it all points towards the same. And then what it solves is it helps translate research into a clear focused product direction. So let's go over to Chat GPT and get started creating our product vision. So here we are. We're going to first create a new chat, but within the same project, so we can always refer back to our product vision in a new chat. Then we're going to paste in our prompt and the prompt is the goal. The goal is to create a clear and concise product vision for the Passion Sports app based on validated market research. The role is you are an experienced agile product leader helping to define early stage product direction. And let's give it some context. Context is we conducted market research on sports fans consuming news and statistics, and the research shows that users feel overwhelmed by too much fragmented information. They want faster understanding, not more data, AI summarization, AI summarization and personalization may create some differentiation and engagement and retention are critical business drivers. We are at the beginning of product development. We need direction before defining the strategy and the roadmap. So the actions that we want, what we want Cha GPT to do is develop the product vision statement, ensure it focuses on customer value, not features, clearly describes target users problem and value and ensure AI is positioned as an enabler, not the product itself. And the format we want CHAPT to give us for the information is as follows. Short vision statement, one to two sentences structured vision board, including target group needs, product direction, and value for the user and the business. And what we do want hATTPT to do is focus on outcomes, not featurest. Keep it simple and directional and ensure strategic clarity. We don't want to list the backlog items at this stage. We don't want to overemphasize AI as the main goal, and we don't want to be vague about the value that we're providing. Other thing I'm going to do is because I like the board, the vision board to be a specific format, I'm going to tell it to create that as an image because that format tends to stand out well. But before I do that, let's generate text because sometimes chat GPT gets confused. If you ask it to do two things, it gets confused. So let's go. So here we go. So the vision statement we've got Passion Sports app helps modern sports fans quickly understand what matters most by cutting through noise, delivering clear personalized insights so they can stay informed without feeling overwhelmed. So, that's great. No mention of AI specifically, which is good. So we can always fix it, we can always change it. This year, it's AI. In the next five years, it may be something else, and a vision statement should usually take us forward for at least the next five years. This keeps it nice and open. This leaves it open at the moment, and we can see the vision board digitally native sports fans who follow multiple teams, leagues or competitions and feel overloaded by fragmented news feeds and updates. That's good. Needs faster understanding of what's important, clear context instead of scattered data, relevant updates without constant searching, simpler way to stay informed daily. And the product direction is create a personalized sport companion that filters, prioritizes and explains sports content in a clear and concise way, helping users focus on what truly matters to them. I think that's going to be key. Speaks to the personalization and then the value, the user value, saves them time. It reduces information overload and increases clarity and confidence. That's fair. The business value is higher engagement through relevance, stronger retention via daily habit formation, clear differentiation through intelligent personalization. And I think that's good, as well. So what I like about this is it covers all of the issues we found in our research. It's vague enough that it doesn't tie us to a specific feature, but it's certainly enough that it covers what users most care about and what users need to resolve. I think one way we could improve this is by adding some clear understanding of what value we need to get as a business. Add a bullet to the business value to mention the expected revenue in the first year, based on research. So I've updated it to say add a bullet to the business value to mention the target revenue of 100 K in the first year, based on research. Let's see that. Under business value, it now says target revenue of 100 K in year one, validating product market fit and monetization potential. So let's update our vision a little bit and anything that goes with it. So I've created a prompt to say, update the product vision to focus the app on the top sports in the world. Update the vision board, or as needed with the minimum updates possible. And this, again, top sports in the world is slightly vague at this point, but what it does is it will filter down into everything we create. So let's go. So now it's updated division. Passion sports helps fans for the world's top sports quickly understand what matters. Most by cutting what matters most by cutting through the noise and delivering clear, personalized insights, so they can stay informed. I think that's great. And it focuses us on personalized insights. So now the next thing I want to do is create an image that shows this vision board, so we can clearly show it to any stakeholder or any of our customers. So here's the template that we usually use. It's got the vision statement here. Here's one I did before. Develop a world class sports website, and then it's got the target group needs product value. But what we're going to do is we're going to get Chat GPT to fill in this template using the text that we've already got. So if we bring up our vision board example, going to go to Files, we've got the example vision board here. So let's open that. Now we've attached that, and what we're going to do is to use a template to get this information into this image in this form. So the prompt is using the attached vision board image as a template, slash example, keeping the same headings and layout that are already in the image, create a new image for the passion sports based for passion sports based on the vision statement and vision board text you've created above. So the interesting thing here is we're using what's actually known as a vision within Chat GPT, a different thing that can actually see and analyze this image. So let's start. Okay, so here's our new vision board, Passion Sports app. Let's have a look at what it says. It says Passion Sports app helps fans of the world's top sports quickly understand what matters most by cutting through the noise and delivering clear, personalized insights so they can stay informed without feeling overwhelmed. Okay, so ticking the box there, target group, digitally active fans of the world's leading sports who follow multiple teams, leagues or competitions and feel overloaded by fragmented news, stats, and updates. Okay, so that's cool. We can always split that up more. We can make it specific age groups or whatever we feel is necessary. Of what needs, people who need faster understanding of what's important. Clear context instead of scattered data, relevant updates without constant searching a simpler way to stay informed daily. That's very, very accurate. I like that. And then it's mentioned personalization, so that's something we could do. So maybe we'll go back and look at that. And then the product, create a personalized sports companion focused on the world's top sports that filters, prioritizes and explains content in a clear and concise way, helping users focus on what really matters. There's also some things around value here. You can see it was repeated over here. So I think Chat GPT has made an error there. We'll clean that up. Let's read in Win value. It says, use a value saves time, reduces information overload, increases clarity and confidence in understanding sports events, and then business value, higher engagement through relevance, stronger retention via daily habit formation, target revenue of 100 K in year one, validating product market fit, monetization potential. So that's pretty good. So let's do that. So in this prompt, I've said in the attached image, there is some repeated text. To fix this under the product column, remove the value text, bullet and all text underneath. Be careful to only remove that text. Do nothing else. Excellent. And that's done a good job there. So you can see this area is clear now, and the other thing we want to do is add a bullet point regarding personalization to here. For this prompt, I've said in the needs column under a simple way to stay informed daily, which is talking about this area here underneath that. Add another short sentence in the same style as the sentences above that mentions personalization. Be careful to only add that text, do nothing else. So we've been really careful here because I've noticed Chat TPT can often get a little bit overloaded and confused. We're actually going to be looking at this one, which is this one we corrected. And my style of doing this is always to download the image and then re upload it into the chat. And we should now have a fully fledged vision board with all bullets, boards full. So there you go. It took a little bit of back and forth to get the bullet point update generated. So now we've got a full vision board generated completely by AI in a way that we can show any of our stakeholders. Now it's your turn. Using the same prompt structure that I showed you in the lesson, generate a product vision for whatever app you have Fintech, budgeting app, health and fitness tracking, business to business, software as a service, and analytics tool. These are all examples, and then ask yourself, does this vision clearly describe who it's for, what problem it solves and why it matters? If not, keep refining until it does. So that's that. Have fun with that, and I'll see you in the next lesson. 11. How to Generate the Product Strategy: So in this lesson, we're going to talk about how to generate a product strategy. We know who the products for, what problem it solves, how AI creates value, potentially, why it matters to the business. But a vision, which is what we've created is not the strategy because the vision says where we want to go and the strategy tells us how we'll get at a high level and how we're going to win with this app. So today, we're going to use Chat GPT to help us generate a focused outcome driven product strategy for the Passion Sports app. So the feature that we're using within Chat GPT is just the normal chat in Chat GPT. We're going to use the normal chat feature with structured reasoning. Why we use it, we use it to translate the vision into more focus, better focus, and the right priorities. And what it solves, it prevents a vague strategy and feature overload because the vision we've created is good and targets us, but we need to be a bit more targeted in various areas for us to call it a real strategy. So here we are within Chat GPT within the app strategy chat. Let's put it under strategy, and that's where we can continue in here or we could continue under the vision. I think we'll continue as part of the vision because we're actually going to expand that into the strategy. So let's paste in our prompt. So in our prompt, the goal is to create a clear and focused product strategy for the Passion Sports app based on the defined product vision. The role is you are an experienced agile product strategist, helping define a focused and outcome driven product strategy. Context, use everything you know from this product project, including the market research, vision statement, and vision board, including everything in the attached vision board image we created. So I've reattached the vision board because I made some updates updates on it. We are in the early stage development and need clarity before building the roadmap. And then the action is to define the primary target segment for launch. To clarify the core problem, we will focus on solving first. Three, identify two to three strategic differentiators. Four, define three measurable product outcomes. Five, outline key strategic choices, what we will focus on and what we will not focus. Six, clarify how AI positioned strategically. And then the format is to provide target segment core problem focused, strategic differentiators, key outcomes, which should be measurable and strategic choices, what to do and what not to do. And then AI positioning statement as part of it, the dos and don'ts. Do make clear trade offs, keep the strategy focused, tie decisions back to customer value, and things we don't want Chat GPT to do. Don't try to serve all sports fans, and I'm going to make this a bit more explicit. I've made this a bit more explicit, say, only for the top ten sports in the world. And that's just to give Track DPT some kind of focus. And don't turn this into a feature list. Yep, don't over emphasize AI as the goal itself. So now we've got our prompt. Let's go for it. So here, we've got our updated strategy or our strategy together. And let's go through quickly what it says. So it says the target segment is digitally active fans of the top ten global sports who follow multiple teams or competitions and consume sports content daily via mobile. Initial launch focus fans who actively check sports updates at least once per day. Our problem focus is sports fans are overloaded with fragmented news, stat and updates across multiple platforms. We'll focus on solving how might we help fans of the world's top sports quickly understand what really matters each day without having to search, filter, and interpret everything themselves. And the strategic differentiators is clarity and personalized relevance engine. Daily essential experience designed to become a sure, high value daily habit. And then key outcomes is engagement to one greater than or equal to 40% of active users. Open the app at least five days per week within six months, greater than or equal to 35% day 30 retention rate for new users. And monetization, as we said before, 100 K in one year. Then strategic choices, we're going to focus on the top ten sports globally. Prioritize updates, personalization, straightaway. That's a big one. Building a repeatable daily habit and then mobile first experience. What we're not going to focus on is covering all sports, deep statistics, databases, becoming a live score first or betting centric platform. Not going to focus on maximizing the volume of content for SEO. We're being quite targeted and we're not going to focus on building broad social community features at an early stage. And the AI positioning statement, AI is a strategic enabler that powers intelligence, filtering, summarization, and prioritization. And I'd say personalization, as well. Helping users understand what matters fast. AI is not the product itself. It's the invisible infrastructure. So our strategy is not necessarily to make AI the buzzword we went for. Our strategy is not necessarily to make AI the buzzword, we went on about, but it's more to give the user some features. We may mention it, but the main thing is to give the user some features that they need. So this is our strategy. I think this is pretty good. I was going to update it to mention personalization, but it's already mentioned. So I'm kind of happy with so one point of note is when we created our vision board image, if we take a quick look at a quick look at that now, under the vision statement, all of this is known as the strategy. So the strategy for the product is actually within here. So what I want to do is I want to slightly update these columns based on the strategy now so that this is aligned with our strategy we've just created. So here is our prompt to update the vision board. It says, based on the above strategy, make subtle updates to the sentences in each of the columns in the attached vision. Change the minimum necessary to align with the strategy, and I'm going to say minimum text. Be careful to do only what is asked and keep the same style. Do nothing else. Let's make that update. So when I ran that, Chat GPT actually started creating some text, and I had to be clear here, and I said, Forget my previous instruction. I'm asking you to update the attached image so it knows I'm talking about an image. Top of that, I set image mode by coming in here and clicking create Image, so there's no way it can misunderstand what I'm doing. So let's run that again. So now we've got a new vision board generated. Let's download this and compare it to the previous one. Here we are looking at the before and after. Now that we've updated our vision board to be in line with the strategy, and the vision statement has remained the same. No changes made. Everything's still in line. We're still focusing on delivering clear personalized insights so that the users can stay informed without feeling overwhelmed. So no issues there. These areas have now changed more in line with our strategy where before it says digitally active fans of the world's leading sports, this has narrowed down a bit. Digitally active fans of the top ten global sports who follow multiple teams or competitions and consume sports content daily via mobile. So it's got a little bit more narrowed down. And it's also added fans who actively check sports updates at least once per day and feel overwhelmed by fragmented apps, notifications, and so forth. So it's got some of what's here, but it's slightly updated, as well. These areas remain the same. It's still for people who need these things in terms of the actual product, instead of just saying focused on the world's top sports, it's now narrowed down a bit. Focused on the top ten sports that filters, prioritizes and explains content in a clear and concise way. The main thing here is the top ten sports. Again, that didn't change. But for now, this narrows it a little bit. We've removed some of the statement here about increasing clarity as we move into its own dedicated section, and we've also explicitly now in our strategy, added our AI positioning, which is great. So if we zoom in a little bit here, we can see it says, AI is a strategic enabler that powers intelligent filtering, summarization, and prioritization, helping users understand what matters faster. So it says, AI. AI is not the product itself. It is the invisible infrastructure that enables clarity, personalization and scaling. We're making this our strategy that we're not necessarily showing just about AI. We're showing about we're talking about benefits it's going to give our user and also give ourselves. So that's it. This is our new vision board. So now it's your turn. What I want you to do is use the same structure, the same prompt that I gave you at the beginning to generate a strategy for your app or for your product or whatever your product might happen to be. So I've thought, again, you know, ideas like the Fintech app, fitness tracking app, et cetera, and then ask does this strategy make clear trade offs? Is it focused enough to guide real decisions in your organization or for your app or for your product, whatever you're creating? If not, keep refining it until it does make clear trade offs, and it's focused enough to guide the decisions because apart from the vision, this is going to guide how you do what you do. And the strategy should feel slightly uncomfortable at this stage. If it doesn't if it doesn't exclude something, it usually isn't focused enough because you can't do everything. In your strategy, make it so that you actually have to make some core decisions about how you're going to do what you need to do, and you may need to exclude something in order to do that. So that's that. Have fun creating your strategy, and I will see you in the next lesson. 12. How to Generate Product Roadmap Clusters: So in this lesson, we're going to show you how to generate a product roadmap. So at the moment, we've got clear market research, we've defined our product vision, and we've got some structured product strategy and a vision board to go on so that we can go on with it. Nice. So now we're going to answer the big question. What do we build first and when? So if we're building this app, there are going to be a number of features. We haven't even created a product backlog yet. We're not at that stage yet. We're at the stage of just deciding at a strategic level what we need to build first and when. Some people already have a product backlog, and so they use that and group items on the backlog to work out what's in the roadmap. However, what we're going to start with is the strategy, and then we're going to build everything from there. A roadmap isn't the feature list. It's a goal driven plan that connects strategy to actual execution. And today, we're going to use the AI to help us to do that. So let's talk about what a product roadmap really is. A product roadmap and a strong roadmap at that should be goal based, not feature based. So we need to think about what our goals are. Before we think about the features. It focuses on outcomes. It uses measurable metrics. It has realistic dates and only includes high level features, and it's regularly adapted. And if it becomes just a long list of features, then that's really what we call a backlog, which is for another day, not for now. So let's go over to Chat GPT and start doing what we need to do to build ourselves a roadmap. So before we go on to do the roadmap, it's important to realize that I did ask one question after we created our strategy and vision. And I said, for the vision board and strategy, how would we organize the roadmap in terms of different sports? Do we first pick a sport and do it a sport at a time, or do the same functionality for a few sports or something else? Don't do the roadmap yet. Just tell me in a few sentences and tell me why. And I did this because I wanted to know if it thought I should be building a number of sports all upfront and with just a few features or maybe just one or two sports upfront with more features. And it responded. You should not build one sport and try it first. You also should not try to fully scan across all ten sports at once. Recommended approach. Build the core experience once, launch with two to three top global sports, then expand sport coverage incrementally. Why? Your differentiation is clarity and prioritization. Which sport should we start with in the UK, given that I'm in the UK? And it said, surprise, surprise, number one, football, first. First, the non negotiable largest fan base, it said. And the second one was Formula one. And it said strong strategic fit because of massive UK interest and a number of other reasons. So that's given us a bit more direction before we go on to actually build a roadmap. So that makes sense to me. Football and F one, that sounds good. And then what we can start to do is build a roadmap with a little bit more direction on and what it's based on. So that led me to adjust my prompt, which is why I did that first. So if we look here, we can see that the prompt is the goal is to help structure the Passion sports app standout features into meaningful roadmap themes aligned with our product strategy. So we've got some features that I've put into this prompt that we feel for the MVP for the minimum viable product are good ones to start off with. The role that we're giving AI is you are an experienced agile product manager helping prepare a goal driven product roadmap. Otext we are building an AI enhanced sports intelligence minimum viable product that aggregates sports news and stats, personalize insights, and uses AI to highlight trends and explain what's trending and why it matters. That's what we're building. And now I can say we will start with the football, soccer and Formula one in that order for our minimum viable product. Our strategy focuses on solving information overload, delivering fast, trustworthy insight, creating AI led differentiation, and driving engagement before monetization. So here are the standout features, news aggregation engine, statistics integration, profile setup, so you can have your own profile on there, personalization logic, insight summaries. So when we're talking about stats, we use AI to give you a bit more insight. These are the things that came out in our market research. And now, match change alerts, push notifications, basic analytics, premium insight, and then subscription management. And then what we want AI to do is to group these into logical themes or epics, identify dependencies between them, and suggest the logical order in which these themes should be built and explain the reasoning. And this is even before we create our roadmap. And the format should be theme name, included items, dependencies. Why this theme should be at this stage. So before we even build a roadmap, we know what kind of groupings AI is recommending and in more order. We can then play with that before we build the roadmap, and the dos and don'ts, do focus on outcomes, not task or sequencing. Do consider strategic priorities, don't create detailed user stories at this stage, and don't assume unlimited capacity in terms of building. So let's get started. So here are the themes it came up with. The first theme is core sports data Foundation. So this includes the sports news feed and the match and stats integration and why it comes first. It says we need reliable, structured sports data before we can even personalize or generate insights. Theme two is personalization and user setup, and this allows you to set up your user profile and then have content personalization in there as well. It says this comes second because the strategies are about reducing overload, and that only happens when content feels relevant. Theme three is AI powered insights. That includes AI insights summaries, and then match change alerts. Why? Because now we have data and we know that we know what users care about. We can explain what changed and why it matters. This is our main differentiation. Theme four is engagement and habit building, and that includes push notifications to the user and basic performance tracking and dashboard. Why? Because notifications only work if the content is valuable, and it supports forming a daily habit. So the user starts to have a daily habit, and they will use it every day because they have that habit. Repeat usage and measuring engagement and retention is something we're interested in. So we can do a lot more like this. And then premium and subscriptions include premium insights and subscription management. Why? Because our strategies are engagement first and then monetization. We only introduce the premium features once the users coming back regularly. And the insights clearly provide extra value. And we see signs of our product market fit. So only once we know it's working for the user do we start charging some kind of a premium. So the logical order this says is core sports data foundation and personalization, user setup, AI powered insights, engagement, and habit building, and then premium and subscriptions. And this keeps the roadmap aligned to the strategy of data, then relevant insight, habit. Oh, so this is the first part. It's the first part we've got. We've ordered some themes, and this can go into our roadmap, and we can mix and match things as we see fit. At least we've got some kind of order to how we should do things and good reasons why I think it makes sense. 13. How to Generate the Product Roadmap: Now we're actually going to create the roadmap, and so here's our prompt. The goal is to generate an outcome based roadmap for passion sports. The role that we want the AI to take on is you are an agile product manager designing a goal driven roadmap. Context is we want to use previously created themes on our strategic priorities and our strategic priorities, which are which we want to validate. We want to validate demands, increase engagement and create AI differentiation. We also want to introduce revenue after retention improves the users to remain in the product before we start thinking about revenue. And then the action is for each theme to define a single outcome based goal, to make the goal measurable, and three suggest appropriate metrics. Four propose a realistic time frame at quarter level. So we don't want to think in terms of months or days yet, just a quarter. Five list three to five high level features required, and the format should be a quarter. So which quarter this should happen in the name, the goal, the metrics, the date, and the high level features associated with that. Do use outcome based language. Do include measurable targets. Don't list detailed functionality at this stage. We're just at the roadmap stage and don't overlap more than one goal per phase. And that's that, so let's run it. Okay, so here is our roadmap in a text form. Have changed a few things. The only thing that I really change is that I've changed the inner text form. I have changed a few things. The only thing I really changed was the name for each quarter, because I wanted it to be a bit more relevant and something that the team could understand straightaway. So here's our roadmap. In oir quarter, quarter one, the first thing we want to do is foot and we're creating football with the CR UI, the core user interface. The product. The goal of that is to validate initial user demand for personalized football sports intelligence experience. So do people want this? And we're looking for 5,000 plus downloads? This is the minimum viable product. So 5,000 downloads, greater than or equal to 60% onboarding completion rate. So that's 60% of the people that go to onboard actually convert, greater than or equal to 30% week one retention of users coming back for more. And greater than or equal to three sessions per user per week on average, and that will happen in ir quarter. So that's months one to three, and the features in there are football, news feed, football, match and stats, data integration, user profile setup, team league selection and personalized football content, also basic analytics tracking. And then for third quarter, football with AI insights, and the goal is to increase engagement by introducing AI powered summaries that reduce the information overload for football fans. And then we've got the metrics for that here in a similar vein, and that's for ir quarter. And the high level features are AI generated football match summaries. What changed, and why it matters, insights for football match change alerts, prioritization improvements and inside engagement track. And then for ir quarter, we've got football plus F one, so Formula one. In this release, in this quarter, we're calling it multi sports. That's going to strengthen the habitual usage, making it more likely that the user is going to come and use this over and over again and validate the multi sports stability by adding Formula one content. And so metrics are 45% or more weekly active usage, five plus days back, and there are some other metrics down here as well. And the features are push notifications, which are insight driven as well, Formula one news and data integration, F one AI summaries, refined daily briefings and engagement analytics dashboard. And for firth quarter, the name of this one is premium UI. And the reason is because we're including and validating premium insights. In other words, you would pay for some insights here, data or something interesting that you wouldn't get unless you paid for it. So that's a subscription. We get subscription revenue without negatively impacting engagement because we've already given them so much value before, we've got our metrics here, and that's in o quarterur. And the features will be premium insight, AI subscription management, so you can change your management if you can manage your subscription. Advanced AI trend analysis, exclusive deep dive explanations and much more explanation about the content. This is what users said they wanted. And then revenue analytics tracking. So that allows us to track our revenue. So this is the current roadmap. It gives us enough to know these are the kind of things that we want either to release in quarters or at least to get done in different quarters. We can work out whether we're going to release it or not. And there's only one more thing that I would do is it's nice to have this in a much more visible way. So similar to what we did with the vision board, we're going to put this into an image or into a template that makes it easier to see what's going on where we can fill in the date, name, goal, features, and metrics. So let's do that in Chat GPT now. So let's go. So here I am about to create the roadmap. I'm selecting this image and prompt. See the below. See the below roadmap, which is down here. It looks a bit jumbled, but it's the same roadmap we created. Analyze the attached image, then create a new image using the roadmap text below as follows. And we have to be really accurate here, I've noticed. Otherwise, the AI doesn't get it right. So I've said put each quarter in the correct column of the date row. So we have to be accurate in what we say here. Each quarter in the correct column of the date so that's the date row and put each quarter's name information in the correct column of the name row. Put each quarter's goal in the correct column of the goal row. Goal row here, put each quarter's feature information in the correct column of the features row. That's here. Put each quarter's metric information in the correct column of the metrics row. So let's go, and here we go. Here's our roadmap. So if we click on that, let's see a bigger version of it. It's got the date important to us, our various names here for what our different releases, our goal I duplicated the goal. I'm not sure why it's done that, but that's something we can edit out. We've got the goals and features and the different metrics relating to each one of these quarters. So once we've sorted that out, we've now got a roadmap that we can use, and that's in a form that makes it really clear for everyone. Okay, so now we have our corrected roadmap in our roadmap template. So now it's your turn. What I want you to do is to generate your roadmap for your app or your product, whatever it is that you're building, and then ask, is this roadmap outcome driven or feature driven, and what would improve it? And that way, you can work with AI to find out how best you can improve your roadmap. Use the prompt I gave you earlier to actually generate the roadmap, and then when you're finished, you can always ask additional questions to test it and make sure that you've got the best roadmap possible. And that's that. See you in the next lesson. 14. How to Create Persona Segments: In this lesson, we're going to talk about how to create personas using the feedback and insights we've already got. So already, what we've done, we've done our market research. We've got our user comments, which we've analyzed. We've done a competitive analysis. We've created a vision and a strategy, and now we're going to answer the critical agile product question, which is, who exactly are we building this for? We've got some ideas, who we're building it for, but we want to narrow it down even more. And if we don't answer the question clearly, what we risk is building features for everyone, which usually means no so what we want to be really targeted about who is who we're building for, and then we can help them the best. So personas help us to build empathy with a particular type of person, make better decisions about the backlog, prioritize correctly, and avoid building for ourselves, which you can often fall into when you're passionate about something. So today, we're going to use Chat GPT to turn all our research and feedback and strategy into a concise, useful persona or concise, useful personas for the Passion Sports app. So what does a good persona look like? So a strong persona is based on real research, not guesses. It's concise. It has a clear primary goal. Good personas describe motivation, not just the demographics, and they help guide the product decisions. They can evolve as we learn more as they're not just fixed. So the first thing we're going to do is turn our research into insight clusters. So generally, what we do is we cluster around specific pieces of insight, and that helps us to generate the personas and we use Chat GPT to extract patterns from the comments and reviews, market research, search trends, and some assumptions about engagement. Who's going to engage with the product? So let's get into it. So we're back at hat GPT. Let's paste in our prompt. And if we zoom in slightly, our goal is to identify clear user segments and recurring goals based on research for the Passion Sports app. Role is, we're telling the AI, you are an experienced product manager helping define user personas from research data. The context is we are building an AI enhanced sports intelligence app that aggregates sports news and stats, highlights key insights using AI, personalizes content, helps users understand what changed and why it matters. So actually, it's probably got this context already. Currently, the research insights show that users feel overwhelmed by sports information. Many many use multiple apps. Users want faster understanding, not more data, and fans follow multiple teams or leagues. And the action is, number one, identify two to three clear user segments. Two, describe their main goals. Three, identify their main frustration, highlight the core reason they would use the app. And the format we want is a segment name, demographics at a high level, behaviors, the main goal of the persona, the main frustration and why they would use the app. So that really helps us narrow it down. Dos and don'ts, do base the segments on research themes. Do focus on motivation over features. Don't invent unrealistic details. Don't design a product yet, and I'm going to add on one more. Use also the attached vision board. So let's attach our vision board. Let's run that. Here we are with our segments. So the first segment is daily multisport follower. The demographics are 18 to 45-years-old and follows multiple teams or leagues. They check sports updates at least once a day, use two to four apps, follow multiple competitions across different sports and scrolls frequently, but rarely needs long articles fully. Their main goal is to stay fully up to date across their teams and leagues without spending too much time. Main frustration is information is scattered across apps and feeds, forcing them to search and filter constantly, and why would they use the app to quickly understand what changed and why it matters across multiple sports in one personalized place. A segment name for the next one is called a time constrained professional fan, and I won't go through all of these in detail, but you can see that we're starting to build up a pattern of different segments or different groupings or types of people who might have different needs. So if we just remember that this one was the daily multisport follower, the next one is a time constrained professional fan. 25 to 50-years-old, checks sports during commute, lunch or evenings, rarely watches full matches live, relies on highlights and summaries and wants reliable sources and accurate data. Their main goal is to get a fast, trustworthy overview of key matches and development. So if we compare that to this goal to stay fully up to date, you can see that there's a little bit that's a little bit different here. That this one wants a fast, trustworthy overview of key matches and developments, whereas this one wants to stay fully up to date across their teams and leagues. So you can see how there's some differentiation between the different segments, different types of people. And this goes on for all the different types of user that we can see. And in this case, we've got a daily multisport follower, time constrained professional fan, and then an insight oriented superfan whose main goal is to understand deeper insights. Not just scores, but implications and trends. So these people want more of a deep dive into particular things about teams rather than just like an overview. So a different age group. So now we've got our different segments. Overall, it says the strategic pattern across all the segments is all three segments share one core motivation. They don't want more sports content. They want faster understanding and clearer prioritization. And that aligns with the vision board, clearer context. AI as visible intelligence, invisible intelligence, more personalization and more depth into particular things rather than breadth. Rather than lots of different things, we want to go deeper into specific things. So that's how segments done. So the next thing we need to do is pick a primary persona, because the primary persona helps us think about who we're primarily building the at for. Even though we may have just three at the moment and maybe more personas in the future, we need to center around a primary persona. So let's go into the AI and let's run this prompt. Based on the identified segments, recommend one primary persona. Explain why and describe the risks of not choosing a primary persona. And notice I'm not doing the whole goal role action context thing because we have the context from before. It knows what we're talking about, so let's just run it. So here we go. We've got our results back from the AI. Number one, recommended primary persona is the daily Multisports follow up. So let's see why this should be the primary persona. So the reason is it says it's got the strongest alignment with the vision. The vision emphasizes cutting through the noise, personalization, multi sport coverage, daily habit formation. It actively follows multiple teams and leagues, feels overwhelmed. Checks sports daily and already uses multiple apps. So it aligns with all of these things in the vision. And then the largest immediate market opportunity. So it's not a niche segment. It represents digitally active mainstream fans, especially in the UK and globally with football first. So they're saying it's the largest market opportunity. Next one is strongest retention potential. So they check sports daily, which means they're going to be easier to keep coming back to the app. There's a bit of detail around that and best foundation for monetization. So these people consume content frequently, care about relevance, and they value the time savings. And then the bottom line is, choose the daily Multisports follower as the primary persona, designed primarily for them and ensure the other segments still receive value. But do not optimize around them yet. And then there's some sort of detail about them. So if we go back up to the top and just look through the personas, the question is, do we actually agree with that? It's always good to check ourselves. So I think, I mean, just reading what was below, I think it's right that it aligns totally with the vision. And if we were to think about a daily Multisports follower and then compare, if we were to build for a time constrained professional fan, the main or the next segment, which is an insight oriented fan, the difference is the main goal is to get a fast, trustworthy overview of key matches and developments and not necessarily depth, but an overview. And the goal for this one is to understand deeper insights, not just scores, but implications and trends, and this one is depth, whereas, if we go to the one that has been suggested, they're fully up to date across their teams and leagues without spending too much time. So let's think about that. So thinking about that, I think it makes sense because I think anything that helps people to stay in one app and fulfill their needs without spending too much time, I think, is going to be the winner. I think in the world we live in convenience is everything. And so the key thing that convinced me here much was time. And thinking about myself and the people I know, I think they'd rather have one app where they could stay fully up to date across their teams and leagues first before they start having in depth information about anything else. So for now, I think that makes sense. We can always change our mind, but that makes sense as our primary segment. So the next thing we're going to do is we're going to add the personas into a little template. 15. How to Generate the Product Backlog: Okay, so now in this lesson, we're going to learn how to generate the product backlog using AI. So now we've got our market research competitive insights. We've got a clear product vision. We've defined our product strategy. We've got our personas, three personas, mainly, and we've got a primary persona, as well, most importantly. Now what we're going to do is translate all of that into something actionable, something we can work with with our team to actually build a product. And that's called a product backlog. So the product backlog is where the ideas become structured, testable product work. We're not going to brainstorm random features. We're going to have a list of features that we're going to be working on that we're happily aligned with that are happily aligned with our vision and our personas. It will be using all the research, the strategy, the persona, and making sure that it's actually clear value for our users. To generate a focused backlog, we use what I call the three Rs of user story writing, the role, which is who it's for, the requirement, what do they need, and the reason why does it matter? And then we'll generate it in the user story format that we usually use, which is as a role, I want requirement, so that reason or return on investment. And so what that does is it makes sure that we're focusing on the key things that we need for the product, which is pleasing our user, matching the feature set. People in Agile don't like the word requirement, but I just use it because we know what we mean. We know it's a requirement that is actually flexible at this point. And the reason why we are doing it, we need to be focused on that and not lose track. So we're doing that all using AI right now. Okay, good. Okay, so here we are, and we're about to generate our product backlog. The goal is to generate a focused minimum viable product, a product backlog for the MVP, so a minimum viable product, product backlog. So we're just putting out something to test and make sure that our users actually want this aligned to our vision, strategy, and primary persona. The role is you are an experienced agile product manager, helping translate strategy into a backlog, into a structured product backlog. The context, use the primary persona and everything you know about the Passion Sports app. At this stage, we're just going to use the primary persona because that assures that we're going to please our key user. We're going to generate something for our most highest priority user, and we can always add in things from secondary personas. And the constraints is, this is a minimum viable product that we have limited development capacity. We must support learning and validation. I'm actually going to remove the limited development capacity because I don't want to think about development so much. The fact that it's a minimum viable product is enough, and it must support learning and validation, very important. And the action is to generate 12 to 15 product backlog items, epics. We're limiting it because it could, you know, generate too many or too few. Focus only on minimum viable product level capabilities. So it's very important that we're building things to validate that. It's worth putting this product out. We need to validate that. It's worth putting this product out in the first place. In short, items are testable and outcome driven. Avoid technical implementation language, and I'm actually going to change this slightly going forward, and you'll see how I do that. And the format is use proper user story format, which is, as I spoke of the three Rs as a role, I want requirement, so that reason. And that makes sure that it's all encompassing why we're doing this, who we're doing it for, and what we're doing. So return as a numbered list and do not prioritize yet. We want to think in stages and do that step by step. So let's go. So here we go. We've got our product backlog back. It says, Here's a focused minimum viable product, product backlog aligned to the primary persona of daily Multisport follower. And designed for learning and validation. So the stories we've got here, let's just skip through a few of them. As a daily Multisports follower, I want to select my favorite sports and teams during onboarding so that my experience feels immediately relevant. That makes sense. So now the user knows which sports and which teams the app is going to focus on. As a daily Multisports follower, I want a personalized home feed showing key updates from my selected sports so that I can quickly see what matters. As a daily Multisports follower, I want concise summaries of major events so that I can understand what happened in under a minute. So I want clear indicators of why an update is important so that I don't miss critical moments. I want live score tracking for selected matches and races so that I can stay updated throughout the day. Short AI generated insights explaining key changes, for example, momentum shift, standings impact, so that I can understand the bigger picture. Simple timeline view of major moments in a match or race so that I can catch up quickly. Optional SMRT notifications for major events, so I can stay informed without being overwhelmed. Daily top stories brief across my sport so I can get quick overview in one place. I want to switch easily between football and F one views so that following multiple sports feels seamless. Trending topics across my selected sports, so I can see what other fans are focusing on a lightweight way to give feedback on summaries and insights so that app improves based on my needs. I want to save important updates so that I can revisit them later. And I want clear distraction free interface. I want a clear distraction free interface, so I can scan information quickly. I want the app to remember my preferences across sessions so that I don't need to reset my interests each time. So just looking at that, that looks cool. It goes into the main things of personalization, speed of understanding, cross sport simplicity, and AI powered insights. So it's lightweight, but it's still differentiated from other apps, and it's got built in learning loops for validation. So, you know, it can learn as it goes. You can give feedback. So just at a glance, I think this is really good. And the beauty of it is it ties to our every all of this ties into our original vision of what we wanted for the app. So if we wanted to refine the backlog there, there are a few things we could do. We could review the Blog and remove the items that are not essential for an MVP, a minimum viable product, identify overlapping stories, suggest clearer wording if needed, flag any story that is too large, and I'm going to add to only do the above steps if they are necessary. Don't do them if they're not necessary because sometimes Chat GBT does things because you've asked it to look at things. So I'm going to tell it not to. So now it's your turn. I want you to do is use Chat GPT to generate a product back log for your app, whichever rap you picked and remember to use the three Rs when you're generating it. Use the prompt that I gave you earlier, but remember that it should also be generating stories in the form of as a role, I want requirement, so that reason. Then from there, you can start to work out why should what you should build first and why, and we'll go over the ordering and the structuring of that in a future lesson. But the main thing is now you know how to build a product backlog that remains in line with specific personas. You can add more personas when needed, and all of this stays in line with your vision and your strategy before you even build the product. So see you in the next lesson. 16. How to Generate the Product and Release Goal: In this lesson, we're going to learn how to generate product or release goals for our product. So we've now completed our market research. This product vision strategy, we've got a roadmap, and we've generated a product backlog. So what else is next? What we must understand is what we're going to achieve, not only in this product, but in our first release of the product. A product manager or product owner, one of the biggest mistakes you can make is building features without a clear outcome or a clear reason. For this course and in my own way of working, I usually set a product goal, and I usually set a release goal. These are often but not always the same thing. Now, the product goal often is aligned with the release goal simply because you may be creating some major update to the product which changes the direction of the product. At this early stage, our goal really is the same for the first release because we're creating a minimum viable product and that we want that to be our first release. And it's also what the product is about, or it's going to encompass what the product is initially. So different from the product vision is the next step down the product goal. So in our case, not always, but in our case, the product goal is the release goal, and every major roadmap milestone represents a meaningful step forward. Every release must deliver measurable value. So the product goal is going to give us tight alignment, measurable progress, and we can treat every milestone as really significant in meeting our product vision. So that's what we're going to do for the Passion Sports app. We're going to create a product goal and a release goal. First of all, let's talk about the product goal and the release goal. In some frameworks, the product goal is long term. The release goal is short term. In our approach, each major roadmap milestone is strategic. Each release is going to represent a meaningful product evolution. At present, that could change, but that's how we're thinking right now. And each release gets a clear product goal. This keeps the roadmap outcome driven. So let's generate our product goal. Here's our prompt. The goal is to generate a clear outcome focused product goal for the upcoming release of the Passion Sports app. Role is to act as an experienced agile product leader focused on measurable user and business outcome. The context is use the attached vision roadmap and everything you know about the Passion sports app. So let's remind ourselves of we've seen the vision, so let's remind ourselves of the roadmap that we created earlier. So here's our roadmap. And because we're going for a product goal and we're talking about the upcoming release, we should look at the fact that ir quarter, which is the first thing that we want to attack this quarter, we want to concentrate on football, and our goal is to validate initial user demand for a personalized football sports intelligence experience. Football news feed, football match and stats data integration, user profile setup, basic analytics tracking. We've got some specific metrics as well. 5,000 plus minimum viable product downloads, greater than or equal to 60% onboard incompletion rate, and greater than or equal to 30% week one retention. This is what we're looking for at ir quarter, but we've also got other quarters, AI insights, football, and the other sports, and then the premium UI. This is our roadmap. And so this is what we're going to be uploading into the AI to work out our product goal, slash Release goal for Release one. So we're going to attach our roadmap. We're also going to attach our vision used as part of the context. And the action is create one clear product goal for this release that focuses on a measurable user behavior, aligns with the business objectives, reflects the roadmap milestone and avoids listing features. The format is one concise sentence followed by three measurable success indicators. Focus on outcomes, not outputs. Tie it together, tie it to user behavior and ensure it guides the order of the balog. Do not list features, do not describe implementation and do not create multiple goals. Okay, so here's our product goal for the upcoming release, which doubles as the release goal for football the core user interface. So the goal is to enable daily multi sport football fans to build a habit of returning to the app to quickly understand what matters in football. So that's cool. We're saying that we want this our ideal primary persona to start building a habit, habit of coming back to the app. And the reason is so that they can see what matters in football. So it's loose enough that we could change what features that we want. And how will we know if it's worked? Well, 30% week one retention rate. 40% of active users open the app at least three times a week, 60% onboarding completion rate. So they start onboarding, they actually finish what they started. And this goal tests whether users come back regularly before we add AI features or monetization. I think that's fair. So I think we've achieved a product goal that aligns with the roadmap and the vision as well, so that's good. Good as a release goal. It doubles as a release goal, too. So a best practice thing to check when the AI returns as our release goal is that a strong product or release goal should be measurable. It should reflect business value. It should change the user behavior. It should guide prioritization, and it should fit on one line. If it reads like just a list of features that you're aiming to release, then that's not correct because actually what we want to do is we want to drive the decisions about the backlog based on the release goal. We want to keep going and looking at the backlog and then look back at the release goal and then say to ourselves, is our backlog of what we're going to deliver this release in line with the release goal? That's what we want. At first, I was wondering, is it measurable, but that's why we've got some stats underneath that that makes it measurable. So if we go back and look at our release goal, we can see that we've got these success indicators that make it really measurable. Usually, I do like to have these as part of the sentence. However, in this case, for now, I think this works. And we can always refine it back into the sentence if we want to. The main thing is people know exactly what we're aiming to do for this release. We've got flexibility with the features as long as they meet this release goal, and we've got something to measure it by. We'll know when it works as long as it hits these success indicators and it meets the release goal. So now it's your turn. Use Chat GPT or whichever LLM you want to generate a product or release goal for your app, whichever app you may have gone for. And when you're finished, ask, does this goal define success or does it just describe the work you're doing? It doesn't clearly define the success and it's measurable and all the things that we've said before, then keep refining it until it's a sentence. In a sentence, you can tell someone what defines success for your first release. So, go for it. I look forward to seeing what you come up with, and I'll see you in the next lesson. 17. How to Order the Product Backlog: This lesson, we're going to go over how to order and sort your product backlog of user stories. So writing user stories themselves aren't hard, but ordering them correctly is something that takes a bit of skill and understanding of what you're doing. So as a product manager or product owner, your job isn't just to manage task. You need to decide what should we build first and why, and always order the backlog with that in mind. So backlog ordering determines what value gets delivered first, how fast you hit your product or release goal, whether your roadmap makes sense, and whether your team builds strategically or randomly. And of course, we want to build strategically. We want to build strategically, not randomly. So in this lesson, we'll use AI to help us order the product backlog based on various factors like impact, user value, business value, strategic alignment, and release sequencing. And we're going to do that using everything we've already created for the Passion Sports app. So some things to think about when ordering a Blog, need to think about the impact on the product or release goal that we're aiming for. Value to the user, value to the business, alignment with the roadmap milestones, what dependencies do we have? And what's the feasibility of delivering certain things at this point. And this is where the AI can help, not to replace the backlog to maximize impact on the current product slash release goal. Role is to act as an experienced agile product leader specializing in value driven backlog prioritization. The context is to use the roadmap, vision, strategy, Blog, and everything you know about the Passion Sports App. And then bear in mind strongly the upcoming release. And the action is for each Blog item, assess the impact on the release one goal. Assess the impact on release one goal, and for all of these things, we're assessing the impact as either high, medium or low. We're assessing the business value, assessing strategic alignment, identifying dependencies, and then suggesting a recommended order. And then at the end, we're going to briefly justify the reasoning. The format we want this in is a table with a column, story, impact, business value, strategic alignment, dependencies, recommended order, and then reasoning. Prioritize alignment with release one goal, consider user engagement impact and factor in any dependencies. And the dos and don'ts do not reorder randomly, do not prioritize based on novelty alone, and do not ignore the roadmap sequencing. So we're taking a whole load of factors into account, and we're going to end up with a table which allows us to see using a high, medium or low score, how they factor for each one of these points. So here you go. Here's the ordered back log. Selecting favorite sports and teams during onboarding, we've said that's got high across the board, no dependencies. It's ordered as number one, and these are basically already in order of what it thinks is the most important. Then next is the personalized feed with major updates. It depends on onboarding preferences because once you've picked what sport you're going to be most into, then we can personalize the feed. Unified multi sport feed. Again, that depends on the personalized feed because once you can personalize the feed, then you can have one feed with lots of different sports in there. The next live scores tracking concise daily summaries, and we've always got what the dependencies are on the right, clear signal showing why an update is important, then short explanations of impact, chronological list of major moments, smart notifications. Remember, user preferences, and then the lightweight inept feedback. So now we've got a list of all the stories that we should do, ordered by impact, business value, and strategic alignment, and we know what all the dependencies are. So now we've ordered our backlog by those factors, we still need to apply human judgment at every point. So before we finalize the order of the backlog, let's ask, does this clearly support the release goal? Does it reflect the roadmap sequencing, and are we building the foundations before we enhance it? We don't want to start getting into the fancy stuff until we've got exactly what the user needs. We chasing trends or are we doing it for good reason? Are we biased towards existing, exciting AI features? We need to be careful of that. And the other thing is just to bear in mind that the backlog will change in order all the time. That's exactly why the product owner exists or product manager exists. We can use AI to do that, or we can do that ourselves, but this isn't a one time thing. So sometimes user feedback can change, the market data evolves, the strategy can change. And then there's new dependencies which change the order that we do things in. The main thing is now at this point, we've done our market research. We've got our vision, strategy, roadmap, and release goal, and the Btlog is aligned with all of that. So now it's your turn. What I want you to do now is use the exact approach that I've used and the prompt that I've given you to order your backlog for your app, whatever it may be, and ask Chat GPT or whichever LLM you're using, which Blog items directly drive the next release goal? Because those are the Blog items that you should be working on for your first release. And then you can refine it and say, what did AI overestimate or underestimate? Can actually think about that yourself, and you can use your own human intelligence to work that out and always update your backlog based on what you think is right. The AI is just there to help you. And that's where real product leadership happens because not only do you use the power of AI to get you there quicker, but you use your own smarts, experience, and intelligence also to lead AI in the right direction. So have fun doing that, and I'll see you in the next lesson. 18. Jira Setup Overview: Okay, guys, welcome. This is Paul from Passion Consulting, and I'm about to show you how to get started with Jira from scratch. I have done this before many times. However, this is my first time doing this on the latest greatest version of Jira. So I'm just going to do this just to show you how easy it is to do this from scratch. So here I am at at Lasian website, at lasian.com. What I'm going to do is I am going to get started installing Jira, actually not installing Jira, but setting myself up with Jira. So what we do is we go to products. We pick Jira software. Jira software is what we use for project teams. Jira service management is what's used for things like incident management and tracking services like website Let's Gun Live. So this is what we'll use because we're creating a project. And then this is Jira, so we'll go to get it free. And so it's already selected. So click next. And now we just enter Email. Click, sign up. Login to continue. So I don't know if it's gonna let me in because I haven't given it a password yet, but let's see what happens. I'm guessing it'll tell me to go back and find my password or something, or it's going to set me up with a password. Looks like it set me up. So I'm going to do is create a project, a passion Budget. So let's call it passion project. I don't like dashes, so just call it passion Budget. I do not want to receive promotions. Sorry at lesion. And this will just go on to create the URL that we go to when we want to get into Jira. So click Go to Jira. So it goes on to ask. Tell us a bit about yourself. So I usually tell it that we're a software development company because that's what I'm used to doing. I tell it that my role, my nearest I'm actually a scrum master in this case, but let's say project manager because that's the nearest. What are our main to us. So I'm going to say, coding, writing reviews, project planning. I'm going to say productivity tracking and reporting, and I'm going to say testing and design. I'm not sure if it uses any of these. It might just be feedback for them, but let's just fill them in, give them some help. They're giving it to us for free anyway. Then it says, select a template for your first project. So sometimes we use Cam Ban, but this is a Scrum project, so let's choose scrum. And there are a lot of things you can do to still set up a Cam Ban style board anyway, so we don't need to worry about not having the board. So we chose a scrum. So now it's going to create our account. A simple as that, easy as one, two, three, we're in. 19. Jira Menu Overview: Welcome to Jira. Which best describes you? I'm not familiar with ira. I'm somewhat familiar with Jira. I'm very familiar with Jira. So I like to start like a beginner just to see what it says. And because many of the people many of my students and the people watching this are going to be not familiar with Jira. I'm going to pick that one, and I'm going to say start. So it's going to give us a bunch of clues as to where to start. Well, because I've done this before, I don't necessarily need these. So the first area to look at here is the backlog, which is the product backlog. Now, before I go into that, I'll just quickly draw your attention to the side. So on the side, we've got a timeline, which is essentially going to be our roadmap. It used to be called Roadmap, and this will show us how far along we are in our roadmap, how many sprints we've done, and some titles which show us the outcomes that we've achieved. The Blog, as we said, is a list of all the features that the product owner requires for the product. And then the board is where we'll go on to work with the team to go through our workflow stages so that we know when every task and every product backlog item is complete. 20. How to generate a CSV Product Backlog from ChatGPT for Jira: Okay, now we've helped our product owner to create the product backlog, and here it is. We've got a bunch of epics, further refinement necessary, yes, but we're at the first stage. What we want to do is we want to get our product backlog into a tool of choice. Every team uses a different tool, and for our team, they're using the popular tools of Jira and Confluence. What we want to do is we want to get this into a format where we can import this into Jira. We asked Chat GPT to do that and to tell us exactly how to do that. A simple instructional prompt, change this backlog into a format we can import directly into the latest version of Jira. I'm careful to tell it that it's the latest version of Jira because Jira updates very frequently, and often Chat GBT has enough information to know if something's changed and it needs to change the format. Let's run. It turns out that as we're using the free version of Chat GBT, it says, No available models. No available models support the tool in use. Try starting a new chat instead or try again later. That means we've run out of credits, and we need to start a new chat. This is a good excuse to show you what to do if you get in this situation. If you're working for a business, you probably won't, but what you need to do in this situation is start a brand new chat and make sure that you've copied everything relevant. The other thing that I would say is this is a good excuse to start a project. And what a project does is it keeps the context of all the other chats going forward. It may not have done it now because we've run out of context. We still create a new chat, but what we'll do first is we'll create a project. And the way to do that is to go here on the left, new project, and then we'll give it a name. This is the Passion sports website project. Excellent. And what we can also do is we can move that previous chat into the project. Going forward, it will remember all the context of what we've done and save us repeating ourselves. If we do repeat ourselves, we'd get even more specific. There you go. We've dragged the product goal writing chat into the Passion Sports website project. Now, anything that we do in a new chat will be able to reference information in this chat. For the avoidance of doubt, were placed in the epics, and we'll go back and we'll grab the instruction, and we'll change the instruction to say to focus on the epics below. Change the below backlog into a format we can input directly into the latest version of Jira. There you go. It tells us to input into the latest version of Jira, we need a CSV with an issue type, Epic name, summary description, priority, and labels. And here is the CSV that we can copy. It gives us some notes to say that priority is optional. Labels help filter. Epic name is what Jira used in the Epic panel and boards. The format is UTFA. It says it can do us an Excel format. I know the CSV will work fine. Let's stick with that. Let's ask if Track GPT can generate the file for us. And there you go. We've got ready made CSV that we can download that we can import straight into Jira. Let's take a look at it. Let's download it. There you go football Epics Jira Import. Let's find that, open it and go to the LibreOffice. Open that. And there you go. Here we've got a bunch of epics with an epic name, summary, and a description. ChatCPT has given it priorities, which the team can work out for themselves whether they're true or not. It said, like, I want real time football scores that I can stay instantly informed during live matches and given that a high priority, whereas team and player statistics, I can evaluate performance at a glance, it's given a medium. These are the product owner can double check. You know, Chat TBTs used its own mind on that one. And it's labeled everything as football because we've got cricket epics to come. The important here is, as we import this into Jira is knowing what these mapped to in Jira. Now let's go over to Jira and import these. 21. How to Import a Product Backlog into Jira: Okay, we have helped our product owners generate a product backlog in Chat GPT. We've exported it to a CSV. And now what we want to do is we want to import that into Jira, so the whole Scrum team, whole Agile team can see the product backlog that we've generated, and we can start to refine the backlog. The key thing we need to know is number one, how to import it into Jira and number two, how to map the fields correctly. So correct information goes into the correct fields in Jira. Let's go over to Jira. Here we are our Jira Batlog. Now, the way to import these so that the backlog items appear here in our backlog is first of all, Jira is always updating and changing. It turns out that there are more than one way to get here. And depending on your project, you'll either have to go up to settings in the corner and then go to system, and then you'll be able to find a way to import, or you need to go you'll need to create a space, even if you're not creating a space and then go down to import data, and that lands you here. Once you're here, you need to switch to the old experience currently. There may be a way to do it in the new experience, but they're changing the experience for how you import, and they advise you while we work on building more capabilities, we recommend that you use the old experience to import data. Let's switch to the old experience. And in here we want to select CSV. And then here we want to choose our CSV file. And there it is. We don't need to select anything in advances fine as it is. It was generated in UTFA and then click Next. Choose the project. We're going to put it in our passion Sports website project. We don't want to change any of the email stuff or the date format, so just click next. This is where we need to map the correct field. If we go and look at our CSV, we know that we want essentially to do two things. The way Jira works is it has epics separately from stories, and we map stories to their specific epic. These are actually story descriptions, and these are actually the names of our epics. And the way it works in Jira is that we're going to need to import these epics separately first, and then we're going to import these stories and relate them to an epic. And that gives us the advantage that as we create more stories, we can relate them to the same epic. If we had more than one story that relates to football scores, as we often do in the real world, there might be a table scores tables for football with different stories. As we break these down, we can relate them to the epic. Let's generate a bunch of epics first. And then we'll come back and generate our stories. The way we generate epics is we're going to need an epic name which has pretty much got correct here. So let's map that. Epic name in our CSV is going to map directly to Epic name in Jira. And then the issue type is going to be one of Epic. I happen to know that in Jira, they've updated it so that it's actually called W type now. Keep an eye on this if this changes, but right now, the issue type actually maps to what we call work type, and that will be one of epic. And that's really all we need but in the description, when you're creating acceptance criteria, it's actually handy to have something there to start off with. And if we flip back to the CSV, what I think works quite well for epics is this summary. So we'll use that as our description for the epic. This description will be for our stories that relate to the epic, as you'll see. So now, we'll use this as the description for our epics. We've got priority here. Let's see if we can map that across. Okay, we're not mapping a description because that's going to be for our stories. The epic name in the CSV is going to be the epic name in Jira. The issue type in our CSV is going to map to work type in Jira. We're not labeling it at the moment. We could do that. Let's just do that as we've got the information. Let's map that across. The priority in the CSV is going to map across to the priority in Jira and what's called the summary in a CSV. I can see here it said that a Jira summary field mapping is required to enable the input. If you go in here, summary will now map to summary. Let's go. Now it's asking us CSV filled issue type is being imported as issue type. And the value from the input that it's getting is epic. In Jira, that's also going to be epic as opposed to any of the others. And that's correct. What we're importing is not an asset, a bug, a feature, story sub task or task. It's an epic that's done that correctly. Click Begin Import. You can see it's done the input, but it hasn't, bringing in the epic name. Let's go across and see why. As we can see, there are a number of epics have been created here. Let's have a look and see what it's done for them. If we choose one of the epics, you can see it's got a title, and this title is what we're going to use to refer to our stories. That's giving us everything we need. We don't really need to worry about anything else at this time. We're not worried about that error. As long as we've got a number of epics here, they're named in a way that means when we come to our backlog and we our stories, they'll appear here with the correct epic name, that's all we're concerned about. Let's move on and import our stories now. Some of the names here are actually not correct. For importing our stories because we've created epics and Jira stores things stores the epics separately from the stories, which is correct. We've done them all as one, and we've related the stories to the epic names. But what we need to do is create a separate file that actually shows these stories with epic names. Let's do that now. What I'll do is I will save a copy, and we're going to call this football stories. Okay. So let's close this one. Let's open up our stories CSV. Okay, great. The first thing we want to do is we want to change all of these to issue type story. That's that. These are stories with a specific epic name. The other thing we want to do is we want to switch the summary and description because they're called opposite things in Jira. This is actually the summary, which is what we see at when we open up our backlog, the first thing we should see is a list of things that look like this, which is the summary. And the description gives us more information is usually where we put our acceptance criteria. And what we're going to do is we're going to expand this out to create acceptance criteria. Switch these around. But that's much better, and we can leave everything else the same. Let's save that we're going to keep it as CSV format. Now, let's go back to Jira and this time, we're going to import our football stories. The way to do that is exactly the same way we did the epics. We go to spaces, create space, even though we're not going to create a space. And down at the bottom, we've got import data. Switch to the old experience. CSV choose our file. This time we're going to choose football stories. Next. The project that we're importing into is the same project. Change nothing, hit next. And now we can do our mapping correctly. Description in the CSV maps to the description in Jira, their epic name, should map to the same epic name. The issue type in the CSV will map to the work type in Jira, and that's now called story. Labels will be the same. Labels in the CSV maps to the labels field in Jira, priority, same maps of the priority field in Jira and summary maps to the summary field in Jira. Everything's mapped across now nicely. Let's go next. And now it's just checking that the issue type will be one of story, and that's correct, too. Let's begin the input. Again, it cannot input the epic name. I've got some issue here where it's locked. Because it's not allowing stories to have this epic name. At the moment, I don't think there will be a problem, so I'm gonna proceed. We'll go to the next stage, have a look at our board. And our backlog. And there we are. We have a number of stories all imported in if we have a look. This is the which is why we need to switch those columns around so that it's in the right place. And this is the description that we can later expand into acceptance criteria. Now we've got everything imported in. We can move on to the next stage. The only thing that hasn't happened is the epics are not set, and this is what I was trying to do on import. Now we can set those individually. We would just go through one by one and set them either here or up here to their correct epic. If you view all epics, choose a parent, you can see in here, for example, this one's related to football scores, and you choose real time football scores here, click Done. You see it appears here. Now we've got our stories imported, and we've got our epics imported that we can relate them to. We can go onto the next stage and start refining our backlo. 22. How to Generate Acceptance Criteria with Rovo AI: Okay, so now we have a full backlog imported into Jira. It's time to start refining, but before we do, usually, we need to help the product owner to develop acceptance criteria and to refine the acceptance criteria a little bit just to make it really clear for the teams. And these are the kind of things that we do usually before we even go into product backlog refinement. So as we can see on our right here, we've got a story. As a football fan, I want real time football scores so that I can stay instantly informed during live matches. That's excellent. However, in here, we have no acceptance criteria. And this is where we can use the power of Jira and ovo to generate acceptance criteria and refine it, as well. Really powerful. So Jira is where we manage our workflows and manage our backlog and everything that goes along with it so that we've got user stories and our product backlog in here. And ovo is essentially the feature set that allows us to work with Lysian Intelligence, work with the AI and the LLM that Asian has created in the background, to use all of that knowledge about our project and wider knowledge as well, from the LLM to automatically generate acceptance criteria and many other things. So let's use that right now. So the way that we're going to access the AI is by typing in Ford Slash. So we can do that here. And here you can see there are many different features that we can access. But the one we're going to use is Ask ovo, because that allows us to type in prompts that will generate whatever we want in this space here. So if you click Enter on Ask ovo or just click on Ask ovo, what it will do is it will bring up here where we write with ovo. In other words, we can type our prompts in here. Or we can choose one of these options, and this will use AI to generate whatever we want from one of these options. So, for example, here, all we've got here is it says real time football scores, and we're going to choose improved description and see what it says. So you can see here it's now saying, first of all, generated a user story. We kind of already got that up here in our summary. As a football fan, when there are live matches, I want to receive real time football scores so that I can stay instantly informed. Up here, it says a football fan. I want real time football scores so that I can stay instantly informed during live matches. So it's generated its own version of the user story description. Context, this issue addresses the need for a feature that provides real time football scores to fans during live matches. Acceptance criteria. The system must provide live updates of football scores. Scores should be updated in real time as the match progresses. So I like part of this. We don't need this part. I like some of the context, and we need much more acceptance criteria. But it's a start, so I'm going to replace what's already in there, which is this with what Rover has generated here. So replace. So I'm going to remove this part because we've already got the description. We don't need the word context, but we can keep that this is what the issue addresses. We'll keep the title acceptance criteria. We don't need other information. So now we've got a couple of bullets. I'm going to say I want to expand this acceptance criteria in more detail. So what I'm going to do is again forward slash. But this time, I'm going to ask Rovo because I want to be specific with a prompt that does exactly what I want. So I'm going to say, Okay, so I've changed this to say, update the existing acceptance criteria and expand it with more bullets, giving a detailed description of what football scores panels should be displayed and what their functionality should be. So now we just hit Enter and see what Rovo comes up with. So here we go. So now we've the system must provide live updates or football scores. Scores should be updated in real time as the match progresses. The score panel should match the current score for both teams. The panel must show the time elapsed in the match. It should show the names of the teams playing. The panel should indicate which team is playing at home and which is away. There should be a section for displaying key match events such as goals, yellow cards, and red cards. The functionality must allow users to click on team names to view detail statistics and player information. Score panel should refresh automatically without requiring a page reload. Users should have the option to customize notifications for specific matches or teams. The panel must be accessible on both desktop and mobile devices for ease of use. So there you go a lot more detail there, so I'm going to insert that. So here's our acceptance criteria. What we don't have is any performance related data, and we don't have any non functional criteria. So, in other words, things that you may not be able to see, but you'll know that the performance is better. So I'm going to ask Rovo to do that. Forward slash again. And I'm going to say as a prompt. So now we're going to grab our product goal and put that in here. So I've said, add some non functional slash performance related acceptance criteria. Ensure it also aligns with the product goal, and then I'll put the product goal in. So let's see what Rovo does here. There you go. So now it says the score panel must load within 2.5 seconds under normal network conditions. The system should handle a minimum of 10,000 concurrent users without performance degradation. The score updates must be delivered with a maximum latency of 2 seconds. The application should have a 99.9% up time to ensure reliability during live matches. Core panel must be optimized for both mobile and network devices to ensure a responsive design. The system should be able to scale horizontally to accommodate spikes and user traffic during major matches. The data refresh mechanism should consume no more than 1% of the total bandwidth available to ensure efficient data usage. So again, this is using our product goal, which is why it's got the 2.5 seconds speed here. I like that it's got stuff around concurrent users, which is exactly what you'd see in industry. It's got stuff to do with latency, 99.9% uptime, so uptime stats that other kind of things we do talk about technology, industry, data refresh. So I think this is very professional. Again, in a real situation, a product owner would cast their eye over this, and so would the team to make sure it's all doable and also that maybe we want higher standards, maybe we want 99.99, but for now, I'll insert that. I'm going to give it a different heading. And now that's looking pretty good, pretty large, but we would definitely split this into multiple stories. And on top of that, this is just an example to show you. It's not necessarily we would have all this detail in here, but this is just to show you how we would get there. And so, also, the other thing we want is some edge cases. You could think of these ourselves, but as a starting point, we can get a lot of this in ovo done by Rovo because that knowledge is already out there. And so why waste our time trying to reinvent the wheel? So let's put in some more acceptance criteria. Okay, so I've said, add some edge cases under the title Edgecases. These should be based on real known edge cases for a website working over the Cloud. So I've given it some more context here. Let's see what it comes up with. So look at that. Lots of detail Edgecases. Users with slow Internet connections may experience delays and square updates leading to a discrepancy between real time events and display scores. So it's not actually telling us what to do. So I think it's interpreted what I've said as just tell me the edge cases. What I really want to do is I want to put in some acceptance criteria for what to do in these edge cases. Okay, so I've updated it to say some criteria for how to handle edge cases under the title Edge cases. These should be based on real world knowledge of edge cases and best practices for how to handle them for website or cloud related websites, especially for sports websites. Let's hit Enter. Let's see what Rovo comes up with. So edge cases, the system must handle scenarios where a match is delayed or postponed, providing users with timely notifications and updates. In the event of a network failure, the score panel should display the last known score and a message indicating that updates are temporarily unavailable. The application should gracefully handle cases where a match goes into extra time or penalties, updating the score panel accordingly. So, you know, we've got a number of different edge cases here. We can pick and choose. We don't need to necessarily use all of this. But as you can see, there's a lot we can take from that, so let's insert it. And there it is. Save that. Now, obviously there's a lot in here. What we could do is make it more concise. We can leave it as it is, or we can make it more concise. This, you know, looks pretty detailed, and the fact that we would need to split this down is actually irrelevant at this point because it's meant to be an epic anyway, a huge story. But let's say if we did want to slim this acceptance criteria down, we could just get ovo to do that, too. So double click to get in. I'm going to copy this for safekeeping. And we could say to Rovo. So let's say we wanted to reduce to the top three performance constraints and be more concise in the acceptance criteria in edge cases. It's a little bit wordy now because of the amount in there. Let's see how we get on with that. There you go. So we even made that a lot more concise. As I say, I think we're fine leaving things as they are. We don't really want to lose any detail. We can just split into different stories. I'm going to discard that. And we're just going to keep with what we've got and break it down more later. So there you go. Now we have a full story. We've helped our product owner to improve the acceptance criteria, get it much more detailed, performance, edge cases, everything that you need, really. And then this is even before the team see it. So we can now go to the team. Hopefully, they'll have a lot less work to do, but they can always add their expert eye over everything. 23. How to Organise Stories into Sprints: So in this lesson, we're going to talk about how to organize stories into sprints for the upcoming sprints. We've got our release goal, and ordered product backlog. We've refined everything, so we're using the invest principle for our user stories, and we've broken our stories down, not necessarily into tasks, but we've broken our stories down smaller. Now comes the practical question, how do we divide these stories into real sprints? As a product manager or product owner, this is where planning meets reality. We're preparing for a release, and we need to decide what fits in Sprint one or Sprint two or whichever sprint we're in. And today, we'll use Chat GPT to simulate two professional agile planning methods, velocity driven planning and capacity driven planning. Now, velocity driven planning uses the velocity, so how quickly the team are working, how many story points usually they're able to deliver in each sprint, and we want the average of that, which is the average velocity. And capacity driven planning uses the real task hours and is usually used to commit to a single sprint. Velocity driven planning usually is for forecasting future sprints. So we're going to use two different methods to try and predict which story should fit inside our sprints. For a forecaster for many sprints ahead, that's just to get some idea to show our stakeholders the current plan for the release. We're going to use Method one to show our stakeholders the current plan for the release, velocity driven planning, and this assumes the best predictor of future performance is past performance, and we use a historical velocity to divide stories into sprints. So to do that, we're going to need some data about previous projects. So let's go and look at some example data of the kind of things we would use to predict what stories are likely to go into which sprints. So here we can see the team velocity for the last three sprints, and the assumptions are that we're doing two week sprints. We've got a stable team composition in that time. So this isn't a time when we had lots of movement, lots of different team members coming in and out of the team and some variability due to onboarding and integration work. So there's always slight changes, but not massive changes to the team. So now we've got here, we completed 26 points, Sprint two, 28 points, Sprint three, 24 points. We know that the average velocity was 26 points, 26 story points. The variance is low and the team is relatively stable. So 26 story points is a reasonable planning baseline. A reasonable number of points we can use. I've also got sprint capacity data, and we won't get into that just yet because we're using Method one. So now using this information, we're going to work out which stories go in which sprints. So the goal is to divide the Btlog into sprints using historical velocity. The role is to act as an experienced agile product owner performing release planning. The context is average team velocity is 26. So yeah, so it says 20. That should be 26 26 story points per sprint. So the context is that the average team velocity is 26 story points per sprint, and bear in mind the release goal and ordered Btlog previously given. The action will be to allocate stories to Sprint one and two. Do not exceed 26 story points per sprint. Preserve Btlog in priority order and explain the reasoning. And I'm going to add to be as concise as possible, the format should be to return Sprint one, Sprint two, remaining Btlog and release progress summary. So we're going to be able to see what goes in Sprint one, what goes in Sprint two, what remains, and what's remaining. Let's go. So here we go. Very quickly, it said sprints one, Sprint one max 26 story points, and it's given us these stories for Sprint one in priority order. If we add the points together, it should not come to more than 26 story points. Five, 13, 18, 23, 26. So it actually totals 26 story points perfectly. We're going to have team selection during onboarding, a personalized football feed, live match score tracking, and concise daily event summary. Importance indicators, what matters. So in here, we've got our first five stories in Sprint one. Now, obviously, in real life, there'd probably be a lot of other things that we'd be doing here within even our first story. If this was a new app, we'd probably be setting up our environment and things like that. But this is just to give you a flavor for what we can do, and it's given some reasons for why we pick these stories. This delivers the core habit loop. So, in other words, the users, the user is going to form some habits coming in every day. After they've selected their team, they've got a personalized feed and live scores. They've got a reason to come back every day and form a habit. Daily utility live scores, fast understanding and clarity all in the first sprint. So that's the reason why we've gone for that, and that all supports the release goal as early as possible. So the AI is aware of what we want to achieve, and that's why it's picked that. So now we know what's in Sprint one. Sprint two, we've got AI Impact explanations. So using AI to explain the impact of the stats, match event timeline, smart match notifications. I will persist the user preferences. So once you pick them, it will keep them. Edit your favorite team after you onboard. Edit your favorite team after you onboard, and then you get inapt feedback on your insights. 813, 18, 20, 23, 26, and then the remaining batlog nothing. So all 52 story points fit cleanly into sprints based on the historical velocity. Then we've got a summary of progress. After Sprint one, we'll have the core value proposition done, habits formed by the user, testable and personalized football experience. And after Sprint two, the AI differentiation is added. The engagement mechanism is active and the feedback loop enabled. Minimum viable product ready for measurable release goal. So that achieves what we set out to achieve according to doing it using method one. So method two is to use capacity driven planning using the real task hours. And this method says, based on real task estimates, what actually fits inside the sprint. Now, usually we would only do this for the upcoming sprint to give us some idea of what the team is happy to commit to or take on for the upcoming sprint. We wouldn't usually use this for many sprints. However, I'm just going to show you exactly how it works, and you can use it as you see fit. So the sprint capacity for the team is the amount in hours that the team can take on. So the assumptions are two week sprints. We've got five developers and ten working days. We've got six effective hours per day because even in an eight hour day, lots of other things come up like meetings, et cetera. So in Sprint one, we've got five developers working across ten working days because it's a two week sprint. So 510 working days, 6 hours per day. And then calculation is five times ten times six equals 300 hours total capacity. Said Sprint two, assume one developer on planned leave for two days, slight overhead for integration testing, and then effective adjustment is four developers full time and one developer working eight days instead of ten days. So the calculation here means that there's two means that there's 28 8 hours of total capacity in Sprint two. So we've got less hours of capacity in Sprint two. So the summary here shows sprint 1,300 hours sprint to 28 8 hours available of development time. So if we go back to our backlog, what we can see is that we've got hours recorded that have been estimated. So assuming these have been broken down into tasks, we've got the amount of hours worth of tasks next to each one of these items. So now what we can do is we can forecast how many stories can fit into each sprint based on a number of hours. The goal is to divide the backlog into sprints, use real task level estimates. The role is to act as an agile product owner and development team performing capacity based sprint planning. The context is use the capacity in hours and the total task hours for each story up here in the table above, and the action is allocate stories to Sprint one and two. Do not exceed the sprint capacity. Preserve the Blog order and highlight overall allocation risk. Compare outcome to velocity based plan. And so by the end of that, we should know which of these stories fit into which sprints or if we're according to task hours, if we're taking on too much. Let's do it. So it said Using Sprint one capacity of 300 hours, Sprint two capacity of 28 8 hours, preserving the Batog order and using total task hours from the earlier table, it says that Sprint one actually everything fits inside of it, and it's a total of h 2 hours capacity, 300 hours and remaining, we've still got 98 hours left over. So we could actually take on more stories. And you can see it's, you know, kept the order. It's told us how many hours each is meant to be how many hours each is meant to be. So this kind of demonstrating this is kind of demonstrating what we know about capacity driven planning is actually it's not a good use to be able to say how much can be taken on at this point. Things often change from one sprint to another anyway, but according to this one, we could take on absolutely everything and do it all in Sprint one. In Sprint two, it would be completely empty because we'd have done all the work in Sprint one. There'd be nothing remaining, and the observations are there's a significant under allocation the total batlog effort is two oh 2 hours total to sprint capacity is 58 8 hours. And this minimum viable product fits comfortably within one sprint according to this way of doing it. So the velocity plan requires two sprints, and the capacity plan, everything fits into one sprint. And even the AI says that even though the hours seem to fit into one sprint, do not compress into one sprint. So the velocity based plan is strategically safer. This kind of reflects what we always thought, which is that we should really use the capacity based planning for the upcoming sprint. We should talk about it in sprint planning where there's a lot more detail. We may get a lot more tasks, and they'll likely take up a lot more time than we think at that point, and then we can make a commitment as a team for that sprint. And then just use the velocity based planning for future sprints. Now it's your turn. What I want you to do is take your own backlog for your product. Use the historical velocity to forecast, use real task hours to validate it, or to see what you're going to commit to for the upcoming sprint that you're going to be working on and allow the team to verify and update this prediction because this is literally a prediction, just to kind of get you on your way and maybe save some time in a meeting, but it's not the final commitment by any means. There are real humans involved, and they need to verify everything, and that's how professional product owners and product managers use AI as a decision support tool, not as the final decision maker. So, have a go, enjoy and see you in the next lesson. 24. How to Pick a Sprint Goal for the Upcoming Sprint: In this lesson, we're going to learn how to pick a sprint goal for the upcoming sprint using AI. So now that we know how to generate a product backlog, order it by impact and alignment and organize stories into sprints, one mistake that many product owners and product managers make is they fill the sprint with work, but they don't really define the clear outcome. And everything needs to be targeted and aligned just to make sure that we get the right product. The end. So a sprint goal is what we use to make sure that everything's aligned for the sprint in line with the goal, the release goal, the vision, and the strategy that we have for the product. And the sprint goal isn't just a list of stories, it's the real purpose for the sprint. So in this lesson, we're going to use AI to help us align that purpose for the sprint. So the sprint goal will help us identify the right focus for the sprint, align that focus with the release goal, and ensure the sprint delivers meaningful progress, and we're going to do that for the Passion sports app. So now that we've defined the product and release goal, we've ordered the backlog based on impact and alignment and allocated stories to Sprint one and even Sprint two provisionally, a good sprint goal is outcome focus instead of task focus. So what's the outcome that we want? It's achievable within the sprint. It contributes to the release goal. It provides flexibility in how the stories are implemented or wiggle room. So it doesn't mean that you have to stick to the exact same stories if things change. Although you should try to, but if things change, you can adapt it in line with the sprint goal, and it helps the team to make decisions if the scope changes or anything else changes as long as they align with the sprint goal. And so if anyone asks, What are we doing this sprint, the sprint goal would answer that question. So let's get into it, and let's use AI to help us do it. Goal is to propose a clear outcome focused sprint goal based on the stories allocated to Sprint one. The role is you are an experienced agile product coach. There's no real reason why we have to say product coach instead of product leader, but a product coach is usually more senior this just shows that it's someone experienced in product from an agile point of view. The context is we have decided to use the velocity driven plan and the stories that fit in Sprint one for this plan. If we scroll up, we can see that the velocity driven plan that we did average 26 story points per sprint. And so we decided that at this stage, we would use that experience that we have because our team has done a few sprints for other things before. So we're doing the team selection during onboarding, personalized football feed, live match score tracking, and concise daily events summaries, and then the importance indicators that tell the user what matters. So that's Sprint one. And that's what we're going with here. And we've said, bear in mind the release goal and everything you know about this passion sports app, as well. We're doing this all in the same conversation, and that's why we can say if you started a new chat, you'll need to upload or paste in release goals and all of that stuff. Or if you switched on memory, it will remember in between conversations. And the action. So one, we want to analyze what outcome these stories collectively enable. Two, we want to suggest three possible sprint goals. Three, we want to recommend the strongest one and explain why. And the format should be three sprint goal options, clearly marked the recommended option and provide short reasoning for each, and we want the AI to be as concise as possible while doing this. So what we should end up with is with three options we can pick from, and we can look at the recommendations and make our own decision. Here we go. So the analysis is that Sprint one stories collectively enable that the users can select and manage their favorite team. They receive a personalized football feed. They see live match scores and timelines. They understand the match importance and the AI explanations, and they receive smart notifications. And the core feedback loop is enabled, so we can get feedback from them, and then we can use that to develop the product more and then get more feedback. So the outcome enables that a user can on board, follow their team, and experience a complete personalized match experience for the first time. So the sprint goal options are sprint goal is to enable users to follow their favorite team and receive a fully personalized live match experience. Option two is the sprint goal is to deliver the first end to end personalized football experience from onboarding to live match insights. And option three is to validate that personalized AI enhanced match tracking increases user relevance and engagement. So the recommended sprint goal of these three is that we say that the sprint goal is to deliver the first end to end personalized football experience from onboarding to live match insights. And the reason why given it's given us is because it focuses on endo and outcome, not features. It clearly ties to a release goal, and it emphasizes working product, not components and provides a clear sprint review narrative. So something at the sprint when we get into the sprint review event, that's what we'll be talking about, and it leaves room for learning without prematurely claiming validation. Option one is good, but it narrows us down more because it talks about features. And Option three assumes validation that may not yet occur within one sprint. So validate that personalized AI enhanced match tracking increases user relevance. Yeah, we can't do that. In fact. I'm not sure why it gave us that option. So if we go back and we look, I don't think it's definitely between option I think it's definitely between option one and option two. I really liked option one because it says, enable what we enable them to do to follow their favorite team and receive a fully personalized live match experience. Or I agree. So if we look at option two to deliver the first interim personalized football experience from onboarding to match insights, that actually that leaves enough room for maneuver that as long as we do have a personalized football experience, we've achieved our goals from our vision and for our release goal, and onboarding and live match insights is enough to say, Well, as long as you can onboard someone, give them some insights and personalize it for football, then you've achieved your goal. Whereas this says you should specifically be able to follow your favorite team. And I guess following your favorite team is the part where it really is explicit to that feature that you're building. You may want to change. This totally depends on how important it is for you to reach your business goal of doing this specifically. If you really need to reach this business goal ahead of the release goal, then go for this one, I'd say. But if you want to allow yourself wiggle room, and in most cases, you do, then I'd go for this one. So let's say we go for this one, in this case, so now it's your turn. Before your next sprint, pause and answer, What outcome are we trying to achieve? Which stories collectively enable that outcome? If we delivered all stories, but no outcome, would it matter? It should. Does this sprint meaningfully move us towards the release goal? If you can clearly articulate the sprint goal in one sentence, you're ready to start the sprint with focus and confidence. So keep refining until you can do that and you've achieved your goal. So have fun with that, and I'll see you in the next lesson.