Ultimate Google Analytics course + 50 practical tips | Pavel Brecik | Skillshare

Ultimate Google Analytics course + 50 practical tips

Pavel Brecik, Web Analytics Evangelist

Play Speed
  • 0.5x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 2x
22 Lessons (4h 40m)
    • 1. Introduction

      2:07
    • 2. Basic data description - dimensions and metrics

      1:49
    • 3. How the measurement is done

      7:35
    • 4. Bounce rate, time and uniqeness

      6:12
    • 5. How to use the tool

      21:29
    • 6. What to remember when doing analytics

      1:02
    • 7. Socio-demo data

      11:57
    • 8. Browser data

      7:52
    • 9. Mobile experience

      26:02
    • 10. Site content

      24:09
    • 11. Site speed

      11:49
    • 12. Internal site search

      20:31
    • 13. How to use UTM parameters

      23:06
    • 14. Traffic sources

      21:11
    • 15. Landing page and keyword

      10:40
    • 16. Keywords

      11:30
    • 17. How to evaluate traffic sources

      13:20
    • 18. Goal setup and funnels

      19:02
    • 19. Product analytics and enhanced e-commerce

      18:20
    • 20. Segmentation

      14:07
    • 21. Google analytics access rights

      5:36
    • 22. OUTRO

      0:29
26 students are watching this class

About This Class

This course is designed to understand how the Google Analytics works as a tool. Teaching is based on 50 practical tips from the interface.

Transcripts

1. Introduction: Hi, guys. My name is problematic, and I'm a Web analyst in Public Analytics Lecter. The goal of this course is to teach you how to use Google Analytics as a tool. But what is more important, how to leverage the data to drive your business at first place we won't cover topics is how to create an account or set up a basic measurement, because you can find thousands of tutorials about it. And it's not the place where the money and value is generated. After completing the course, he should be able to identify what should be quickly improved on the Web site from a user experience point of view. Very fire agencies, online media spends you decide, sir, data check and evaluate campaigns. And in general, you should have very decent dated really ability. The course is designed to help you answer three questions. Why, how and when something happened. And it's for everyone who wants to make the right data decisions. We won't cover all of Google Analytics report because there's just too many of them, and plenty of them have very small potential impact. I personally use probably only 25 to 30% of reports. Instead of that, we'll focus on analytics principles, and the reports that can make the huge impact when we will interpret the data correctly during the course will go through 50 practical tips. They are based on my experience from both working in agency and now as an in house Web analytics lead. Some of them come from mistakes I made when I was starting with analytics, so we don't have to repeat them. I'm sure many of tips will be applicable to your data on business, too. The first couple of lessons are slightly more theoretical, explaining how the data is collected and describing the basic terms you need to know. But then we'll spend most of the time in the interface. And one more thing. Always check the lessons, description and resources because most of them will contain an extra material for studying or short please. So what's master? This powerful too 2. Basic data description - dimensions and metrics: Google Analytics is nothing but one huge table I'm sure everyone has seen at least once in your life. G itself is basically a visual layer upon that huge table, and every time we change something in report configuration, it just brought us a different data set just because it's stable. The very basic data differentiation is on two types that mentions and metrics for easier understanding. Them engines are text, data characteristics, and metrics are numeric data characteristics or to simplify to a Tom's. Dimensions are letters and metrics are numbers. So let's have the first look into interface and see the dimension metric relation. Okay, so here we are. For the first time in the interface, you don't have to log in into it because we will explain it in more detail in less than number five. So now were only going to show the relation between them engine and the metric. So I'll go into random report, let's say to acquisition all traffic and source medium, and what we're actually see here right now is the chart with the metric users. But what we are going to focus on is the table below it, where we can see some numbers and some letters. And as we said them engines are letters and metrics are numbers, and this is exactly what we're seeing here. Source. Medium is our dimension, which is Google organic, direct, non. You chew referral. And here are our metrics users, new users Session Mount straight and this is exactly it. So this is the relation between them engine and metric. 3. How the measurement is done : Google Analytics Measurement is based on cookies. And don't worry, not under wants you buy in the store intact. Terminology. Cookies are small file stored in your browser, and I stress the word browser. They have limited size, and they don't store any information about the pages review before or how many times have already visited the website. For purposes of Google analytics, Cookie is just identify of user or to be more precise, browser. So, yes, the number of users we all seen our accounts is, in fact, the number of browsers that visited our website. Let's have a close look on how the G a cookie looks like. So this is how Google analytics kooky looks like. Pretty simple, right? We have a couple of numbers in here separated by that. Let's see more detail what it actually means. So G a 0.1 dot two is every time the same and is the version of Google Analytics measurement coat. It's every time the same, so we don't have to worry about it that much. Then we have to mawr number sequences. The 1st 1 is actually nothing by the random number it's generated by JavaScript measurement coat which you all have on your website. And then the 2nd 1 is the first time stamp. This is actually the time when you for the first time, was it it particulary website. And these two numbers together make something we called client idee, which is user unique identification and as we know, the user is nothing but a browser. So this is actually the client i d. That measurement code takes every time you visit a website and is able to recognize you as the same user who is already being on the website. We already know the DJIA is one huge table to make it work correctly. It's of course, necessary to have the JavaScript snippet on every page of your website. Once it's loaded together with the rest of your websites coat In very simplified version, it only sends information that some users who we know is a browser has you certain page at certain time. Time is represented by timestamp, which is nothing but time format and this is it. There's off course, some more information sand, but for understanding of how it works, it's OK to remember just this right. Imagine a simple scenario. He attacked the your address into a browser, and your brother is now downloading and parsing in HTML. Coat. Part of this girl is also a Google Analytics cult, which ensures measurement and once is loaded. It asks your browser a question. Is there a Google Analytics cookie? If he has, it's day for gland I d from it and knows it's a returning visitor. If not, it's a signal for a G A coat that it has to create a kooky and take a client. I d from it. Either way, we now have a Google Analytics kooky so we can send all the information on Google Analytics servers. This is exactly how Google Analytics Circle looks like we're calling a J server. We're sending their on information that the reviewing bait we want to send it into this particulary Google Analytics account you a hyphen. 123456-1 We're sending their information that this is our client i D, which we took from G A cookie. And we're sending their own information that we just view this particular girl called Contact Page. Once all this information hit Google analytics server, they are parsed. There and start in a database which we know is a huge table author, that we are able to see all the beautiful charts and tables in the interface. And now let's have a look at the very basic metrics definition. We have to start with the users as humans, really people joking with your website, and now we have to make a switch. Two. Who is user from Google Analytics Perspective. They are browsers, yes, browsers. And even if you use the same computer on which you have three browsers from Google Analytics perspective, he will be taken as three users. Every user can make a session during which he or she can view multiple pages. So this is the very simple definition off how we got from users as humans do page views. And now we're going to look how decision is defined, Decision might seem, is something so obvious that it wouldn't even think about its definition. I think it's worth remembering off how exactly the session is defined in Google analytics is the group of user interactions with website and buying direction. I mean mostly page views. Some evens like adding products into cars or a transaction and here we have awareness example of how rescission can look like we have a user viewed four pages he started at nine o'clock, and we have exit at 9 25 Next thing good to know is that there are three possibilities off when a session can expire. The 1st 1 is midnight. So if you are browsing website around the midnight, the first page view after midnight starts a new session. The second case is when you arrive while different campaign, either by clicking on a banner or clicking on a different result from a search engine and the third. And I think the most common case is so called in activity window. This window is by default, set to 30 minutes. Let's have a look on the second example to see what I mean. As you can see, the second session starts very much like the 1st 1 up to the second page you, as we can see, there's exactly 40 minutes window, and as you said, by default, the activity of Indo is said to 30 minutes, which means that with 1/3 page view starts a new session. The activity window can be adjusted in the interface, it can be either shortened just a few seconds or prolonged to couple of hours. There's one more metric. We have to define its conversion rate, and we will use it very often during this course. Sometimes it's almost the Holy Grail for all the marketers. Is the percentage of sessions that completed ago a goal can be purchased, sending the league reading more than five pages on block and so on. In general, though, is any action we define and take its component by which we evaluate our website. A quick example of how it's calculated. Let's assume we have 1000 sessions from which 25 purchase something 25 divided by 1000 is 2.5% the higher the by their try to remember his one, because we'll use it often in the upcoming lessons. 4. Bounce rate, time and uniqeness: There are a couple of metrics used very frequently. When it comes to evaluation, bounce, rate time and something that contains unique in its name, they all have to be perceived in context. Otherwise, it's very easy to misinterpret them and make a wrong decision. The 1st 1 is bound for it, one off the most common metrics in general. I personally love it, and again I have to stress it has to be used in context. It shouldn't be used. Escapee I Let's look how the bombs rate is defined. It's simply the percentage off single page sessions. It means that is the number off sessions during which only one page was wheeled. Regardless how much time you've spent there. The most common usage is when we look at the number with individual page or set off similar pages, like all product details or or category pages. A lot of marketers still consider high bounce rate as wrong thing. The first thing we have to mention that the metric itself doesn't tell us anything at all. Let's have a look at the page with contact information. Let's assume a simple scenario. Your potential customer is looking where you're offline store is located. He found the information on left the website, which is okay. It's natural that this contact information page will have high bounce rate. It's quite obvious that there are pages with bounce rate very close to 100% and it's completely OK. On the other hand, if we have a product, ditto page with, let's, say, 1000 page views in the very high bounce rate, then it's definitely worth to do the further examination. In a better case, your measurement is broken, but they're not affecting your business directly. In the worst case, there's really something wrong with particular product detail page time measurement very commonly used and by many marketers, favorite metric, even though not many people really understand how time measurement in perfect this. It's based on times them. And let's have a look on simple illustration. How s obvious thing this time? It's measured in G A. So time measure Man is based on the time stamps, and timestamp is just a time format. The way Google Analytics works is that it subtracts the following time stems for articular pages. We're going to look on a three examples and see what time will we calculated in Google Analytics? So we have example. Number one way. We have a session consisting of four page views. We started a wizard at nine o'clock and we exited website 25 past nine. Now we're going to show how Google analytics will subtract Time stems for particulate her pages. So we start with the baby one. We have to subtract five past nine minus nine o'clock. It's five minutes, and this is the time measured for Page one for Page. Do we subtract 10 past nine minus five past nine, which is five minutes, and this is the time measured for page do for Page 3 20 past nine, minus 10 past nine. It's 10 minutes, and this is the time recorded for Page three, and then we spend five more minutes on page for. But because there is no further timestamp, the time measured for page for will be exactly zero minutes zero seconds. So to sum it up, we have the first time stem, which is at nine o'clock in the last time stem 20 past nine, which gives us session that lost it 20 minutes. Example. Number two. Let's assume a scenario user started a session at nine o'clock. Then he viewed the second page at five past 9/3 page at quarter to 10 and then he left at 5 to 10. Now, as we know how in activity window works, we see that between Page two and Page three, there's 40 minutes window, which we know is the moment with musician starts. So we actually have here two sessions. The 1st 1 started at nine o'clock and ended at five past nine, which gives us five minutes session duration, and then then is 1/3 session that started at quarter to 10 and ended at 5 to 10. We only hear have a one time stamp at quarter to 10 which gives us the session that lost it exactly zero minutes and zero seconds. The simple rule worth remembering is that we don't measure the last page. And now, as you know how the bounce rate is defined, just check for how many sessions your average session duration is exactly zero minutes zero seconds. Plus, in the non balance sessions as mentioned, the last page is never measured. Yes, it's probably an affair, but this is how the works. So I guess from now you won't rely as much on this metric anymore. There are much better ways to trade content than time, Let's say, even with information about what percentage have user scrolled. This requires, of course, slightly advanced measurement implementation, and you'll find a tutorial for it. In the description of this chapter, there are a couple of metrics n g a that have worked unique in their names, for example, page views and unique pages unique. And, as in J means that something occurred once per session. It's not per user or person. Think else it's birth session, so if you view one page three times during one session, it will be counted into page views metric three times and into unique page views just once . 5. How to use the tool: that was a theory. It's time to finally get to the interface and see what everything is in there. So I'm not going to prolong it on. Let go straight into it. So here we are, guys. In other theory, we're going to spend most of the time in the tool. We're going to explore the Google merchandise store data. If you want, you'll find link in the description so you can add into it and see exactly the same date I will show you, or it's completely okay to look on your later. So, uh, want to log in? You should see something similar I have. Right now. I'm in the home tap, which is kind of a dashboard. Quite new in Google Analytics on you can see some basic data from your account s What's your audience overviews from real time data. How we acquire users and so on and so on. Thistle is just the extraction off the data we have in all the report, So yeah, feel free to use it as kind of Ah, quick, look off how you're going. But but do not spend a lot of time in here because you can do any deep analytics in here. So let's see what we have next in here. We have a main taps. It's 2456 of them on every of them is set of Ah, quite nice amount off. Report the 1st 1 It's scold riel time, so if we click on it, we see another 246 steps and the every main tap. Let's call it mean tap as the first step called Overview, which is again something like a short. There's board off what is below it now, where we are in the real time report, which, as it's quite always from the name, is showing us what it's a real time situation on the website. We can see what is ever be by split, which now we can see is 2/3 off mobile users. 1/3 of desktop weaken. See off how many page views are here per minute or per second. Andi. Then we have some top active pages of referral stop keywords or or publications, so it's not much awful analytics you can do here where it makes sense to use this report is , especially if you let's say, redesigned website and you want to check if your measurement works fine or you started some new feature on the website to see if you measure it. This is the place to look for. But I said, not much often analytics. You can do here, so I use it. I would say, I know twice a year at maximum. Let's jump into the audience and here we're going to show off. What controls do we have here again? As in previous case, the first step is called Overview. Let's click on it and see what we have. I will believe the comparison time because I have it. And now So this is something you should see. Once you click on the overview, this is how most of the reports and Google analytics look like. So we have some a line chart in here. Andi. Then some more metrics on some very basic information, exactly as in previous Case three. Overview. Tap is the dashboard off? What Google fund below it. So let's have a look on the controls. Definitely. The first thing you want to know how to set up or change is a time range. You can do it on the tub, right place on the top right corner. Sorry by clicking on the date, selecting they train, she want to see clicking on Apply waiting for a couple of seconds, and you should see the data for selected time. Rich A Z you can see now I'm looking on the user's metric, which shows me by day that I had through three thousands users than 3.62 point seven and Salon, which is the daily granularity. So this is so far, very simple and easy. But what we're looking at is just the one date range, which pretty much doesn't tell us anything, right? We know that since eighth off May until 19th of May, we had 30,000 users who made 38,000 sessions. We don't know if it's a lot if it's not a lot if it's OK if we were supposed mawr or less or will they had. So this is exactly the place when we want to know how to compare two dates right on, and how we do is exactly as in the first case, we have to click on the top right corner and click on the compared to check box by default . Google analytics is trying to compare Thea original date range with the previous period and that can be very mistaken. Let me show you why. Uh, I'll leave it as it iss eso. I expect it to show me the previous date range. I click on apply. And this is something especially beginners do a lot. And let me explain you why It's not a good approach. A zoo. We can see both off the lines look pretty much the same, but one of them is slightly moved to the right or to the left Depends. Um, what you prefer this'd is exactly the problem because we're not comparing, like for lax here. Ah, the rule of thumb is that we always should start with the same day in comparing. So the first day off original period isn't Tuesday and the last one is on Sunday. This is exactly what we should have in a comparing date range. So we don't want to compare the the previous period by the custom one. And now we click into the first previous time range Ah, field and we start with the Tuesday as in the first case and we end on Sunday, which is here and let's see off. How would you look like in a graph? Makes much more sense, right? Or at least I hope so. So, uh, especially the worst case, which happens quite often that people lose something like this. They'll take the data from, Let's say, Monday Q. Two Thursday and compared with the previous period, they click on apply. And what it does is basically that we're comparing the I would say strong part of the week from Monday to Thursday, with with the Thursday to Sunday, which are much, much out in most of the case says this is the days during the week where you don't have that much traffic. So, uh, just to remember, compare likes for likes, I hope so far it makes sense. So I will unclipped the check box to have just the one line to show another controls we have here. I will click on apply, and now what is possible here in the overview tab, which is Ah, that board. We have a couple of metrics in here. Users new user sessions, number of sessions per user. It's possible to plot any metric you want to just by clicking this small grade drop down menu. We have a users by default so we can select every session, duration, bounce rate, new users and any of these metrics. So let's say we click on the sessions and we see the sessions. It's also possible to click on any of those small widgets. Teoh plot the metric on the main graph. So if we play, let's say on new users. You can see that now we're looking on new users. Eso this is it. One more thing to show you is something we called granularity, right? It's just in here on granularity is something is basically how many dots you're going to see here. So it always depends on how long they trained you select it. Eso Now we're we have ah, 17 days, so it completely makes sense to have the daily data. If you have a shorter period, let's say two days anyone to see the intraday granularity than it, of course, makes sense to look on the hourly data. Let me show you what it does. So we're changing the granularity to hourly data so you can see that we have many more, uh, time spots in here So it's not that easy to orient in here to to see off what actually is going on. So if we're going on the day, then it makes sense. Of course, if you would select weekly granularity, we would have just three dots, which also doesn't tell us much. So just use a common sense in here and don't use an hourly granularity when you select three months data, on the other hand, then use weekly data when you have just 10 days. So common sayings just works here. So this was the first look into the interface. And now we're going to see how to use the report. It also contains a table in here. Okay, guys. And now let's move into a different tap in which will show also a table. So from the audience main tap and then overview, we're going to demographics and again overview, as in previous case, Also overview. Here, ISS a dashboard off war will fund below which in this case means in age and gender taps. So, uh, in this overview, we see what is our age and gender distribution. In case of Google, its most of the users are between 25 34 and the almost 70% of users are males. It's possible that you don't see these data in your account. So let me show you what you have to enable in your admin section in order to see it. So on admin section is always on the bought left corner. So click on this small wheel. If you want to enable it, you have to click on tracking info, then data collection and scroll down to advertising reporting features. And this is the bottom that has to be turned on to see the data. So if you don't have it, please turn it on. And in couple of days, I think it's something between 24 48 hours. You should see those demographics daytime your account. So this is the place to to enable it. Um, let's jump bank to to the data. So we were in the audience demographics and overview s so far, it looks pretty much the same as the main overview top in audience and what we're going to see right now. ISS the age tap. So click on it. And now it it looks slightly different, right? Because now we don't see only the line chart, which is nice. But we also see a table here. If you remember from the second lesson, the very basic data differentiation is on them engines and metrics. And this is the first time we see it in the interface. So the dimension here iss eight and the metrics are users new user sessions, bounce rate, Bages procession, every session, duration, transactions, revenue and conversion rate. So the controls are exactly the same as in previous case, but we have a couple a couple more opportunities for what we can do. So let me scroll down to the table because this is actually where analytics is being done. So the first thing I will do, I will just minimize this this last step, Teoh. Yes, this is it. So, through this report is pretty simple because it only has six lines. Don't worry. We'll have plenty of them that will have couple of thousands of lines. So the feature I'm going to show you now we're be very helpful. And it's just the filtering. So for filtering the table, you have to use this small field if you just collect into it and you type something, it's pretty straightforward to use. So, um, let me, for example, type 25 included, and you can either press enter or click on this button on. If you wait for a second or two, it'll filter you only the lines that contain the tax you've typed into the filter field. So, yeah, you can filter just line you want to see. And also the line chart is now pre filter to Onley the data. Ah, you see in the table. So this is pretty useful. One more thing I want to show you here is that it's possible to see the two metrics in the same line chart. Now we have the users, and if you want to add one more, we click here, and we can either use this drop down clicking or we can type something into it. So let me type, let's say sessions came selected sessions. We wait for a second year, and now we see it so we can see the lines are copying each other on, and this is something that actually makes sense. But don't or you'll find ah, a lot off metric. That won't look that nice. Ah, so far, we're still just showing the interface, the controls off it. So bear with me for a couple more minutes and then we're actually start with 50 practical tips. So don't worry. Now we're not showing any table, just showing the interface itself. Eso We now know how to plot to metrics. Let me disable the filter off 25 on. Now we have again along the lines that are available by default in this report. Ah, if you want to, let's say, order the data differently by some different metric. It's very easy, and the only thing you have to do is just to click on it. So now we can see this report IHS ordered by users. Ah, and it's ascending. So if we won't do, let's say, sorted by decisions, we click on the sessions, and this small arrow is always telling us by which metric ISS all report sorted if you want to. If we, for example, click again on the sessions, look for what happens now. It's descending so you can play with it. Explore the data you have been. Worry a size promised will go to the practical tips in just couple of minutes. So, uh, this is it, Ondas, We said this is our dimension and these are our metrics. Let me show you one more thing in this lesson, we're going to open. Ah, the friend report, which is gender. We are still in the same section, were in the audience. We're in demographics and now we're engender. Ah, this is probably the easiest report we have in Google analytics because it only has two values in his damp engine. So males and females in the previous one, we were able to see our what's our age distribution. And now we can see what's our gender distribution, right? It doesn't make much sense, Teoh do some filtering in here because we don't have two lines. Right? But what I want to show you is that it's very nice to see awards, our age distribution and also nice to see what's the gender of distribution. And what is even more cool, I would say, is to see both this information on one place. And it's very easy to do in Google analytics. And so far we will only use e using the naming the dimension and metrics. And from now we will use primary and secondary damn engine. This is exactly what this is exactly the feature we're going to use to break down the gender by age. So we have to click on secondary damn engine. And now we can either try to find it somewhere from the multiple damn engines that there are here. I think it's more than 100 or we can just simply top here and it'll find out h so each Here we are. We click on age and see what's going to happen. Okay? And here it is. Ah, we had only two lines. And now we have, as we can see in the bottom, right corner. We have 12 lines and this is exactly what we're looking for, right? The breakdown off our gender by H so a zoo we can see our top users Groups are males between 25 34. The 2nd 1 males 35 44 on this is it. So there is also possible to do the filtering. But please remember every time you type something into this white field, you Onley feel filtering primary them engine. So if I will drive here female, This is what's going to happen, Right? But if I want to filter the age, I can't type it here. What I have to do is let me is able to filter. We have to click on advanced filtering, which is here and here. We can actually select which them engine we want to filter. So let's assume we want Teoh filter the age that contains 25 and then we have to click on them engine select, the one we want to filter, which is gender in our case. Uh, yeah, this can be slightly tricky, right? Because if you want to filter only males in here and you tap mail also, female contains male, right? So it's not that easy to feel there. So this is ah, drop down menu we can use for slightly and I'm stressed very slightly advanced filtering. So if we want to fill their only males we can use begins with and time here Well, it can see you. You can also select it from here. So we tied year, we click apply and we only should see the one line and this is exactly what happened. So again, it's possible to sorted by any metric you have here. You can plot on the top any metric. You want to or not. Just one, but two. And ah, used the granularity that fits to your ah, time range. So this was it. Now you should be able to use the tool. And we're going Teoh, the next lesson where we will start with the practical tips. 6. What to remember when doing analytics: first things first Analytics is nothing but way of thinking. Every analysis is only as good as their hypothesis or question. You asked. Before you dive into data, the very good characteristics off analyst is being sceptic, which means they do very fine numbers from multiple point of views. Secondly, you have to perceive the data in context in under the second lessons, with the explanation of high bounce rate and contact or information page. That's exactly what I mean. Context is something I will stress couple more times your next lessons. Another one is that we don't measure the reality because we measure the browsers and we can't measure everything to G A. By this, I mean we can't measure P II Information such as name, surname, email, phone and the national identification number. This happens will receive a warning message from Google with some period of time to fix it . 7. Socio-demo data: As you noticed. There s plenty of socio demographic data in your tool, and you might be curious how Google loss it. It's taken from your browser and activity, specifically DoubleClick cookie. It means that based on your search input, history, clicks and banners and salon, Google's algorithm profiles you. If you want to know what actually Google thinks a few, you'll find the link in the description, and they'll be surprised that it can be slightly on precise. So, guys, now we're going to show up what it actually means to do. The analytical thinking cried. It's quite the beginner's mistake that a lot of people just see the one number and then they're going to act based on it. Which is a huge mistake, I would say, and let me show you a very simple example, which is, which is, unfortunately still very, very common. Let's assume you are trying to find out who is your targeting group and you'll open and demographics report and go to overview and you'll see here. Okay, The most of the users are between 25 34 they are male, so this is probably my target group on which I should run my campaign and it could be either I know some banner campaign and or you want to pay more for these guys in search engines or whatever online campaign it ISS. Uh, if you will look on Lee on that data, that would be a huge mistake. And let me show you why. Because this is exactly where the blood, the place where we're thinking in contacts is very, very handy. So let's maybe click ah to the age report and scroll down of it. A So we can see. Yes, the age group between 25 34 is the top one, right? 43% off users. So you might think, Yeah, this is These are the users I want to target. But if you will look on how much they actually convert, which means completing a goal in this case, it's a transaction. You can see that the average convergent rate is 3.22% and the top 1 25 to 34 is three born 29 which is about the average, which might seem nice. But if you will just look on the second line and the four Flynt. You can see that especially in the second case, it's 4.92 which is a huge difference, right between 3.29 4.92. And even the target group on the for the age group on the fourth line, which is 45 to 54 has significantly higher conversion rate, which is something we actually want if our goal is to to let se increase the sales or any any conversion we have. So this is This is, I would say, very, very simple example of how a lot of people are being deceived by the by the data because they don't even go one step further. So it was very simple example of how to get from the our top ah age users group to the ones that are are probably, ah, the ones it should and actually generate the money to us. Right? Eso we might change an opinion that maybe 25 to 34 isn't our top targeting group. It might be 35 plus and what's actually even better to do not just going by one metric and always use a bit more to maybe combine it also with not just a different metric, but also a different time engine. So if we use the same one as in previous case when you were just showing the controls in the interface and time here gender Ah, and we're going to show all of the lines which were going to do here scroll a bit to the right way, we can see that our absolutely stop targeting group is the line do and then here, this one. So they are the males between 35 4445 54. So our top targeting groups, when it comes to generating the revenue and transactions are males 35 plus definitely not males 25 34 which is 30% of our users. But their conversion rate is actually how below the average or, let's say average because it's pretty much the same. 3.2% on lets me check. What about females? The top one is lying for, which is 3.5%. So even even the females 25 to 34 are better targeting group than males 25 34 from conversion rate in port off you so That was a very simple lesson to show you what it means to think analytically not to see just the one number and one dimension, but actually to see it for multiple point of views because, as you could see, it was very easy to misinterpret. Interpret the data and not to do the right decision. So, guys, we're going to stay still in the audience report, and this time we will focus mainly on socio demo data. Ah, this is actually the lesson in which will start with practical tips and leads. Jump to the 1st 1 We're going to geography reports and the language one. Ah. As you can see, we have abbreviations as values of the language. Them engine E and U S E N g B E s. E s friend Germany, Portuguese and so on and so on. How to use it. Ah, the values in this dimension comes from your user's browser language. So if I, for example, live in Czech Republic and I have my browser set in Czech language, this is exactly where the value comes from. So I will be here a Z user. We see that CS I assume this is the abbreviation of Czech Republic. So now how to use it? Ah, you might look on this quite sceptic because Okay, What am I going to do if I know that my user speak a different language and that's exactly the point. The we're going to find out who exactly are users are to do to make easy for them to to browse your website and I'll give you a use case. So this is actually a tip number one. I had a client who runs his business in real estate development, and they were starting a new product. So we were trying to figure out who should be or could be the our potential customer. And this is this was actually one of the last reports. We're looking at two to find some a process information and you actually found it. We found out that about 10% off users ah had very in language them engine Russian one. So it's quite the huge group here in Czech Republic. But he was just it was just 10% of users. But as we as we saw in report, their conversion rate was much, much higher than the average one almost three times if I remember correctly. So what we actually did it was that we translated the whole website into Russian language. And then the miracle happened just two or three weeks after that. The client called as that it worked perfectly. And he has. He already now has sold out about 2/3 off the whole project, so there was very easily done and we didn't need to do some some like deep dive analytics in it. So please check it. And if you find out that there's some significant ah, not native language or the language in which you have your website check, if there's some significant number of users who not speak your language and try to find out , maybe they really do convert more than you would expect. So this waas this waas tip number one Ah, let's let jump into the second report, which is location. What is worth to remember or know here. Iss Ah, how Actually Google Analytics is able to recognize the location. It's based on your Internet service provider I P address. So on a country and the state's lever should work perfectly, But it sometimes may happen that your provider is sitting somewhere, I don't know. Three or 500 kilometers away. So, Butte, on the city level, for example, you might seem like you're from somewhere else, not a to place you're actually sitting at right now. So anyway, tried to find out where do your users come from? Ah, it's possible to drill down much deeper, not just on a country level. So let's see of how it works. Ah, we're click on the United States. Ah, now we see a regions which are states and we can go even a level, or maybe to deeper. We click on ah on ah region. Now we see the city, and now how to work with it. Try to check of how many of your customers or users in general or what is better. I will take customers as we can see that the conversion rate actually differs a lot if we looking on a different different cities. I guess number six year, 12.46% which is a huge conversion rate. And it's a Palo Alto. But what is more important trying to check and this is step number two. Where do your customer slip you d live mawr in a small or big cities who they are again, it really helps to understand who your customers are, because this is this is exactly the main point of this section. You won't understand who your customers are on again. Dip number three. It's very helpful when you're deciding off where to run your online, but mostly you're offline campaigns. If you do some billboards or another another out of home campaigns, this is exactly the place where to look for the data. Eso if it actually makes sense to run some out of home campaign in I know ah country, which is Oh, our city, which is no a couple of 100 kilometers from your hometown. This is exactly the reporter. They will help you to up to, to know if it makes sense to do or to run such a campaign. So please check it. Ah, drill it down and understand we're customer is Ah, let's say we have another tip here how we can use the exactly same data and this is ah probably works much better in a smaller countries, I would say in Euro, but also us can be a case if you're thinking about if you have some pickup points, your offline so user can buy something and you have a pickup one where you can come on and actually pick up goods. So if you want to know where they live or were they buy mostly. This is also again the place where we can have the data. So please check this data. It really makes sense to spend a couple of minutes in here to see what you already have. Teoh, help you make a better decision. 8. Browser data: again something that might seem technical. On the other hand, it's very useful. The first case, when it's handy, is when creating personas or figuring out your business archetype s the brother type can help you identifying where visitors are. Let's do a quick use case off what tone of voice you should use on your website. Okay, so we're stilling in the audience technology and grows A and O s step. Ah, and what we're going actually to do is a very simple think. So Ah, what May men, if I navigation on what we're going to do here is to just compare the conversion rates between the different browsers. Ah, and this report, this is the metric on the very right. It's here. On as we can see, very nice example are the first lines chrome and safari? A. So we can see. There's a huge difference between their conversion rate, right, 3.85 0 born 75. And the case we're going to show is quite simple. I hope guys, you know that every browsers renders your website slightly differently is because they just have a ah different graphic core which can cause small differences off how your website looks like. And sometimes those small changes can cost you a lot of money. And let me show you one example, right? I wouldn't personally expect that much of a difference between those two conversion rates, right? It's, like 56 times more between the chrome and safari, right? It's quite always Is there something wrong? And let me show you Why is it probably so, uh, I'm now? Uh, right now I'm using the chrome. So let me google a merchandise store. So Britain I store I'm clicking on the link. I'm selecting Europe. And now I'm seeing a website, actually. Right, So I have a navigation here. There is a carousel with different products. Every couple off second there, Seenu on so I can interact with it, right? This is I guess how the baby is supposed to look. And now let's do exactly the same thing in safari. So I I'm going to change browser. Here I am. Zoom it and I will google the same thing Ruble merchandise store. So you can see that I'm actually doing exactly the same thing at the results from from Google. I'm clicking on the 1st 1 s so far, so good. I'm selecting Europe. Um okay. Nothing is happening. Oh, yeah, it is. And this is actually what I have right now, right? The Onley element on the website. I can interact with his navigation. There is no carousel, which is quite strange. Freddie can see, and I'm as a user now, it could be quite confused. Is it? Okay, Is the way the page is supposed to look? I don't think so. So let me scroll now a bit. I'm scrolling, scrolling, scrolling, scrolling And here this is actually the first place I can see the Carol. So as you can see, it's probably something is probably broken because I don't think that Google guys want this page to to look like this right on. I now hope you've got a point, right? Ah, the pages just doesn't display or render correctly in a safari, which Ah, let me go back to two here to analytics and to the data. And I'm not saying that this is the only reason why there's so much difference between conversion rights. But I hope you've got a point, right? There's something broken and please do exactly the same thing with with your website just opened this report and check the very right a metric if you don't find something something very similar. Also, a good thing to do is to click on a browser. On what happens is that you will see browser versions eso the same browser, but the difference worsen off it and and do the same thing. A. So we can see. The average conversion rate is 0.75 the 2 4/5 line is one born 32 which is about average almost twice right on. This is probably the newest version off a safari which probably this place to upset correctly. But yeah, it can be that they made some ablation browser and now it looks fine. But this is something definitely, or developers should check before they roll out website because a so you can see it can cost you a lot of money, right? The five time last conversion read comparing crow, and so far it's a lot, so please do the same thing. Now let's go back to all browsers and I'm going to show you one more thing. Please do the same thing with screen resolution we have to click in this vertical navigation to screen resolution on this is the data will have, right? I hope you're familiar with this values in in, ah, screen resolution, dimension and and again do the same thing. Look on the conversion rates off resolutions, right? You can see it's it's again a lot of differences between them. If I just compare the 1st 2 lines. I'm sorry. There's like 10 times more conversions from the 2nd 1 even though the 1st 1 is the top one when it comes to to users, right, So there's probably something not displaying correctly in this resolution, but on the 2nd 1 works just fine. So please do the same thing. This is really a low hanging fruit. Sometimes it just fixing couple of lines of code and and then you will see a miracles happening on your website. Also, a good metric. To check here, either with browsers or with resolutions, is a bounce right, because if there's something really, really broken, you should see significantly higher bounce for it. And as we can see, it also differs here. It's just from the top 10 lines. It's sometimes between 26 and 58 almost 60% right, which is like twice amount of users leaving the website after winning just one page. So this also can be a hint for you that there's probably something broken with it. Also, a good thing to do here is to use a browser a second, the rhythm engine. And why is it so? Because you can compare whether this resolution is broken in general. So the low conversion rates will be for all the browsers or just one specific browser. So just time here browser. Uh, and you have the data with which you can work right now. Ah, so this was guys. Step number six. We will show maybe how to extend this information in the next lesson when we will show how to use a device category dimension, which is also also extremely important. So please, guys, check this one. This is really low hanging fruit 9. Mobile experience: one of the most important topics in this course. First off, all check your mobile traffic. Share an answer yourself. Question. How much attention do you pay to it? And what is even better to check is more about your picture progress over time. It wouldn't be a surprise if it's almost 50 50 comparing to desktop, and if it's still just 30% it deserves a lot of attention. Don't be scared if you will see lower conversion rate there. It's quite normal anyway. It shouldn't be three or four times last, then asked up. If so, you have probably the first place to start work on for this year. It's in a secret that's still a lot of people do. The research on mobile and purchasing desktop Very simple rule works here. If you suck on mobile, there's probably no purchase on next up, either. Again, good metric to check. Here is bound freight. If it's enormously, hire a mobile, you're probably doing something wrong. There can be many reasons for that slow mobile page. We will cover this topic in next lesson. Broken mobile experience in general, missing important elements when creating responsive urban and so on and so on. So let's dive into data. Okay, so now we're going to move from technology tap to more while. Ah, in this report were actually going also to work with the overview tab, which we mentioned a couple of times. It's just the dashboard off what we find below it. But this one is you can see has only two taps. So, um, again minimize the navigation. And now I really want you to be focused right now because this guys is really important. First of all, ask yourself a simple question. How much attention to pay to your mobile experience? Everyone loves smooth mobile experience, but not everyone has it so on. This is exactly what we're going to to explore in this report. So this is Ah, dip number seven. Ah, check. What is your move? I'll share a So we can see in a Google's case for the past two months. April in May. It's about 28 I would also count tablet as a mobile experience, which is about 30 to 33 32 year percent off mobile users that come here and what is even better to do is to export Ah, long term data and see, what is the progress off it to see if it it's kind of Ah, very slowly increasing. Or if it's if it's massive, you're in black like, let's say, two years And hey, we're going to show one more technique on this is exporting the data. Ah, the first thing we have to do is Teoh applause. Some sometime them engine here. What I think makes sense. If we're going to look on a long term named Data and I would say long term, let's say two years. So let me at first select the two years period. So it's a uh yeah, this is a nice thing to show you. You don't have to every time click on the left arrow to get to like, let's say, 2016. Sometimes it's much easier to just click into the field and time there. Ah, number. So this is what I'm going to do right now. I'll click. Apply on what happened is that I don't see any data, and I don't know why is it so um uh, So let's do it again. Click. Apply again? I think so. I think something is broken in the interface, so I'll just do the clicking. But anyway, it should work normally. And as you can see, I have it selected. So I didn't know why doesn't work. So let me click on the May click Apply now. Still no date s. So this is really weird. So yeah, let's maybe the boss do past three months. Okay, Now I have reader. So let's do the same thing again. Let's say since April not to overload the interface. Okay, Now we have it anyway. Sort of some back. Uh, okay. And what we're going to do right now is to use the second the rhythm engine, and I recommend here to use something with month A Let's pride month here, and we have a couple of opportunities in here is this is a day of month. We're not going to use this one. Ah, and we definitely want to select something from month index, which is in extra month month of the year, which is just a number between one and 12 or months off a year, which is exactly the foreman we're looking for because we have in the same time, engine both year and month eso We're going to select this one. Ah, and it'll give us the data for the whole period. Right? Since April 2017 until the end of May 2018. Ah, so we now have to a lot all the data which we're going to do by this drop down menu. And as we can see, we have 42 lines, so 50 will be enough. We'll wait for it. And now if we want to export it, we have to scroll up to the very top. And this is our our button export Click exported it. We can sell like the former, which we like. You can do it directly to Google Sheets to excel or CSE file really depends on where will you work with it, then? If it's excel or if it's double Oh, are some different tool really depends on you on. Then do some some small game with it plotted in a time to see off what percentage it was one or two years ago. And what is it now? So you can somehow extrapolated into future to see off. When will you, for example, have the 50 50 share between desktop and mobile? It's quite possible that he already headed right now. So then, in that case, you should be even more focused on it. Okay, So I'm now going to cancel this this month of year. Them engine on may be one more thing to to show you, Uh, if you if you're not sure what exactly some metric means. Feel free to use that small Aiken with a question mark. And if you only who were on it with your with your pointer, it will show you what exactly does it mean? It the It's very nice documentation. Eso Ah, a lot of times there's learn more button, so feel free to use it if you're not sure what some metrics mean. And I also use it quite often. If I'm not sure by off what, exactly? Some metrics or them engineer. So ah, don't hesitate to use it anyway. Now let's go back to how to work with this report again. We're going at first to look on conversion rate between the devices a Z, we can see the average for the past year in year. In one month it's 2.60 and again we can see that there is difference between desktop and mobile, and it's not small difference. Actually, the conversion rate between desktop and mobile it's 10 times less, and I repeat 10 times. That's really, really a lot I would expect if you're if you're mobile, experience doesn't suck. And it's quite nice. I would say that you should have at least half off the desktop conversion rate also on the mobile, which isn't the case at all. So, uh, please do exactly the same exercise. And if you will find that you you really are 456 times less than on desktop, this is probably the first places you should start work on, because just check what is your percentage off mobile searches in your country, For example, here in Czech Republic, it's over. 50% of the searches are done on a mobile, so that's a lot really a lot on may be the first thing to do here. It Z let me just sort in the date range so we don't have to wait for a long time for later . The first thing to do, which is also ah, the tip number eight. Ah, let's say we'll filter us now. We know that the mobile doesn't work well. Also, during past two months, it's still like 67 times less, six times less than then on desktop four point 4.0, and comparing to 0.71 we're going to filter only more, while see whether it's a global problem. Or maybe there are a couple off quick wins we can do. And what we're going to do right now is to use the secondary damn engine screen resolution , which, with explored in a previous lesson, if you remember with the browsers. So let me do this Onda again. We should see the values which with which we are now familiar on. Let's see off how how different the conversion rate ISS the other is 0.71 we can see that there are differences, right? The first resolution, which is 3 62 6 40 0.36 the fourth line, which is for 12 to 7 32 It is actually five times better than this one. Eso please do exactly the same thing to see if it's if it's a global problem on all school resolutions or a T least some of them work better. Ah, and this is really, really important, guys. And now we're going Teoh Tip number one. Sorry, Tip number nine where we're going to simulate the mobile experience and please do the same thing. It's actually easier than you might think. What we're going to do, that we will leave here. The data we have with the screen resolution and we're going to see off what actually is the user experience in a more about So a good place to start here is to have a Google chrome as a browser and using the consul. Don't worry, if you're not familiar of how to use that you were going to show you right now. So the first thing I'm going to do here is again, as in previous case with the browsers, I'm just going to Google. The Google merchandise starts a real merchant. I star Ah, Blick, and you can do it also in a previous case. But let's do it from here. We have to do the right click and click on inspect in English. And if we do that, we should see something like this. Which is ah, developers Consul with lot off codes and and possibilities to check. Go have your network works and sources and so on and so on. The only thing we have to do right now is to click on this small Aiken, right? It's ah Tuggle device to toolbar. And if we click on it right, there has to be a little sorry on. We'll go back to the original window. It simulates us of how website looks on a different resolution. So you can either select specific mobile device which you would like to inspect or just click on resolutions that are pretty fun here. And they simulate like small, medium or large mobile phone eso. This is guys, what we are going to to inspect right now as we can see the first line, which is 3 62 6 40 This is our resolution that we're going to expect because, as we can see, it really sucks. So it's not the resolution itself, but about the merchandise store. So, Google guys, if you can hear me, please fix it. We're going to lobby for that. So, um, let's go to a different that where we have ah website and we're going to click on international on, and we're going to get as close as possible to 3 62 6 40 Resolution which and, um console ISS 3 20 pixels. This is the closest one, and this is actually what we have here. So, um, I'm not a u x guy, so please don't crucify me if I say something wrong. I'm just trying to do a user testing. So the first thing I see is that there are two hamburger navigations. So this is actually the first place where I as a user, I feel quite confused. I don't know why they're two of them. OK, so the 1st 1 is about my account on the 2nd 1 is actually a navigation. OK, so the first thing I don't think it's supposed to work. The way I see it, it's the Caruso with the products. Ah, it's clear that one of the dots is over the text, which I don't think it's supposed to be, the way I see it. And if I tried to scroll down, I can see also that a shop now button has a dark background which definitely isn't supposed to look like this. So So, uh, I can imagine that not all the people will notice that there is a bottom so they won't click on it. Um, and let's go further. Maybe let's try to use a navigation. So I try to click on view all and see what was actually going to happen. Uh, yeah. Now I should see some content, which is not happening because I think we're waiting for quite long on the morning, but we will cover ah, site speed in the next lesson. So So let's now continue and our user testing. So Okay, I see some products in here Prague list, and I want to filter them. So, uh, clicking on a button. And now I have possibility to filter it by price, which I'm going to do. Um okay. And now, as I'm moving with this slider, nothing is happening. Actually, on the left side. So I would expect that I would see what is my my minimum price, which I can see right now. So again, Okay, let's use it and see what's going to happen. So far, nothing. I can see that I can scroll in the upside, but there is still light overlay over the product lists which doesn't allow me to click on any of the products. So apparently filtering doesn't work properly. I would say so. At least we will observe a couple off touchpoints that that wouldn't satisfy me. A Z user so far. So maybe let's do still. Still nothing happened. Ah, about the filtering. So maybe let's dio the second case and look for a second to the data and we can see that actually, the the fourth resolution, which is four plenty to 7 32 works much better. And to be precise, comparing to the 1st 1 is five times better. So let's maybe try to see off how beach will look like when we get close to this resolution for 22 7 32 So, uh, maybe try to reload the page because still nothing happened after filtering eso. So maybe I will open a different tap with the same content merchandise store. A clicking on the result I will simulate more. Well, devise, which is a blue. Okay, so we have it on and let's see a page so clicking again on international. Now let's change the resolution. So we we want to get somewhere close to 4 12 just through it one day. 3 75 and 3 25 I think this is exactly what we are looking for. So let's use this one, okay? And now you can see it off. What happened? Actually, finally, the corrosive is looking as it's supposed to look, I think because we have the buttons in one line, nothing is overlaying anything else. Let's scroll down to see what happened. Okay? And here we are with with shop now bottom. Now it looks okay, or I think a designers want to look it. So So, yeah, we can actually now quicken shop. Now it's much more usable than it waas in a 3 20 resolution, which just which is just confirming that there is a difference between those two resolutions, both for mobile maybe. Let's try to use also product category. And the filtration is to see it might work here. But that's OK. Oh, as you can see, it visits march. Foster on now. Actually, the filtering so again filter. Uh, okay. Anyway, also not even not even here. If we use the price slider, the minimum price, it is changing right now. But it's Xas some dealer in it. Anyway, It's still not working, I would say as user. So after I use, it's let's see if something is actually going to happen. Still quite slow. I would say, Um, again, it allows me to scroll, but nothing is is happening. Um, all right. So, uh, this was a getting confirmation that Okay, finally, we have it. Ah, at least it worked after quite a long time, I would say 32nd maybe even more. But we can see that that its significance difference right between ah of these two resolutions. And I hope, guys, do you have a point? Right? Just check the data and do the same exercise for you because this is really a low hanging fruit. You can do that in literally a couple of minutes and see what will works quite well on. DWhite really sucks. And where you're losing money. So please do the same exercise in your case. So there's one more thing we're going to show. Ah, we're still in the audience report in a mobile type, and now we're going to collect two devices. Step. It's going to enhance our view on a mobile experience. We've been showing in a previous tip. So I'm just now going to minimize the navigation to see all the data and let's see what we have first thing to mention. All of the data we have here are only for mobile devices, which you can see right in here that it's just 43,000 users, which is approximately 32% of all users. So I'm sure you're familiar with the values in ah mobile device in the fall. Did I mention which are the names off their devices? Actually, so maybe the first thing worth checking is not using the exact names, but to switch to the different vertical tap called operating system on. It's quite similar, as in lesson number eight about browsers. Also here is definitely worth checking the operating system, which is ah, basically showing you off. How much android versus I always users you have, and the first usage is very simple. Andi similar as in as in browsers cases, which is kind of a status for a guy who have such a device, whether we like it or not. If someone is buying some some device, and in this case we just use Android and IOS. It is specifications of user estate. So please look of how much Androids birds is Iowa's users you have. And maybe if it's significant in either and greater I West, right, maybe change your call be or the way you design your website. For such a users, the the way how it works can be different in different countries. So depending on where you live or where businesses, it can mean a different thing. So please second, also do have a look on the conversion rate between those operating systems for again have to stress it just for mobile devices. There's not desktop and would, ah, also make sense is to use a secondary damn engine here. Also, foreign screen resolution, which were in ah, a gang case off brothers or to do break down by age is so you can see whether maybe android are older people, and I was younger or wise versa or something in between it. So ah, that was the first thing. And then I'm going to switch back to mobile device info to see the exact names if you want to see what exactly devices people use, this is exactly the report where to check it again can be very helpful if you want to find out who your users are in terms off their status or how they behave, what they think, so that this is the one case. And then maybe what I think is slightly more interesting, switching to another vertical tap called Mobile device Brendan. And if we switch there, we can actually see the brands now the operating systems while the brand, which is again very, very helpful, I think on and let me show you one example where if you find out this, for example, let's say I don't know Sam soon as we can see has much lower convergent right? Just 0.43 the ever jeez open town to do the definitely makes sense to filter probably the same single made. So let me do that. And then he uses Sorry, I kind of wrong fielder Some Samsung here have it and then use the second the rhythm engine as ah rob revising for to see whether this is a general problem, that because of the lower conversion rate or difference between the devices. So let's just see off how it works and considered. It differs and it differs a lot. So again it can be either, uh, device. And for which I wouldn't suspect to be the problem, but probably more of a resolution. Problem assess. We expected as be inspected in a previous case. Let me change them. Engine for mobile that was in full to screen resolution and see how it how it looks. So yeah, it basically just confirms the date that we saw in the browser cases that the 3 62 6 40 is very, very bad resolution in terms of conversion, right, which is assuage can see just zero point on the salvage. It's really, really low. And the 2nd 13 for 12 to 8 46 It's much much by the 1.6 to 1. So, yeah, in case off merchandise store. This is just the information that the mobile experience is very bad on this resolution and much better on slightly bigger one. So please check also this date that can be very helpful and more of ah, information if you're going to the right directions with the resolution or device stops or something in between that feel free to play with different That mentions combination with the brand with the screen resolution as we have it right now, with maybe operating system, maybe with the browser also can be a case. So just work with the information we already have and combined them, and you will definitely find a place for improvement. So that was another tape. 10. Site content: This is the lesson we will talk about website content. First of all, it contains all the pages, so we see which one are the most popular. It's common that you'll find their home page and Carter Page is This report is very nice, but sometimes it's too granular, which means it has too many lines in report. That's the reason it makes sense to implement content grouping. It means that in the moment the tracking beacon is sent into G A and their information about what's type off page it ISS, the very basic content grouping can be just on a couple of page types like home page product. It'll page category page, check out contact information and thank you Page. So let's dive into report and explore it. So, um, now we're going Teoh change our main tap. So we're going off from audience and we're going to behavior tap. Don't worry about acquisition right now. We'll get back to it and upcoming lessons. And now we're going to behavior and to site content, and we're going to spend most of the time in all pages report so a. So I think it's quite easy to understand from name. It's going to be about the pages. So let's minimize navigation. And what we see here is actually report for all our pages on Ah, quite Communists. That or I think not common. But in in 99% of the websites you will see is the most popular page. Your home page, which isn't an exception for Merchandise Store. And also one of the top pages will be some signing Bages Basket or Carter Page is so this is something we're very common. So don't don't be scared if you will see such a thing the first place when it starts to be slightly more interesting. Yes, I think if we if we see couple more lines, let's say 100 on we can explore. What are our top? Are the most popular pages of our website, which, which also gives us a hint off while the users are browsing. What was our what are they basically looking for? So so feel free to explore this report and find out whether you don't have some pages would you wouldn't expect to be that popular and try to find out why is it so nice? Him here is If you want to see if you, for some reason I want to see immediately some page because you've spotted something you don't have to copy the HTML address. Yeah, there's a nice thing you can only click on this small Aiken the small arrow, and it will automatically show you the page, so we'll click on it way. See a card in which I obviously don't have anything, but anyway, it's It's nice. Nice shortcut, which will save you a couple of clicks. So we have, ah, set of metrics here, page views and unique bedrooms. And as we know how uniqueness is defined in N g. A. It basically accounts only if it counts maximum off one time procession. So Ah, this is why Ah, the total number off unique bedrooms is always smaller number than pages. In extreme cases, it can equal, but but I don't think I've ever seen such a such a thing. So then you have, ah, next metrics, like every each time on page entrances bounce frayed exit rate page. With that, we will cover it in couple off moments. But here comes the tip number 11 and it's something called content grouping. Let me explain what it is. Ah, right now we're We're looking on all the pages, right? Every line is one page, which is nice and completely okay. View for a lot of analysis, but sometimes these reporters to granular, What I mean is basically that we have too many lines. A Zoe can see 445. And this is this is exactly the place and the case when having implemented content grouping is very nice thing. Ah, let's at first see what Google has here. And ah, we have here three content grouping types. 1st 1 is brands, product categories and closing by gender. Let's maybe click on the 1st 1 and I think you you you you know what exactly it does. It's it's it's basically grouping your pages by some, some higher level so we don't see all of them, but just a couple of lines. Which group? Let's say, as we can see in this case, brands also, all the pages that are related to some brand are grouped together. There's the first line, is not said, which is the top one with the highest number of pages, and if we click on it It's not said because probably not all of the pages are related to some Brent right home page baskets sign in store. They are not related to brand. But if you, for example, used ah content grouping based the based on page types a sweet X we mentioned so you would have probably 5 to 6. Klein's here, like whole page product detail page, category page, maybe some car pages. Contact Bages and information pages. I would say this is the completely sufficient content grouping for most of the pages, so you can see how they perform, how popular they are, how they perform from from the perspective of bounce rate exit. Great. Ah, how many times are set off the pages, entrances and so on and so on. So there are a couple of cases when having gone 10 grouping is very, very handy. Also, this report is one of them. But then also, one of the upcoming lessons will be about side speed when where this comes very, very handy. So ah, let's go back to pages and one of the things we're going to show, which is also tip number 12 is navigation summary. It's ah different vertical tap. So let's click on it and see well how helpful helpful it can be. This can can seem quite messy if you see it for the first time, but there quite often or common scenario when you want to find out off. Ah, where were the traffic to particular page comes from and where those users are going And this is exactly the report. It will give you answer for this question. So the first thing we have to do is to select the page we we want to explore. So, uh, feel free to use the filtering here. It covers all of your pages, but to so it quite quickly, I'll just click on one of them. So let's click, for example, on this Google redesign office and see what we have here. While these reports tells us it's telling us that the most of the traffic comes from Google redesign accessories, they come to this side and then they mostly going to Google, redesign accessories again, Google redesign office quickly home redesign and so on. So it also tells us off how many people or how many sessions exited here, which is 18% or 19 almost and the rest 82% are distributed by this number. So this is three reporters is basically quite easy, for example, for some particle or campaign badges which you have anyone to find out. How How did user get to this page? And, well, where did they go afterwards? So, yeah, feel free to explore. It can show up to 500 lines, which is more than enough. I would say, I don't think I've ever used more than 25. So this is important. This is exactly what this report's is showing you. Let's go back to explorer tap. Um, you and I have electrical me one based still with click on the old all of them on This is what we have. One technique I want to show you right now is called ah, sorting the report not by the full number or but by default, sorting by by but by weighted average. What I mean by this is if let's say, let's assume a scenario that we want to explore our worst performing landing pages. Ah, the nice approach would be to sword our pages by bounce rate right, which tells us of how maney how many sessions have you'd only one page and exited by particulate page. And if we click on it, we have a lot of pages with 100% bounce rate, but only very small William off pages right for one than we finally have something 101 170 30 day 26 so on and so on. And this is something we want to award right, because these are not relevant numbers. If most of them have just one or two or up to five pages, and this is why different four type is very handy so we can switch from default to wait it . If you want to find out the calculation, you'll find a link in the description. But it basically takes into consideration the relationship between Bounce Trade and the Bay Jews, and it's swords. The pages by by this waited ever it so we can see it's not like, ah, from the highs to the lowest Bob, but by weighted average. So the way a one of one of the ways how to interpret this is that it's basically showing as the pages with the highest potential for improvement. So as we can see this page Google resigns for by brand new YouTube has, Ah, I would say, quite high volume, off page views and also quite high bounce rate. So this is something we definitely should look on and try to find out why this bounce rate is so high here, right? The thinking in context is absolutely critical here. We do such an analysis because it doesn't have doing anything if it's for example, Contact Page. But if it's if it's not, then probably it shouldn't have that much that that that high bounce rate. If we go to adjust to the third line, we can see there's a huge bombs red right, 80%. But then, if we if we look what pages, it is a fake use. So this is probably quite OK because people are looking for some information about the shipping or or the payment, and they find the information and they leave, which is which is okay. And it's not that that ah, medical problem for us, but anyway uses this sort type on. There's one thing I have to stress out. It only works for the percentage metric right, So if I will click on entrances. Ah, it wont be available, so it only works for for percentage metrics. Feel free to use it. It's very easy to do. For some reason you will want to sort something, and the weighted average won't be available, and you'll get a lot off pages or perfect sources or anything else. You can also do Emanuel filtering not on li buy them engine, but also by metrics. So let me show you what I mean. If I click on advanced, we know how to filter primary and secondary damn engine. But it's also possible to filter a metric. So if you just click on the drop down menu, you can select any metric you want to. So let's say entrances. Ah, greater than 100 you can also feel there filter the report by by metric. So as we can see now, we only have 31 pages, which is not ah, a lot off lines to to explore and to analyse. And it's quite easy and quick Teoh spot where the problem or opportunity can be so yeah, I used to use weighted average another metric we have here, which is also tip number 14 ISS percentage off exit or maybe exaggerate. This is easier to understand. Ah, you might think that it's quite similar to what boundary it ISS. But it's not. As you can see, the values differ. Ah, the formula. The that works here is that every bound is also on exit. But not every exit is a bounce on its because a different calculation exit is calculated Onley in a case that particular page was viewed and was also lost page off a session. So this is the difference bound for it is only calculated if the bay jewel's the Onley page viewed uring the session, whereas exit rate, it doesn't have to be. So, uh, this is why the number is, ah smaller than, ah, bounce rate on again the way of how to use it. It's probably to find the pages that are the top so called killers off sessions eso sorted by by this metric, um, and again you will find that there might be some ah, some pages with were very small William off page views and unique Pedroso assess. We have it refilled over here for the pages on Lee greater than 100. It actually makes quite a sense. So if I click on apply, I have the same numbers. Ah, and we can see that now It's sorted by the top exiting pages. So do the same thing in your case explored. If you will fund if there is a page with high exit, great on shouldn't have that high exit great again. Critical thinking is, or or thinking context is critical here because if you, for example, find that there is a page with contact information with very high exit great, it's probably okay. If it's a product detail or category page, it's probably now that okay, so the easiest thing to do here ihsaa to see the page to display the page and try to find out why is it so on? If it's now that clear just from viewing the page denies technique is to switch to navigation summary, which we showed just couple of moments ago. Ah, filtered a Their particulate page tried to find out by seeing Where do the users ah come from to this page? And where did they go then? And it can give you a hint. Why is it ah that much? Ah, why is it that much a lost page in the session because they probably are coming from a different pages that are not related somehow to this page, and they don't find what they were looking for. And this is why the exit. So it's nice and easy technique to use a navigation summary for for finding out, off where, where do you kill most of the sessions or which pages kill most of the session? It doesn't have to be the problem off page itself. It might be the problem off. How are the users linked to this page? So, yeah, play with it on trying to find out those those ah, the red spots on this was tip number 14. Number 15 is the lost metric in this report called Page Value. It's my quite favorite one, because what it tells us, As you can see, it's with with dollar. So it's mostly monetary value if you don't have an e commerce website to you and you have some girl CW fund here. Also Willy, but without dollars anyway, what it tells us is how much every page contributors to all of the conversions in this case , off the revenue uh, it was good to know Here is to understand off how page value is calculated. So let me show you a simple illustration off this calculation, and then we'll get back to the interface. So he was example of how page value is calculated. Let's assume we have a session Number one, which starts with Page eight, then is followed by page do and ends with no conversion. Then we have a session number two, which starts exactly as the 1st 1 with Page A and Begbie. But then there is a page see, and the conversion with weather over $100. And now let's see off how the page value for these three pages will be calculated. So we have a simple table, and the calculation is quite simple. Which is the total value off conversions in which particulate er page was viewed, divided by the number of sessions in pit in which particulate page waas well, so for Page A. The total value of conversions off sessions in which page was viewed its 100 divided by number off sessions in which page was viewed, which is to so 100 divided by two, is 50 for baby, it's exactly the same case. So the total value off sessions in which Page Waas Field is 100 divided by a number of sessions in which page was fueled it's too. So 100 divided by two is 50 and now we're coming to the page. See? So, uh, the total value off sessions in which page see will seal this again 100 exactly as in the previous do pages. But the total number off sessions in which Pedro's field is only one. So if we do, I 100 buy one. We have $100. So, uh, this is why BHC has higher in this example. Has higher page better in this example. And it's basically trying to tell us that if there was no Page Seaview, there was probably no convert on. And this is exactly the the logic off this off this metric is the contribution off particulate er page on total conversion value. So this is how credibility is calculated. And now let's go back to to the interface and see off how it can be used. So we're back in the interface and we're going to show how to use page volumetric. Uh, so the first thing you would probably do and I would do to is that I would sort report by this metric. So let's do that and see what we have. It's probably going to happen also in your case on, and you're going to see the 1st 4 or five pages having a very high page value, which is also something we see here. But if we check which pages have this high Beijing, really, we can see it's a payment page order completed, review order and two or three more. Now, as we know how Page Billy Risk Alka calculated, it's quite obvious that this pages have very high page Billy, because every user that converted had to view them. So it's It's normal that this pages have, ah, high page value, so don't spend too much time by exploring is extremely high. And maybe what makes sense is to scroll down a bit blot. Ah, a couple more lines I have here. Ah, no. 250 which is probably too much. But if you plod, let's hear no top 50 and it will scroll down a bit. You'll find out that there might be some particulary pages, having very nice page, really on the case when it come when it's very helpful. Using this one is when you have some content part of your website, which means that you are writing something that's a block posts or some valuable content for your users. And if you want to find out whether this content works, this is ah, great metric because it tells you that once users ah consume this content. This is how much they do or do not convert. So try to try to see which pages you will happen. Your data and it can happen that you will find some pages that you don't promote. That much on your pages are you. Don't link on it that much, but they have very nice Page Valley, which is a signal for you that you should somehow promote them. Thio more attractive pages off or more attractive parts off your pages to Teoh having more visible because this number tells you that once user see it and consume the content, they convert Morton the user that don't so um, yeah, please do it. It's it's very easy to do on this metric will tell you which content to promote in which actually works in which is not what probably makes sense. Also, to do here to avoid some outliers is to pre filter it by by some nice volume. Off Pedro is to avoid some pages that could have probably 23 pages and then some. Ah, high conversion, which was which would distort this metric, so definitely makes sense to pre filter it by at no. Some some numbers did that make sense in your case, and then look on the values. So this waas tip number 15 please use It's very nice metric and easy to use. 11. Site speed: and here we are heavily discussed, stopping recently, and it's well deserved attention. Side speed is currently something. Also, Google pushes a lot off before, too. None of us likes waiting for something, and the base holding is not an exception at all. If you're familiar with the topic, congratulations. If not, you might be surprised how multiplication effect fast loaded websites half let's play a quick game will simulate the three D connection and see how fast or slow the pace alone. Try to stand up and sit in the moment you would abandon the page. There are many studies showing how much side speed effects conversion rate, and you'll find few of them in the description. Couple of things were mentioning regarding fast loading website. It affect almost every discipline off online marketing. First off, all user experience. Smooth user experience is absolute basic when converting your visitors to customers and side speed is sometimes the only thing that matters. Just remember the game we played just one minute ago. 2nd 1 s seal. It's not a secret that Google prefers fast loading websites, especially on mobile devices. It's one of the signals Google definitely takes into consideration when the ranking the website the 3rd 1 performance campaigns. The simple rule applies here. The weather, the conversion rate, the better the performance campaigns, effectiveness automated bidding strategies takes your website general conversion rate into consideration. So why not to spend your money more effectively? Let's have a look on side speed report in detail. So I hope they enjoy the game. And you find it helpful when it comes to appreciating your colleagues or bosses that the occasional time is important and that every second really matters on. What we were going to do right now is that we're going to explore them oil, more details. The side speed reports we have available in Google Analytics. So we have to go decide speech step and then to all review. And, uh, what we have here is definitely worth checking as the first thing ISS, what is our General Basil? It's time with me. Maybe it first slightly changed the day trains to have more data. Okay, this is enough. So the first thing worth checking is general bagels time, which we can see in case off Merchandise store is 4.43 this is so something slightly worse than will user expect. Well, they do expect you. You will find a couple of links and descriptions off. What user or how Quick User, expect your page to be loaded. It's something somewhere between 23 seconds. And you also find out how much off how much revenue or conversions every second of the late costs. You, onda, uh, you might be surprised that three every second of the high can cost you Ah, between 4 to 6% off revenues, which is a lot. So definitely check it. Uh, And if you're somewhere between or above 67 seconds, this is definitely the first place to start with your website improvements. Because even if you would have the exactly same webpage and one was foster and the 2nd 1 slower, definitely the foster one would perform much, much better. Uneven. One second off page loads. Time or sake. Basil time can make a huge difference. So please check this number and tried to squeeze it as much as possible. Ah, this was actually the number 17 would also includes checking your ah visible time by browsers because, as we know, different browser browsers works differently and they can blot your website also differently when it comes to baseball time, so please check check it. And if you find out some huge differences, try to explore that in ah, more detail. Uh, another thing worth mentioning here. Eyes. The payload time isn't the only metric we should follow because what actually basil time means is the moment when all of your contact waas loaded and rendered in front of user. And by content, I mean all all the other HTML content. Then allow internal and external images JavaScript files, see if this falls and so on and so on. It's quite a lot of content, and this number of facial time doesn't necessarily have to mean that this was the first time that user could interact with your website. The moment is actually or normally much, much shorter. So, uh, you don't have to take it too seriously, But also don't ignore this number. It's definitely good starting point off where we might have a problem. So place jacket. And if it's enormously high or not enormously, maybe if it's just only six seconds, try to squeeze it. Ah, next numbers. We're going to check our page timings, which is tip number 18. First of all, definitely worth checking ISS Ah, comparing particular website perpetual time to website average, which is pretty much all these reports tells us a Zoe can see the average for merchandise store is 4.43. Ah, and even here in the first top 10 pages, we can see the 2nd 1 which is good, will redesign. So by brand YouTube is 81% slower than I'm sorry, Paige, All time average which is if we do just quick math about No. 7.5 2nd eso 82% is definitely something Google guys shouldn't be happy with for this particular upside. So you can just going pardon page by page and trying to find out where are the slowest website and trying to fix them because it costs you a lot of money. Ah, and ah, nice thing to have here ISS having implemented content grouping which we already know. What ISS because we explained it, explained it in a previous lesson with all pages and is the grouping your pages by by some attributes, which can be an ideal case for for this report things grouping is based on page steps, which every website should have more than five or six of them, like home page product, people pages, category pages, some cartridges, contact and information that should be it. So if you haven't implemented and use it, you may find that, for example, all of your political pages are much more float much, much lower than, for example, category pages. So this would be your case to start because you can go page by page. But if you have tens of thousands of pages, you probably can't this approach or can. But it'll take you a lot of time. So content Grouping will help you find maybe some global problems that you might have for for particular page types. And I stress again page types. So I just click here on ah content grouping available here, which are brands, and even here we can see it's quite useful because, as we can see, there are many pages included in this content grouping by brands. But even if we compare a Google and YouTube, we can see that the Google pages are off course. Brand related Google pages perform mud bank better from Pedro time than than YouTube. So this can be a hint for your developers. Hey, guys should check our due to pages because they really sucks in terms off baseball time. So Contin grouping can be very, very helpful thing here, especially when having based up so you can find their available problems for particular bass steps. So, yeah, this will spur. And I think, uh, next report we're going to explore, which is also tip number 19 our speed suggestions. Uh, so let's minimize this one. And what we have here is something similar to previous reports. But it has one more column here, which is paid speed suggestions on this. Is there a nice gift from Google because we can see that we have, Let's say, for the first page seven page beat suggestions. And if we click on it, what it does is that it actually redirects us to Google bait speech tools, and this one is called based feet inside. And what it does for us is that it automatically analyzes the website, telling us how it performs when it comes to side speed for both mobile and desktop. And that's not the only thing it does. Is it actually tells you what you can improve and how you should do that. So all that hard work with finding out very have problem in cold. Is this done by this very simple but but very important tool and a So you can see we have here do for 67 optimization suggestions. And it's not only telling us that we should leverage our browser catching, but also how to do that. And, uh, and also not in this particle suggestion, but let me show you what I mean. If we, for example, click on a magnifying javascript, it also will tell us how Maney or percentage reduction it'll do, so we can see that we can buy buy magnifying this file. We can save 65% off facial time, which is the great thing so we can find out what is the best thing to start with when it comes to optimization of your cold or external and internal files. So, uh, yeah, use this file. It's announce or not filed this tool, it's it's great. I use it quite often, I would say, because it does you well. You should have to do if you want. Teoh lower your page load time, So Ah, very nice thing to do. Ah, thank you, Google for that. We're very grateful. So yeah, that was it. A big speed. Ah, if you're again, just want to repeat and stress it again. If you're not familiar with the topic off off page speed, please pay attention to it. It's It's really, really crucial when you want Teoh increase your conversion, right, So that waas based beat 12. Internal site search: side search is something every website owner should have on her or his radar. The reason for it is very simple. This is the date advisors give anonymously and is very, very precious. They basically give you the information while they were looking for. This data can be used for improvement of your product inventory keyword to start to run paid campaigns, changing your information architecture and so on. A lot of money is sometimes paid for such information in external researchers. And what is even by their G A gives you information about how relevant results can you provide to users who use internal search. In general, users who perform aside search have much higher conversion rate than once who don't. And sometimes it's not an exception when they generate almost 50% off conversions. This is the users group you should care about when thinking off retargeting. Now we're going from side speed two sides search. Um, let's go to sites of step and into usage. One. Uh, well, this 1st 1 tells us so. Let's minimize Navigation is basically very simple report. It only has two values inside. Still, state is them engine, and it's dividing our sessions onto two groups with with outside search and with side search on and as mentioned, it's very, very common that the users who use side search are very, very likely to convert, comparing to those who don't on as you can see now, this is exactly the case because in case of merchandise store only 4% off sessions used side surge. But if we scroll to right, we can see that their conversion rate is five times higher than, ah, whizzes who don't use our visits that don't use site search. Ah, and what is even more interesting is, ah, look on the revenue we conceded only 4% off session. Generally those 24 or 24% of revenue, which is 1/4. So this is a very, very small group. The generate significant volume of revenue. So it's It's definitely very important thing to know, because now you can. If you know that your side searches working burger well, you can, for example, try to somehow promoted more on your website because, you know, once the user use it, they are likely to convert, or you can use those users for targeting if they haven't per chased or haven't haven't converted in general on your website. So this is this is quite common case that that small, small group of users can generate that significant amount off revenue. So definitely check this one. This is step number 20 on may be Let me show you off. What exactly aside for searches? Because just to make sure that they were gonna understand what this feature actually means and things that the thing we're going to show is how the data is collected. And it might happen that you don't see here two lines on Lee one, which means that you either don't have this feature on your Web site or you just don't have it. Ah, sat in your G A administration. So let's go to my favorite clothes shell, which is Colombia. Uh, let me show you how the data is collected. So, uh, what I mean by side search is this small feature which a lot of websites half. So we just time there, some some curious and keyword on it. Shall we in shows you the results. So just type, for example, jackets as something we have here really doesn't matter what we die there. But what members right now is to see off what happened with the URL address? A So we can see I wanted to trap jackets, but I made a mistake. But it's okay for for this purpose. What is important to know this is the perimeter in your address, which we have to remember or write down and then set it up in Google Analytics on where this parameter is in this case is actually only the letter. Q. How do you find this parameter is very easy. It's always something that is after a question mark on before equal sign. So in this case, it's cute, quite comedies to have their curie as served something something related to it. But it really doesn't matter what is there. So it could be a B. C D equals decades. So if you don't have it, if you don't have this data in your G A account and you have this feature on your website, try to use side search and see what happens with your old rest and remember the parameter, after which ah, impotent keyword s. So in this case, it's cute, and this is the only thing we have to remember eso Now let me show you how to how to set it up in Google analytics. So we have to go to administration, which is on the bottom. Ah, left corner and ah, it's always set in a view, something So here we are. And if we scroll down to the very bottom, there is ah, site search button, which have to be It has to be blue, which is on here. And this is exactly the place where we put the cure parameter which we remember. Waske you. So the only thing we have to do it just type que here and click on down or safe. If you have I I'm n g a in merchandise store council. I don't have permissions to do that, but if you if you have ah sufficient permissions in your in your account you should be able to do that so we only have to type your cue clicking down or safe. And from this moment, the data will start to collect in your account. So pretty simple in in terms off set up Engl analytics. So let's go back to report. Here we are. So we know how to check it and how to briefly use use this information. So we are here and now we're going to the second that which is search terms. And I don't think I will exaggerate when I will tell that for those of you have this, uh, feature offside search enabled on your website and you also collect this data and Google Analytics, This is you're probably digital gold. And let me let me show you. Why do I mean that? Because you actually see all search terms that your user search on your website, which is I would say amazing because they they are anonymously telling you what they were searching for. And now we can see. Okay, nice thing is to know, off to know what did a search for. Okay, you can see. So in case of Google Analytics in ah Google Merchandise store, the top search term is YouTube, which has almost 900 unique searches. Uh, which is nice and now how to use it. First of all, if you just just exploring this first terms, sometimes it can show you the actual trends for for some topic, which you might not notice if you if you're not familiar with your business or you don't follow the latest trends, this are the date that it can help you to support its something news on the market. And you might not have a there. We don't have content for it or ah, or your website is not prepared for something. You. So this is this is one of the usage of it. But what is more important, I think and makes more sense is actually the third metric, which is percentage of search exit. And this actually tells you off how good results you can provide to particulate er search term. So in case of YouTube, the 1st 1 it tells us that 28% of users leave the website after seeing the results off YouTube search terms. So user came to merchandise store he or she tied the YouTube into internal side search. And ah, almost 1/3 of them left immediately after seeing the results. So as we can see the averages 24.3. So this is slightly above average again, a good place to or a goose start point is to start typing search terms into our engine and trying to find out what the re provide villains results or not. But that would be quite quite, ah, long process, I would say, because as we can see, we have almost 10,000 search terms, so going term by term would be quite time consuming, I would say so what we can do here is to sort and by the top cert exits. And again, if we only do this, we will see that we have a lot of 100% exits body only one sort with which we can see. It's It's quite funny sometimes to see, but users conserve, so I dont know Google it Search eso What? Quit interesting search term or or the ladies or or 55.74. So what actually makes sense to do here is to change the sore type again from default toe weighted one which will again consider the volume off total unique searches. So we have relevant results and as we can see the worst performing, uh, search term or I wouldn't say worst performing. But it's indicator is not performing where well is the word ingress, which I don't know what it is, and I'm not going to go looking for it. But anyway, I hope you've got a point so we can see that 66% off users will search for this particular search term leave immediately after seeing the results. So that's quite a lot. And please play with his data exactly the same exercise so we can see off which search terms users search for any can provide relevantly results. You are. And if you can do that, you're losing probably the most converting users groups I know in in Google Analytics, your side search user. So please do that for your top top search terms. Eso This was This was, ah, metric. That tells us off how how many people are we losing after in putting some search term, and now we're going to do a slightly different thing, which is tip number 22 it means that we're going to switch to different vertical type, which is e commerce. So let's do that and see what they thought we have here. As in previous case, we were able to see off how people are leaving a website after typing a particular search term. Now we can see e commerce stats and what we are interesting in here is seeing conversion rate, which tells us of how ah, people are not people. In this case, sessions converted after searching for particular search terms so we can see the average conversion rate after using after typing some search term is 3.25 we conceded YouTube is working quite well. 7% considered. Android is just 3% and salons hole. And we considered the 4th 1 tour jersey. Ah has very, very low conversion, Right. So I would say that this one would correlate with also quite high exit rate, which we have to check in in a previous ah, metric said inside usage. So, uh, yeah, again do this exercise. Ah, definitely blood lead 100 lines, too, to see which which keywords convert well, which one don't. And what is what is nice to see that Ah, again, this this particular searched ingress If you remember it waas the top one. If we sorted report by exit rate and as we can see here, uh, zero conversion rates. So ah, as you can see, there is. There is a very strong correlation between high exit rate and and a very low conversion rate. So again, do this for these data. Your users are basically telling you Hey, I was looking for this one and really suck on this on this search term. So police tried to fix it or trying to find out about this search term is and and do something with it because you're losing a lot of money by this. So this was a knee commerce e commerce set of metrics for search terms. Very, very brushes data. We're going next, and we're going to switch from search term to search pages, which is our tape number 23 on this issue may see here, and I think it's it's again. Very straightforward. Start Page report is basically report off the pages on which users mostly used searches. Internal site search, which means that the very top page on which users actually type some something into so do into search engine is stop a slash hole. So in 25% of cage off cases users youth side search on this particulary page eso again. It's known as information to see off where probably users may may lose the information or they might not know where they are. So the used sites shirt or or some some completely different reason. But this is something we have to find out once we see the data. Uh, so what actually makes more sense here is to even use the combination off those two dimensions we just showed. So it's a search term and the start page, and this is absolutely amazing data. So if we use is a secondary damage in here with s search term, well, we will now able to see it's, too is Teoh see at what page users were searching for So we can see that our top combination ISS shop a slash home and then chromecast. So this waas the absolute Domco top combination off search term and the page on which they were searching for this search term. So use this combination is it's truly amazing that they have one of the best in the whole Google analytics, So please use it. There's a lot of money in it. Uh, and there's one thing you might notice, and it's the the entrance start page what it is. First of all, if you do any side search analysis, you have to do one thing, and this is step number 24. You have to exclude all the searches which have start page as an entrance on the reason for it. It's very simple because this actually is not using a side search. It happens quite often. Did that if you, for example, running Ah ah, performance campaigns. You sometimes don't have ah, dedicated landing page for it. So you basically use your side surge and you start to link on this lending bed which is actually created by the US by your search engine. And you you you just taking people to that landing bait, which which then is able to Rick way we are able to recognize this page as, ah, a zoo page a start page with the entrance wary of this of this dimension. So this definitely is not a side search or the second case can be when the when some, uh, Robert from from Google or some different search engine is is crawling your website and also is indexing this this space so it can link on it in its organic search results. So if you want to have a clear date out that are on Lee from using the side search. You have to exclude and runs from from any reports. So how to do that? Is there a symbol? We know that because we know how to filter primary or second the rhythm engine. So we just click on advanced. And now we have to select start page and instead of in included, you have to exclude start page containing entrance. So now we click, apply. And now we have ah, clean data. So this is something can work with eso Please do that. I can't imagine that you can have quite a lot of start pages. Ah so called entrances which are not actually start pages because these are the the basically the landing pages used for campaigns or from organic results. So please do exactly the same thing even if you are in ah, search terms report And I don't mean Onley in search pages because those 22 reports or to them engines are stick together. So if you want to have clean that, I also have to automatically use here secondary damage and start page, and you have to exclude it from from results. Otherwise your data can be distorted and sometimes some quite quite heavily distorted. So again, the same thing excluding start page containing entrance. So now we have green data and we can work with it. Ah, again, if we go for one more second to search pages exactly the same tap also works here so we can see from From which page after using ah, the side surge and ideal cases to use your second rhythm engine search Term off How it looks then, from the perspective of convergent rates which keywords are are not keywords but but search terms works fine and which are not. So please use these days tough for for multiple reasons you can change your information architectural Based on it, you can start to running paid campaigns based on a key words. You know that people are typing into your internal search engine. You can start using those users who actually used your searching and your surgeon and haven't purchased it for retargeting. Let's see so very, very versus data. I have to stress it again. So this Waas ah search engine 13. How to use UTM parameters: UDN parameters are used for traffic trekking across all traffic platforms. They are nothing but perimeters. We have to add to the Earl of Landing Page address if you want to travel. It's a BBC campaigns newsletter, social network posts, book boasts and so on. It means all kind of traffic except direct and organic. One. There are five of them. Source. Medium campaign content In term, let's have a look on a couple of examples of how it's used and show it to that will help us to create you tm parameters correctly. So, guys, we're going to show how to use you. Tm parameters. Maybe it first small window from the history. You might wonder what actually you tm abbreviation means on its urchin tracking module. Urchin was a company which waas a predecessor off Google analytics, and it was acquired in 2000 and five by Google. And then he was lately ah developed Ah, as a Google analytics, but for which we are using right now when we love it. So this is what community means. It's still used. So, uh, you, tim parameters enable us to distinguish between various traffic channels. Let's assume very simple scenario we have this your address, which is my website https. Probably thing that sees that slide hyphen slash e and slash and I'm running and paid search campaign in Google Edwards. And once I want to see this traffic and Google analytics, that is the this traffic iss exactly from this campaign, I have to take it somehow. And how I do that is by adding this small fractions off text into finally rural address which will look like this. So by only having this one, we would see this traffic as direct one, which is not ah, because it's from this paid campaign. So this is what we are heading here. The beginning off the final you're our address is the same https about Brett sick that sees at slash en slash And then there is a question mark beauty in medium, equal CPC you teams force equals Google and you team campaign equals g a course. Don't more Now about the question mark underscore signing equal sign because we're going to explain it. Ah, lately and what is more important, we will show a tool that will do it for us. All this necessary special characters in there. So this is what actually UDM parameters are. And now let's repeat them. There are five of them. Source, medium campaign content and term the quick explanation off what every perimeter s and what should be its value. So the 1st 1 you tm source, it's mostly the domain or at platform Name s O, for example, if we have ah banner campaign on ah, New York Times. So this is what should be there as a domain New York Times Or, if we are running the campaign in Edwards, who are there, can can have their Edwards or Google. So this is what we should Ah, but there, Then there is a 2nd 1 You tm medium. This is ah, always on at type. So if you, for example are running a banner campaign, we should have their Benner as well you or if we are writing a Facebook post than there should be opposed. Or we are running a paid search campaign there for B C B C, which is the most commonly used abbreviation. Then there is 1/3 1 you tm campaign. There's always ah campaign name. Feel free to basically use anything there but try to use something meaningful and the reason for it? It's We're simple because once you will do some comparison, I didn't after a couple of months try to name it somehow, easily to understand. So we don't have to think for half an hour. What exactly was this campaign about? So try to use it something in in a way that he will easy understand, like at Easter Sellout summer campaign, school campaign. And so long. Then there is 1/4 1 you team content. This'd is Ah, I would say, the broadest one. There are no rules or recommendations what should be there? But mostly it's for add details. Uh, which, for example, tried to imagine we are running. Ah, better campaign somewhere and we have multiple banners and we want to find out which one works the best. Which one brings the most traffic, and then which one converts the best and so on. So this is the perimeter that helps us to distinguish between those those banners so it could be just some small naming off the banner like blue or red. Or there can be a vendor size or better damn engines or creative name or something similar , so feel free to use it. It's the broadest one, uh, and play with it. And then there is the last one called you TM term, which is dedicated for keyword and key. Word on Lee. Please try to remember this one. It's quite common mistake that people tend to put their something that does not belong there. It's on Lee. It's It only should be used for bait search campaign. And there should be a key word on which you were, which on a cured on which your adults sold. So this is the only type off campaign where you should use key work where you can term perimeter, and the reason for it is were simple because there's a dedicated report in Google analytics for it. And once we put their something that don't belong, there will just create massive number data. So please remember this one. You, tim, term is on Lee used for paid search campaigns. So, uh, now we're going to show a couple of examples off how the finally our address, including you, Tim Parameters may look like so the first example. ISS. Let's assume I'm earning paid search campaign in Google. And this is how the URL address should look like the 1st 1 you tm sores, which we know is the either dooming or at platform name. I'm running it an Edward, So I'm putting their Google, which is a Google platform, adds type is it's ah, search campaign. So the most common ah, medium value for it is CPC, which is a cost per click can either is BBC, but those two are the most common. Definitely. I named this campaign as a G A course because I'm running a couple of more thin. There is a U team content, which is the broadest one we know on, and it should help us to encourage the information about the campaign. And I'm putting their benefits because once my ad is shown, I had there a couple of words describing the benefits off this course. So this is why put their benefits and then the last one, you tim term keyword on a I. This this ad was shown for a cure Google Analytics course. So this is why I put it into cure because it's paid search campaign, and this is how the final your address will look like? So all of the U team perimeters are there again. Don't worry much about the structure off it all of special character, because we're going to show a tool which will do it for us. So, uh, that was a bait search. Another example. Banners. Let's assume a scenario that I bought one million impressions on on the reddit dot com. So again, we're going todo start from the u T M source. So I said I bought it on a ready dot com, So I'm putting their A domain name a Z u T M source, which is Read it. Then there is a style which we know it is a form of the ad, but we shown which is a banner you can either advantage or display for four banners. Ah, better campaigns by the noble 20 prefer anyway, he was consistently. Then there is a campaign in which is, in this case still the same G course and the U Tim content, which is for a detail. I put their the dimensions of the banner 200 to 200. So this is it because I'm running multiple banners. The wider ones, the bigger ones so I want to distinguish between them. And then there is the last one. You Tim. Term keyword, which is empty because, as we know, it's only used for paid search campaigns. So this is how it looks like. Ah, we only have four parameters here. Ah, Now, uh, or currently there's only one off parameters that is mandatory, and it's ah, you team source. So we the minimum off parameters We use this one, which is a damn source. A couple of years ago, there were three of them. Source medium campaign. Now it's only source. So ah, don't worry that we don't have you 10 term here it will work. So that was Ah, menu example. Another one is, let's assume social network activity. Different scenario. Let's assume I have a couple of thousands off followers on the Facebook and I'm writing. There regularly posts about analytics. So again it's on the Facebook. So we use a Facebook, which is the domain as you team source at stop. This one is the post, so this is it now. We are still in the same campaign J course, and there's a detail. I've put their benefits off tagging because in this post I was describing How good is it to tax some campaign? So this is something that we're doing actually, right now. And this is how our final you are all dress will look like with this you, Tim parameters began. Keyword is empty because it's not the paid search campaign. So social network activity So newsletter is the last one in this case, this is slightly more specific because, uh, there is no specific domain we have or a platform name. Right? We sounded. And we There is no website on which newsletter is displayed or some specific platform name . Right? So this is why I would also year, because there's, Ah, this is this is definitely the most common source and medium, uh, values which are used for a newsletter. So there's almost always a newsletter in a source and in a medium email on as a campaign name, I used their date. Uh, let's assume a scenario that I'm sending a new letters regularly. Let's say every two or three days. So date is definitely the easiest form for me to Teoh distinguish between different newsletter. So this is why I put it there and as, ah, you team content I use here g a course. Ah, you might notice that in a previous examples I use G courses A campaign name on this one. I have it as a U T M content. Onda had a reason for that. Try to imagine off how a newsletter can look like there are multiple multiple pictures, some tax logo, maybe navigation on with user can click and then will be linked to my website. Ah, and let's assume I have their some information about a new block. Boast, I have local there. Uh, I have my recent courses that I'm, uh, are, let's say, public lectures that I have. And then I have one short calm there about G a course about this one that I'm preparing on . I want to see of how many people actually click on it. So this is why I use g a course a za part of the newsletter. Ah, as a u tim content on again, as as in previous two cases. Ah, beauty in terms is empty because we this isn't the bait search campaign. And again, this is our final. You are all address, including beauty and parameters. Eso this were for examples. And now let's have a look on couple off deuce and don't, which is also a tip Number 25. The first don't is Please don't use that. Critics thes are the special characters, which England doesn't have many of them, but especially if you live in a in a central Europe or specifically if you some Slavic language. There are a lot of special characters which, if you will use in Eugene parameters, will be translated into something I would say not reserve flight friendly form, which you go on loose any traffic. But you will see some very strange characters in there won't be easy to read. So if you don't have to for some specific reason, please don't use that critics, please use them consistently, which is something we will describe in a couple of minutes on. If you want to have a blank space in Google analytics, which is especially good for for keywords, please use the plus instead of it. Google analytics then will translate. Plus is a blank space in the interface. Ah, you might notice that I use the blast in the first example, So let me just quickly go back there. It was in a bake search and in the you can come here, which you can see. I have their Google blasts analytics blast course, which then will be translated in Google analytics with the blanks basis. So yeah, please use it. Ah, if you for some reason use a blank space in the game parameter it won't work because you 10 perimeter. It's part of urinal address. And the blank space is not a supporting character there, so it won't work. Please don't leave any of parameters empty. As we said, only UDM source is mandatory. Perimeter wants to use it. So, uh, if you want to use only on Guilford to do that even though I recommend to use at least three of them Ah, please don't leave it empty, right? So don't don't do something like it's shown here that you will have you tm medium equals nothing. It might happen that it won't work and it would be a very stupid mistake. Ah, not being able to to find the tax strafing just because this mistake on if you can please only use small letters because Google analytics is a case sensitive tool, which is something will again. Ah, describe in couple of minutes. So, uh, this were a couple of do's and don't. Ah, And what we're going to show right now are two more tips. The 1st 1 which is a tip number 26 is how to use. Ah, you are all campaign Gilder, which is a tool that will help us to build a fully working. You are all address, including you tm parameter. So let's go and and ah, do that. Ah, but first we have to google it, which is easy. Ural campaign builder On the first link, we have inserts results G a deaf girls app sport. Come if you don't see this, uh, link your serves results on war. You'll find it also in Ah, in this lesson description. So let's get going. And the This is it. It's very simple and straightforward. Tedious? I think so. Let's scroll a bit down and see what we have here. As we can see, the first thing we have to type there is ah, your address to which we want to feeling the user ones he clicks on on on ah, link on. Then we have Ah, five UDM parameters, which with which we already are familiar. So, uh, let's use it. I'm going to use the SE mural address, which waas in a previous example. So it's ah, a little bit signup season slash e m and let's assume I'm running the paid search campaign eso um And it's on a Google in Edward. So what I'm going to use here is Google. Uh, you also have a suggestions. You're under some of the Offiah you 10 parameter. So Google is something I'm going to use as a campaign source. Medium is a CPC because it's based search campaign. So CBC and ah, campaign name ISS Let's say the one that I used, which is a g a fourth on then we have also year ability to use campaign turn and campaign content on as I said that we are running a state search campaign, so I'm also going to fill in a campaign term which is a key work on which my ad was showed on. I wanted to have it shown on Google plus analytics both or us, on insurgency. I'm using a plus here because I want to see that the blank space in Google Analytics. So, uh, this was it. Uh, Now we only have to scroll down a bit and we can see here, uh, full year old dress, including all that special characters as question mark underscore on person equal sign. So we don't have to worry about whether we will write it correctly manually. So it's here. Ah, so this is also tip number 26. Please use the stool. I use it every time I want to tax something. And now comes the tip number 27 which is nothing but checking whether it works or not. And we'll do. I mean by this. We have to copy this address, which we can do by this small button. And or even we can use this folding converter, which is nice recommended to use. Especially if you want to use this link on Facebook or Twitter because you don't want to have the that ugly long you are all address, right? So it works exactly the same way. You just have Ah, nice. Or so we can see here once I shorts. And it's something like this. Anyway, we have carpeted with the longer versions and now we're going to test it. So what do you mean by testing? Ah, we have to open a new tab. Uh, place there. Um, address and press enter. And what I mean by testing is that we have to wait until the pages fully loaded and we have to check is that even after loading on the first page only there are also you 10 parameters included in your address. Because this is the only way how you will then be able to recognize this traffic. Google analytics. So it's a very simple test, and it's worth doing. And why I am saying that is sometimes happens. They do have some very 43 direction on your server, which you, as a user, don't even have to notice. But what it sometimes thus is that it cuts off the UDM parameters. So what it would sometimes do is that it would sorry it would basically delete you, Tim, parameters and your final your address will look like this. So there would be no you 10 parameters which you won't lose this traffic user. We'll see the page and will baby toe will be able to interact with it. But you won't be able to recognize the specific traffic, because once the U R l would look like it looks here, we would see this traffic as a direct traffic and Google analytics, which is not true because it was some different type of traffic which we attacked. But because of the redirection that cut off power you 10 parameters we see as a direct traffic. So it would distort basically both off this traffic cops direct on this specific one. So please do this test. It's just five seconds. Uh, and that was actually these are guys you TN parameters. So please try to take as much of traffic as you can. It's always good to see where exactly the traffic comes from. And in the upcoming lessons, we are going to show how to find this traffic in Google analytics and what is more important, how to evaluate it. So this is for duty, and perimeters are there are a couple of things worth remembering. Please use them consistently, which means that you should create a simple methodology and use every time you take something. The reason for it It's very simple. Google Analytics is case sensitive to which means that B and B are two different characters . So, for example, if you will send traffic from your block post to your web and for the first time you take it s u T and medium equals look post. And for the second time, as you TM equals block post, Google analytics will take it as two separated mediums, which will cost discontinuity of data. When you will filter it, you're not going to lose traffic model, take you more time to get the date. I want it. And as we said, time is money and you don't want to waste it. So please use them by a strict methodology I recommend you to create. You'll find one example in the description. 14. Traffic sources: If we want to evaluate our traffic sources properly, it's necessary to understand how Google analytics recognizes specific traffic sources. The following flow chart will help to explain how it all works. So we're going to show a decision of three, which Google is using every time a new session is starting to recognize what is actually a traffic source for particulary session. So this is how it all starts. We're starting with the loading of a freaking coat, and once it's loaded, the first question it asks is if there are you TM perimeters or gcl idea in your URL address, we already know what the U 10 parameters are and how to use them, and we might not know what the gcl I D or will go click ideas. It's very similar to you tm parameters, which is, Ah, another parameter, that it's past your final Ural address from directly from Edwards interface, and it works on Lee. If you have length your Google Analytics and Edwards interface on what it basically does that it automatically ah is transferring or passing all the traffic dimensions directly into a Google analytics. We don't have to take the manually, but anyway, Uh, the principle is is exactly the same. It's checking whether there is some specific traffic source parameter in your URL address. If the answer on this question is yes, then it's taking all the traffic sources that mention on its values from those you tm parameters, which we know how it works. So this is the first case. If the answer on this question is no, the 2nd 1 which Google analytics is asking if there is a referring domain and what we mean by this is it's asking. Basically, if there is some websites from which user clicked on something and came to your website and the answer is no, then it's clear that it has to be a direct session because there is no specific parameter in you are in the Ural address telling us that it's some beauty and parameter or gcl i d. There's also no referring the main. So it has to be a direct session, so the values off traffic sources parameters will be source and medium will be direct done , and the rest of parameters will have the well us and not sad because it doesn't exist. And if the answer on the existing The Referring Domain. Easy as the last question. It asks if the referring domain is actually a search engine that Google analytics can recognize is a search engine the list of the search engines that Google can recognize automatically. It's pretty long right now. In the past, it WAAS, especially for for from smaller or or local search engine, was necessary to pass this information information into tracking code. But now it's not necessarily really recognize is very, very small search engines. So, uh, if the answer is yes, it will take the domain name off this search engine, which can be also Google Organic. Uh, can we ask Yahoo being so any search engine there is on, and as the value of the medium dimension it will use organic, which is common common name on the perimeters. Campaign and content will be well, we will have the will you not said, because they don't exist. Any of the search term that user used will be available. It will take it so we don't take a cured or if it's not well, we will. We will have their well, you not provided in the last game that can happen is if the answer on ah, is it a search engine is no. If not, it will do pretty much the same thing as in case off. Yes, it will take the domain name off this off, this referring domain on into value of the medium dimension. It will pass the value referral, which is a referring the main and the rest of the traffic. Them engines will be not set, so content, campaign and term will be not set. So this is how how this decisions really looks like. And every time Google Analytics Eyes is trying to recognize what what is the traffic source off particle affection? This is how it works. So it's it's good to know one important thing to mention by default. Guia remembers lost known direct traffic source for six months. Every following direct session is attributed to the previous known direct traffic source. I know it might sound complicated now, so let's have a look at example for better understanding off what I mean. So what we are going to show right now is the way how Google Analytic is attributing a traffic sources to some off direct sessions. I can imagine that the previous definition might not be that easy to understand. So I will repeat it first and then we're going to show this example. So by default, Google Analytics remembers the last known directs traffic source and every following direct session is attributed to that previous non direct traffic source. So let's see this example. We're starting here on the first of January, which ah, when we came directly to do our website by typing a website name into a browser, pressing, enter and landing on the website. So it was direct access. And in Google analytics, we will see as a traffic source to this position. Direct flash. Not so, Yeah, makes sense. We can directly. So this is what we supposed to see in Google analytics. Let's assume that the uring this session we signed up to a new flatter and then after two months, we received the 1st 1 on DA clicked on one of the links in there, and it was stag with the U T M parameters, which we know how they work. Eso we and we also know how Ah, the decision three works. So Google Analytics will take a values off source and medium from you 10 parameters and will attribute this session to newsletter email. So again still makes sense. So far, so good. Ah, two weeks after that, we read some article on reddit dot com When and there was linked to our website, we clicked on it. And there, of course, was no you tm parameter. And Google Analytics does not recognize threat it as a search engine, so it will take just domain name of it and into medium value will pass the referral so we will see readied referral as a traffic source off it. Still, So far, so good, everything makes sense. But then, two days after that session, we came again directly, exactly as in the first case. But this session will be attributed to read it referral, which is exactly the definition we set in at the beginning. So Google Analytics remembers for six months, sort by default for six months. Last non direct traffic source and every following direct session is attributed to that non direct traffic source. So even though it was direct Ishan, it will be attributed to read it referral because it was last in six months. So there's another session which ISS seven months since the previous one and its executive same excess type Azaz 4th 1 and the 1st 1 which is direct access. Which means that we typed directly in a browser tab. Euro addresses for website repressed Enter and we'll end it there on in this case, even though it's exactly as in as in a previous case. But the the time window waas more than six months. So this is why then in again will be attributed to direct, which is correct. So, uh, this is this is how it works. And if this window won't expire so it's less than six months, it will be a still attributed to that previous known direct traffic source. So this is how it works. Not that is fair, right? Not at all. I have the same opinion. Fortunately, there are two possibilities. How to get more precise data. The 1st 1 which I really like. Excuse to them. Engine available since August 2017 and still not many people know about it. It's called direct session, and it has two values which tell us whether the session came. Three white traffic source we see or it was because of broken attribution. If you're already evaluating your traffic sources for sometime, this may change your opinion off which traffic sources brought huge traffic volume for my personal experience is heavily distorted, especially for organic traffic. The second option is to change the campaign time out, which will show right now. So let's let's try to to change and shorten the campaign window. If you want to do that, we have to go to Admin Section, which we do by clicking here on the last bottom. Aiken eso We should see something like this. This is administration section and the place where we can set it is on a property level where we have to click on tracking info and then on session settings. So let's do that. And now you should see something similar. As I see here, there are two options. 1st 1 is session time. All this is not what we were going to do. We're going to do a campaign time out. A. So you can see the maximum time out is 24 months by default is set to six months. As we can see here on the minimum is I think one minute, which is also something we probably don't want to do. So I assume it will be some small. Either are there some couple of days, or maybe month one or two, but definitely not six months. So this is the place where we considered this is actually a tip number 28. Try to shorten this window. And now the question comes, how long this window should be and we're going to show reported can give you a hint of how to shorten it. We have to go to conversions report, then to multi channel funnels and then to time lag report. And what this report basically does it. We can see the official definition. It tells us the numbers off the number of days it's took from the first in direction to the conversion, which means, uh, we should dry at first. Please try to filter it only for your main or major convergent that they use should be only the one filter it. And it tells us that the 66% of conversions happened on the day zero, which is basically the first day when user came. It was exactly the same day that ah, he converted. So now we have the distribution in day. So what I would do and it will never be 100% advice. But just to give you handled, I would do I would try to find the number of days or the distribution where I can cover approximately 80 85% of conversions. And this is what I would sad as as ah, a time out window. So if I can see that here is 66% than plus 4 70 then approximately plus 1% I can see the world. When I when I will reach 11 days, I can see that I can cover 82% of conversion. So this'd is this is the time out. I would say I would sat for for a Google merchandise store instead of six months. So if we do 11 or 12 or 13 this is I think it is pretty much the same anyway. Shouldn't be six months. So to give you a hint, please look in this report and try to find out off where your 85% of conversion level Yes, I can imagine that it can be only a couple of these. If it's very quick decision process or it can be a couple of months if you're for example, selling the cars or or a mortgage. So yeah, it will able where I depending on your business, but this is the place where we can find out off how long or for your campaign time out window should be. So please find it out here and then go back to admin section, tracking info sessions, said things and set it in campaign dime out. Okay, guys, we're going to show two more tips in this lesson. So we're going to acquisition that. Ah, and then Warren will cover all off the main taps and one off upcoming lessons. But we're what we're going to show right now. Are two thinks the 1st 1 is quite common case of how people look on the traffic sources. They often go to acquisition all traffic and into channels report, which is nice, but I don't recommend to use. This is a tip number 29 let me explain you why, If we scroll down a bit, we can see that the name off primary them engine is default channel grouping and as we can see it only has eight lines, which means that by some, uh, certain rules, Google groups all of our traffic channels into some bigger buckets. So on. And let me explain why it's not the best view on the traffic sources. Try to imagine that you don't only have Onley organic search engine in your country. I can imagine that in us, Google is dominant. Maybe there's some small small penetration of being or or Yahoo, but there are definitely a lot of countries where Google is not Onley dominant search engine. For example. In Tikrit public, we have another major player, which ah creates approximately 40% of organic traffic. And if we if we use only this view by the fortunate grouping, we would see all of our search engines in one bucket and thats view that can hide a lot of information from you, especially when we we don't even see of how much traffic every search engine brings. But what is more important if we look on a conversion part off it, we can see that organic search year has 1% conversion rate, and now try to imagine we have to search engines on one of them has conversion rate off no half percent and the 2nd 12% on. There's a huge difference, right? So this is exactly information that will be hidden from us if we look on Lee into this reporting for channel grouping. So let me show you do on which I really love and I use it very often and recommend it to use also do which is source medium. Ah, let's go there. And this is something we now should be familiar with, right? Because we explained, ah, the main traffic jam engines in Google analytics, which are source medium. This is something we see here in one dimension source medium. Then there are keywords, campaigns and content. As I said, Ah, don't worry. We'll cover all of it. Ah, in one of upcoming lessons about what I want to show you right now, we can see that there are traffic sources Google organic organic direct non YouTube referral. Some partners affiliates from Google, CPC. Ah, and as you can see, it has 169 line. So this is the much more granular report and the previous one on it. It's not hiding anything from us. So Ah, what I want to show you right now is Ah, Tip number 30. And as I promise, it's about finding out of how much off the traffic is actually generated by the traffic source we see here and how much off it is. Ah, Only attributed to it because of the six month ah campaign time out window which we now know how toe shorten it. But to show you off how much of the traffic is actually distorted in your case is very easy to do. Ah, the only thing that is necessary to do here is to use the secondary damn engine called direct Session. So let me use it. And let's maybe filter only one traffic source to to see it in more detail. So let's filter only Google organic on and what we see right now, iss ah, that it it divides our reports on two parts. Ah, Their accession on Lee has two values. Yes and no on 2 may be easily understanding this. Try to try to imagine it as, ah question the accession. Yes or no? If there is a yes, it means that it was actually a direct session. But because of the so called broken attribution, this traffic was attributed to Google Organic, even though it waas direct session. So you can see that in case of merchandise store, 22% off sessions are actually or Google organic sessions are actually direct one and are only attributed to our old to Google organic. So please try to try to check also, how would how it works In your case, I can imagine that this this number ah, direct session, Yes, for particular channels can be much, much higher. And I wouldn't be surprised if it's higher than 50%. So please do this short exercise and you might be surprised of how much off the traffic is really, really distorted. And, ah, it might change your perspective on traffic sources you have. So then feel free to filter it on Lee for seeing also direct session competing Yes, which will show us. And now I shouldn't use a filter for for a traffic only for direct session. Because once I will I will use, um filter for second or them and indirect session. Yes, I will actually see what is my my actual number off direct traffic instead. The ones that I was able to see without this filter so I can see that it's 40% off sessions are direct sessions, whereas if I don't use this filter, I can see that it sells mid is only 18% azi. You can see it basically doubles our direct traffic. So please do this exercise. Also in your case, it will change your perspective on perfect sources. 15. Landing page and keyword: again very popular report. Landing pages are the first pages off the session, and landing page is very often the place where you have the Onley chance to get users engaged enough to continue to the next page. That's the reason why it's important to understand how to look into data in this report. I'm stressing out that thinking in context is absolutely critical in this report. A simple example. Why is it so common approach is to soar the pages by the bounds right from the highest to the lowest to find the worst performing landing pages? The approach is absolutely correct, but it can happen that you'll find there, for example, contact pages. Is it time for the panic? I wouldn't say so. What is the purpose of the contact pages? Yes, to provide the information about opening hours, address e mails and so on. If visitors find this information, it's completely okay that they close the page. We have a bounce, but it's all right in this case. So the point is, always look on what is the purpose of the page or set off pages and then check the metrics . If we use keyword, a secondary dimension. This is the place where magic can happen. Alright, guys, we're going to show three more tapes. Ah, in a landing pages report eso First we have to get there, which is in behavior, Tap side content and London Bages. Ah, as mentioned landing pages are the first pages in every decision and this is most of the time, The only place where you have a chance to engage your users enough so they continue. So this is why they definitely deserve the separated report. Let's have a look on it. And the the first step in this section, which is, uh, tip number 31 is to check your worst performing landing pages and again is mentioned. Um, thinking in context is absolutely critical here because it will happen that will will find , let's say, contact pages with a very high bound trade or some pages for which high bounce rate is is quiet, normal and organic. So we don't have to panic. Anyway, let's do this quick exercise. So let me minimize the side navigation and let's sort this, uh, report by bounce rate and then we're not going to do there. The full sore type of the waited one so we don't have to filter it is report by some real a number of sessions. So this is what we have now we have, ah, waited sore type by bounce rate for landing pages. So Ah, the goal of this technique is to find out the first performing landing pages, which we can see are the top one. Is Google redesigned saw by brand YouTube. This basically tells us that 60% off sessions that start the session on this particular landing page leaves immediately after seeing all of this one page, which might be OK, but I wouldn't expect if it's a category page to have 60% bounce. Right? So the first thing good to do here ISS to use some secondary dimension here. What I can recommend you to do ISAT first, filter it, which you can do either by filtering it here in this filter field or just by clicking on the name of it. And it'll show us on Lee the one Ah, now we can play with it. So the first thing I would definitely do is to use second, the rhythm engine device category might happen that for some reason. Ah, this particular page can be broken on on on a mobile, which we would see here. And actually it's quite an opposite because a Z weaken see ya. The mobile experience is actually better than then that stuff, which is quite surprised for me. So Ah, as we as we saw in one of the previous lessons about the mobile experience, we saw that it's now that good in in case of merchandise store, so that if our hypotheses waas that it's caused by some different device category, it's not ah, correct hypothesis. So Ah, we can cancel this secondary damn engine and definitely the I think worth doing for every landing page, which you will find out that it's not working as you supposed it. To work is to use secondary dimension with some traffic source data. I would definitely recommend a source medium, which is a tip number 32. So let's do that. Ah, the goal now is to find out whether it's a problem off some particular traffic source or it's a problem in general off over landing page, which this is the report consideration that would help us to to understand where the problem can be. And, as you can see, the average boundaries 58%. And as we can see, it's it's pretty much the same for all the top traffic sources, which we have the double bonus YouTube referral than Google organic. And it's not differing that much again. It's not a is that a problem of particulate traffic source. I can imagine that you will find significant differences in this metric if you will do it for your later. So please do. That is very simple and easy technique to do. You might find that some sources just doesn't work with this landing page on and now have ah lost step in this in this lesson adjusted number 33 we will use another perfect source, them engine and it's keyword, and this will show us off relevant Our lending. Beijing is throop articular keywords because those two dimensions are are linked together on DNA. Now take it from user experience. Perspective User is searching for something. Ah, he or she sees the results from search engine and ah, users are clicking on the results and they expect to have something relevant to the search term or search Curie day they used on our sides are supposed to have the relevant content for them. And if these two things doesn't work, we can expect that we will have a very high bounce rate. And this is exactly what we're doing right now. Ah, maybe it will get back to to the previous combination for source medium and the lending bait. So, uh, there's one more thing I want to show you. I forgot to mention that it definitely makes sense to filter here on Lee bait traffic if you are running. Ah, a lot off paid campaigns in, Let's say Google search engine or some display at it's definitely worth filtering here on Lee bait channels because you're just paying for this traffic. So we wanted to work well and not having high bouncer because you're basically wasting your money with that. So please filter it with advanced filtering and source meeting for Onley containing bacon veins so it can be a CPC, uh, because here and and, ah, feel free to work with it. So ah Google CPC IHS. I'm not sure if it's ah if it's a major traffic source, which you can see is here but but not for this particular landing page. So Ah, yet if you do it in your case, you might find that you have significant amount off. Landing babies, for which you pay for the traffic on is not performing very well, So this is definitely something worth thinking because it's basically a double loss. You're losing the traffic, and you have to pay for it. So yet, dude, for for source medium and do exactly the same thing for a keyword. This is something we were there, and and you can see off how relevant your lending page is to articular keywords. Whether it works, it's not working or or it's somewhere in the middle on DA. This is exactly where bounce rate can again help us. So if we if we sword it by brown trade and we use where did ever it, which we have Ah, this is the place where we're going to see off which particulary cures work and which new, not Google does not run a lot off paid campaigns. So most of the traffic has not said in a Cuba them engine, which we can feel there if you want to. So you can exclude keyword containing not sad for not, uh, click on apply. And, ah, this is what we have, right? So we can now do the very simple exercise and check how good or how bad our cures perform with particular landing page. So please do these exercises we can see in case of Google. If someone types YouTube T shirt and clicks on bait campaign and they are redirected to this particle or page, 83% of them are leaving without going to next page. So, um yeah, as I said, it's a double loss because you're losing a perfect and you have to pay for it. So please knew this exercise and fix as much as possible. You might see a miracles to happen. 16. Keywords: keywords are queries. User stopping to search engine through which they visit your site. Quite simple, right? Let's divide keywords into two groups. Bait and organic. The first ones are easier to get in Google analytics, and we'll show how in couple of moments, whereas organic ones are mostly not provided they used to wear a couple of years ago. But not anymore is you to Google's privacy policy. In general, keywords give us great information about what user search before they visit the website. Just check what they searched for and whether you have relevant content for such a query. If not, that's a nice example off low hanging fruit, as mentioned in previous chapter, If someone is looking for a camera and you're trying to sell him a fridge, it's probably not going to be a long term relationship. The first thing we're going to show before we jump into the interface is what's the difference between keywords and search? Curis. It happens quite often that Ah, a lot of people think that it's the same thing and it's not and there's actually quite a difference. So let me explain. Ah, what it actually means both off them are two separate them engines in Google Analytics. So let's assume, uh, this scenario, uh, search. Curious are, for example, klepto back blue back Large back schoolbag backs for kids. As you can see, all of them are somehow related to back on. This is what actually is a key word cured is something on which on what we bit. So we want to bid on a word keyword in Google AdWords And these are all the curies that are somehow related to it. So we are our Our ads are phone on this particular keywords. So keywords iss something, I would say Ah, an ideal search Curie. So anyway, keyword is basically a group off. Multiple curious, somehow related to the skewered and search queries are the real text curious that the user stopped into surgery and into search engine that actually triggered an impression for particular cured in this case back. I can imagine that a lot of times Ah, search curie can be more valuable information and the key work. So this is why it's important to know that existed there and was the difference. Ah, and please remember the only information from from this'll I that there are two separated them engines search curies and keywords, and we're going to show how to use them right now. So we're going to tell something about the keywords. So first of all, we have to go to the report where we can find something by the keywords, and it's an acquisition tap. Then campaigns. And then we have two taps paid and organic keywords. First of all, I will show you what organic keywords are. As you probably know, uh, organic are keywords that users typed into search engine and they didn't click on Ah, bait. Ah, at. So I expect that the your report will look very similar to this one. So more than 95% off sessions will have a keyword not provided, which is something we described. And it's due to Google's privacy policy not showing this keyword. So, as you can see, it's not very interesting, and the data won't tell a smart here. So let's maybe just keep directly to paid keyword Ah, 50 words, which are more Mawr relevant and interesting. Ah, as we already know, what is the difference between keywords and search? Fury will try their combination, but let's start with the first thing on its ah tip number 34. Do the very simple thing which is just opening this report, Ah, showing some relevant number of lines here and go word by word and try to check whether you have a relevant content for such a key words. Because this is what users are typing and they expect to have relevant content there on. If they don't, they will just leave. Ah, good thing to remember here is that if you're in a paid Cuba's report, you have to pay for the cure. So we want them to work well on the first. The hand or delete for you will be if something does not work, will be a high bound freight, Uh, which you can see the first a lot for, uh, particular keywords, and that I'm pretty sure there will be strong correlation between high bounds, right and local burden. Right? Which makes sense, right? Ondas We can see even for very, very similar keywords were merchandise years and and blast the will plus merchandise. It's quite a difference. And as you can see, even a bounce rate is different here. So the suggested confirmation this correlation works here, so please do do it and check whether you have relevant content for them or not. If not, try to create it. This is the easiest thing you can do on go through. I don't know how much bake campaigns you running. So this is what this is. Determinant off. How many keywords you will have here? Ah, And now, second thing, search Curie, Let's go there. And that's we know these are the real worst. That user stopped into search search engines. This'd something I would. Same or reader friendly, as we can see. So in case of Google merchandise store is almost every time something related to Google Merchant I start right. But if we were if we, uh, and get exactly the same thing using mainly for creating ah content on way scroll down, we may find something. Maybe Sometimes it's quite funny to see what people are searching for. So if we go down Ah, let's see, for example, I don't know. Here are Google products School to teens rail form sticker from Google. Ah, check us out on Google Sticker doesn't have an online store, so you can see it. It's Ah, Why Quite a difference between the keywords and some of them are interesting. Anyway, some some huge business impact from them. So please check it. And these are actually I would same or valuable. Ah, this more more valuable data than than in case of cure. So please use it and let me let me show you one more one more. Think on this is a tip number 35 which is, ah, relation between cured and search curing as what is correct on ah slide a couple of moments ago. Keywords are basically something on. What we are targeting and search crews is something that is triggering the impression for for a cure. So let's do here a simple combination off key word. Sorry. Okay. Word and search Curie and ah, tip number 35. Very simple thing to do. Ah, it can be a lot off a lot of search curies under one key work in one bucket and there on relevant for a user can differ. So if we, for example, try to ah, I don't know killed her it for, let's say, Google store and I'll do the exact match. So include key were exactly matching and go store. Ah, uh, and we can see that there are three off off search Curie. So let me maybe just use something something slightly more different. Ah, And for example, let's use Google merchandise. Yeah, on. We can also see we don't have much data about any where we can see that that bounce rate differs between these search queries which are under one keyword. So so do you do you might be surprised off what everything you are paying for for being You're having your your, uh, available on do this exercise. So you're paying for it. So you won't only this search curious and cured combination that really works and unvetted . And we're going to tip number 36 which also helps us Teoh create ah, list off Negative. Q Bert, what did this is that these are actually keyword on which you don't want your had to be shown because for some reason, you probably will will never have your business for some particular curious. And now you're paying for them, so it doesn't make sense to do pay for them once it's not your business. So do exactly the same thing. Just try to find out the worst performing search curies, and you have two options. You can either create content for it and have your business related to it. That this is, I would say about her case, which is, if it's somehow related your business of creating the content for it. Or if not, just use it as a negative keywords in your in your at settings because it's just wait wasting of money if it's not related to your business. So again, a very simple thing. Do this exercise. Combine keywords and search curie and see what works and what does not work. 17. How to evaluate traffic sources: There are plenty of ways how to evaluate traffic sources. Let's start from the very beginning. At first, always ask what is the purpose of traffic source? I'm going to evaluate. Not all channels are supposed to directly sell something or bring the lead. Their goal might be, for example, just to bring users on the website for the first time. Well, look on couple off scenarios on how to look on the traffic data. It will be mostly about the various dimensions combination, which will help us to identify the possible problems or possibilities for improvement. So, uh, traffic sources evaluation can be quite broad concept. Anyway, In general, we should somehow try to find out what channels works better. Which do not. Ah, but there are a couple off rules we should always follow when we regulate something. So at first let's show where we can find all five main traffic them engines, which are source medium campaign at content and keyword. Ah, we know that the most used one source media, which we find all traffic and then source medium, uh, which is the report we are right now. Then there is dedicated report for all campaigns which is in campaign step all campaigns. And here we can also, in a campaign step in the sub tab. In acquisition, we can find faith keywords, which is which hard Williams from the the keyword them engine. If we want to switch it to add content, we have to click on other. If you want to see this primary dimension on, we just filter in here at ground. It's always under acquisition them engines, so clicking on add content. We now know how to find all five them engines, and now we're going to evaluation. So let's go back to old traffic and source medium, uh, and you may find that sometimes you will want you off or somehow will want you to evaluate traffic sources, and that will be all someone will tell you eso you might start to look into data, and you will not know what to look for. The first thing, which is a tip Number 37 please it. Every time you're going to evaluate something, ask yourself a simple question. What its purpose off traffic source or a campaign or any any traffic dimension. This because this is something that will lead you to find out the ride metrics. Right? Because not all of the channels have the single. Some of them are supposed to bring the new users to Web site. Some of them are supposed to converge all users. And this is this is exactly the question we should ask every time we want to evaluate something. So, for example, it doesn't make sense to compare every channel with every child because it just doesn't work this way. Uh, maybe go to two. Ah, Google CPC, and I will show you what exactly do I mean? So we filtered Onley, Google, CPC and I will use campaign as a secondary damn engine And here we are. We can see that there are about 30 campaigns on you want to evaluate them? Which one works good and which one do not. So maybe the first thing you will do would be to look on the conversion rate and you would say OK, like the 1st 2 of them differently did not work at all. Ah, and you would say the 3rd 1 and the 5th 1 are probably the best one, right? But we do not compare what is comparable and this is this is deep number of 38 thes two are really stick together. What is purpose off traffic source and and we want to compare what is comparable. As we can see, we have two main types of campaigns here. Some of them are search campaigns, and some of them are display campaigns, which is which is a huge difference. Right, because if someone is searching for something, they already have their their intentions. So they want something instead of this play campaigns. Most of the users are probably probably not interested, and we want them to do something. So that's that. That's a difference, right? So if we would would take the same metric for valuation. For both of these campaigns, who would say the display campaigns, they just doesn't work, which is not true at all. So, uh, we have to use more than one metric. Please try to avoid a situation when you will take one metric. For all of the campaigns, it doesn't make sense. So the good starting point is always bond trade, depending on what is the purpose of the campaign if it's a search campaign, so the users are searching for something. Ah, and you, you can bring them to your side and they bound. That's probably not a good think. Where is in case off this play campaigns when most of the time they are supposed to to start the relationship between user and Europe cider product? I wouldn't say hi bound for it is extremely bad thing here or other local Burton, Right? So, thesis what, actually is the different between between evaluating the campaign or traffic sources? Always ask yourself question. What is purpose off this campaign? And based on, did it tried to find a proper metrics for it on these are actually two very simple rules which, if you will follow, will always help you, Dio ah, not to get mislead by the data. So please, please use it on. And we're going to to tip number 39 which is something that will enhance your perspective on the data on its about linking Edwards and Google analytics to have all of the metrics on on one place. Because even though we are comparing the comparable, so let's a, uh, a campaign number three, which is dynamic search. And then let's say number five, we can say that both of them have very similar conversion rate, right? 1.99 in 2.1 It's to another full information. And what will happen once we link analytics and Edwards that we will see one were important metric here, which are costs. So let me show you how to do that. And then we'll go back to the report where we will show the full full later. So we have to go to Adam and Section. And then we are on the property level where ISS product linking part on the 1st 1 is Edwards drinking. We have to click on it, and we have to have Edmund Reid's in both Edwards Account and Google Analytics account. So I don't have ah permissions to do that in merchandise store accounts, which is obvious. But anyway, in your in your account, you should be able to do that. So you just click there you will see the possibilities with which you can choose from from Edwards account we have access to and then you just laying them. Then you were the wait for a couple of hours and you should see all Edwards data in Google Analytics, so you will have full information. So please do that. It's Ah, As I said, it enhances your view on the data, so you have full picture. So this will step number 39. And now let's go and see how it looks. So we're going to acquisition and the Edwards has separated or dedicated tap, which if we click on it, you can see that it's quite rich. We have a lot off a lot of times here eso let's go into 1st 1 which are campaigns. And once we're here, we have the full information. And if you remember, we were comparing dynamic search ATS with Edwards accessories, which had very similar conversion rate. Ah, and, uh, the information we were missing Waas Waas Ah cost. Now we have it. So we finally have the full information. We can say which one works better because they're supposed to be both the same. The goal of those campaigns is same. Teoh, bring the user who is searching something to upset and convert him. So now we have full information so we can say which one works better and which not eso this is in general, something you should always try to do have the full information about any source. You have tried to understand what is its purpose and then try to make some analysts based on it. Ah, so this is it. And maybe one more thing, which is quite similar that we did when we were talking about keywords. Now we also have here search Curis, which, as we know, it's probably more valuable information, comparing two key words only on we can see here that which keywords were paying for and how much was ever CBC and how it can words done. And this is a very nice example off if you remember that one of the tips was about creating a list off negative keywords, which is something we or negative search curies to be precise, which are curious for which we don't want. Teoh have our at shown because for some reason, we either don't don't have our business related to it or we don't want to target them. And this is very nice example to find out where, uh, where this key words are, which they are Onda. The example I'm talking about is the 3rd 1 which is a Google store. As we can see from search campaigns, we have every bonds for a 32% and Google store has 64.37. We can see it's doubled, so that's probably something that does not work. And it's confirmed by 0% conversion rate. Right? If we compare the 1st 21 which are 2.2 and 2.4, we wouldn't expect zero burst and conversion rate. And it's not in, Ah, we have, Ah, 334 sessions year. So it's not that we have a very low number of sessions, so let's try to find out what can be, Ah, the reason for it. So if we Google the Google store, what probably users are expecting is store that google dot com, which is a completely different website, right? It's about abs you can download to your to your mobile phone from Google Story for on Android. So it probably doesn't make much sense to have campaign for a day ski word because sorry, search Curie. Because if the parenting doesn't work, Uh so, yeah, do you get again? Very simple exercise. You may save money or try to find more element. Search cures eso this Waas acquisition and campaign evaluation. One more thing A So you can see we have no annotations here and with our invitation. This is tip number 40. Try to create them and I can recommend it to create as many of them. It's possible. What are invitations? They are small information again right there. That will help you to evaluate something in the past. For example, you had some significant campaign I don't know. Just for 23 days, which will do a peak in your later. And if you will look into data half a year after it, you will not know what exactly what's So we will try defining Inspect. There was some spam or some robot or something else. But if you will have a notation there written that I don't know something happened. Ah, you know, I'm just right writing some damage back there, and it will save it with, uh, date. You will have this information in G A. And trust me sometimes it will save you hours off analytics or hours off wasted time trying to find out what what happened. So please use it. It's very nice thing to do and to have it. So yeah, that wasn't Guess that was. Ah, traffic sources evaluation. Remember, the three rules were mentioned. Ah, it'll save you time and you will make my decision. 18. Goal setup and funnels: one of the fundamental things in Google analytics goals. In general. Goal is every action we define, and based on it, we evaluate websites performance. It can be literally anything, but it makes sense to have meaningful goals, such as transaction correctly, field and sent contact form downloaded price list, new registration fully read article new let their subscription in Salon. We're going to show two ways how to set up a go. Both of them allow us to evaluate conversion rate and the 2nd 1 will help us to find the bottleneck in the funnel the place where we probably lose potential customers. There are four goal types in Google Analytics destination. This is the most common one. If we set up a go, we define a so called thank you page about which we know is displayed on Lee to user when he or she completes ago. So, for example, users purchased something or send a contact form. What we need to know is the URL address off such a page. So if you are an e commerce or early generation business, this is the type of girl you're probably going to half duration the simple one you basically defined time, after which you want a goal to be completed, since we know how inaccurate it is to measure time in G. A. I do know the recommend using it very often. Bages Procession The straightforward one just defined how many page views procession you want to consider as go completion. I can imagine that in cases off purely content websites such as blocks on my newspaper or brand Onley pages, this can be a meaningful goal. And the Final one event. This topic wasn't covered during the course because it's slightly more advanced. But just to give you a hint, sometimes it's not that easy to measure ago by Earl. A good example are quite popular. Pop up windows with simple forms in which you feel in a couple of lines and send it after successfully sending form. The pop up window disappears and there is no specific thank you page that will tell us, Hey, the formal successfully sent you should measure ago. This is exactly a case when events come very handy except page views. It's also possible to send events into G A. Just to enrich the basic measurement by such an event, we can say that even if there is not a dedicated thank you page, we know that something important in this case sending form happened. Following Lee, we can set up a goal based on this event. Let's assume a simple scenario and set it up. Okay, Now we're going to tell something about the goals. Has mentioned girls are something fundamental in every Google analytics accountant. Never business in general on I can recommend you to have leased to goes or two types of goals. First of them are primary, which are your KP eyes for for our business. And then you should have a secondary goal full, for example, not something based on which you evaluate your business, was it? But it's kind of a helpful metric. Whether you're growing or not, it can be, for example, if you are each hop, newsletter, subscriptions or new registrations can be a secondary goal. So something ah, not major, but but nice to have that helps you to evaluate your business. Well, you're going to show in this in this lesson are three things. First of them is how to set up Go. If you want to do that, we have to go to administration section. So let's go there and I will switch into my account because you need and, uh, admin rights to to do that. So if you want to set up the goal, you have to do it for every view. So if you if you're using multiple views, you have to set it up. Every Yucel anyway, we're going to click on the goals on. There are a couple of possibilities. When we're setting up Go, we're going to show the easiest one, which is gold based on a thank you page. We already know what it is. It's a page that is shown to use their Onley when go is completed, so we know once is displayed. It was so let's go and try to set it up. Click on a new goal on. We're going to see a couple of pretty fine templates by bike. Go. There's a lot of them, Andi only think it does if you if you select one of them, is just that it will name your go by one of the templates. Nothing else. So what? I recommend it to do its always quick and custom click continue on then we're going. Teoh. Name it at first. So let's assume we're going to have a very simple goal, which is an issue of order. Ah, so let's call this go border. Uh and it will be a destination, which means that the thank you page is displayed. So this is what we're going to select. Theun, there are three more duration and Bages position. Feel free to use it. I don't use it very often. I think that I only did it twice Urine being analytics in the past couple of years s older , selecting destination. And, uh, this is the set up where we have to do. Uh, first of all, we we have to choose a quantify air eso We will assume we have Bernas. You are dresses on and let's say we will have order Beat the God html so we will know what it wants. This page is displayed. It was a conversion on If I would click now safe from this moment, it would start to count the conversions. So it's very easy to set up this girl. I will not click save now because we're going to show something slightly more advanced anyway if we would set up a goal this way, we will start to see conversion rates for it. Thea Absolute number of conversions we had. So if we would go, for example to Ultra Figgins source media report, which is something we're now familiar with And we would scroll to the very right which is always Ah, Convergence set off metrics from this woman we would be able to select also In you go it created I now have only one here and if we would created, we will see another goal so we could select it and then see conversion rates and the absolute number of conversions for for ago. So this is another simple thing to do. Ah, if by any reason you don't have any goal in your Google analytics, they should be one of the first things you you will do in your account. Because if you have no goals atomic, you probably cannot evaluate anything. So please do that as one of the first thing. If if you don't have it. So this was the first thing we wanted to show. And now we're going to tip number 41 which is advanced. Go set up and ah, don't worry. By advanced, it doesn't mean that we will have to call something. We just have to tell slightly more information before goal happened. So what would you were again? Going to admin section against goals and we're going to create in you go. So again, Gustin, goal. Continue Order completed the destination and we could continue on. We know that in a previous case we know that our thinking pages ardor it did not HTML So this is exactly the case we had in a couple of moments ago. And what we're going to do show and set up is now funnel. And what this almost every website has couple off off stabs during which users I have to go once they convert. And the typical case is check a political process depending on your website. You have probably some something between three and five steps there, and I'm pretty sure that if you don't have ah you eggs guy on your team or someone who is who is very familiar with user experience, a tely east One of these steps were a slightly more problematic in terms off that you will lose a lot of potential customers. There just happens on DWI will show how to find out whether you have such a step or not on . And the only thing we have to do right now is to tell the previous steps to to Google analytics. So we will not only see the final number off conversions we had, but also where is our bottom, like in a funnel. So let's assume we will have three more steps there, and at first we have to name them. So we're assuming we're in a tickle process and we say that let's say the first page will be card. Ah, 2nd 1 will be breath and the third Juan will be, Let's say remember? And now, exactly as it is In a case of thank you Page, we have to tell Google analytics, What are you Earl arrested? So let's assume we have nice. You are else. So it's car html, then breath html then Damon html. So, uh, there's one button, uh named required What it means if we click. Yes, it would only count the conversions that have to go through all of this page is so it won't be possible to enter funnel from the second step or the third step, which I leave it up to you. If it's possible in case of your website to enter the conversion funnel in one of the ah, I would say later, steps off off process We're defining right now, so I leave it up to you. Let's assume that in case off this girl, it's necessary or it's not possible to enter the check out without skipping or it's impossible to skip one of the steps. So this is what it does. Ah, so if we would save it now, we will see again, as in a previous case, the absolute numbers, but also funnel eso if we would click save Now we start to see also a funnel visualization . So this was set up. It was a tip number 41. And now we're going to show where to find the funnel visualization report and how to read. This is this is number 42 so let's see where it is. We have to go to conversions over. At first, I have to switch again to merchandise store. So here we are on, we have to go to conversions. Two goals and Then there is a dedicated top funnel visualization. So let's go there and see what we have. First of all, we have to select the goal, which has the final steps. Eso in case of merchandise store its purchase completed. Uh, committee, my step. And, uh, what we see is the funnel here, right? So this is the type of report you will see once you set up a funnel, which is amazing. And it tells me how many people am I losing in every step off the funnel, which is truly amazing with this report? You This is really a low hanging fruit. So once you once you fix this steps, when you will find out that they're losing in one of the steps, customers can fix it. And most of the time are those fixes are quite simple and easy to do. You just don't know about that. And this is the report. It will tell you. Hey, you've got a problem in step number three, for example. So the first thing t check as your funnel commercial right, which is always the number on the top of the funnel in case of merchandise story 27.13% which is not a lot. I would say that the that a good number is if you will get someone to 50%. This is something definitely every every business should be able to reach on. I'm not saying that you can't can't go to 70. You can. I've seen a couple of websites having 70% final conversion rate in way. Anything be below 50%. There's a lot of space for optimization. So now let's show you how to read this report. I think it's quite straightforward. Anyway, let's look on the left and right sides off it. So it does us from which pages people came to funnel step and then once they laughed, it tells me why and how they left. So as we can see in in case of the first step, which is card, we can see that only 37% off people who enter the card made it to billing and shipping on that. There could be many reasons for that. Why they didn't leave, why they didn't continue. And this is something you're ux guys should do. So So this is the report. It will definitely love because you're They don't have to go through all of your website trying to figure out where the problem can be. Your basically telling them, Hey, cart page is my problem. He's trying to fix it, try to find out where the problem is and we know that the from the 33,000 people lad there , 9000 exited website, which means they clicked on ah on the red button and they they stopped the session. Theun 6000 went to sign in page five points having to basket and 2000 to store on. We see exactly the same thing for every step, right? So we can see that from from Ah, 2700 people, half of them left without seeing any baby then and not coming back and completing ago. And exactly in this report we can go step by step and trying to find out where the bottleneck kiss, which we can see it is definitely the first step. Now we don't know what exactly the problem is, but we know where the problem is. So this is what this this reportable tenet of this date I would just I would say amazing. So I heavily recommended to set up this type of gold because if you would have only the first type of it that you would say, OK, I only want to measure the final step, which is 14,000. You would say OK, this month A Have I had, Ah, 14,000 conversions and previous one waas I know 17. But if you if you see that in quite quite ah final part of decision process you had 53,000 people and only 27% of them completely the purchase. You just know that there's a lot of money on the table. So this is the report. It will tell you where to start with optimization Where is there a low hanging fruit? So please do that important thing to know all the data you set up for Ah, for a goal will start only from the the U set up there, the go into the future So we don't have this data historically. Anyway, there is if you're if you're somehow we're interested in having this data immediately. You can do that because as you saw, this gold top is based on sequence off ages and this is something we historically have if your measurement wasn't broken. So if you want to created the historically is actually quite easy because the only thing you have to do is go to behavior report and side content. And this is guys. Tip Number 43 which is creating funnel historically and on Lee. Think you have to do is to filter the page you're looking you're looking for. So, for example, let's assume it's a basket base, so we will filter basket html on, Uh, then we will ride down somewhere. Number off page views. So this is what we are looking for because basically number of unique page views is the number off sessions that viewed a articular page. This is exactly what we are looking for. So if we know that the first pages let's basket, then second page. If that's a four, we will write down again the unique pages for a store. Then I think there's something with Order Order completely. So if we go exactly step by step right down the unique page views we can created historically to have the same data, and then we have to find out where the bottleneck case until your young's guy. This is This is the place where we should start with optimization. So these were the three tips. When it comes to go set up, please do all three of them especially funnel visualization and then try to recreate his data historically by ah, writing down unique bedrooms for every step off your check out process or card or if you're in the lead generation business, the final steps off your your contact form or whatever you have there as a final conversion step. So this waas gold set up. 19. Product analytics and enhanced e-commerce: If you're on an e shop, Product analytics itself can be done. Also in G a report called Product Performance. We're going to show a couple of examples which can help you increase your sales possibilities off product analytics are changing in the moment you have implemented enhanced e commerce mogul. Even though you don't have it, we will show a couple of simple tips of how to sew more. No working. Teoh. Look off about what can we do with product data we have in Google Analytics? So we have to go to conversion step e commerce and product performance. Ah, I would divide this lesson into two parts. First of them will be for those of you who have not implemented enhanced e commerce yet on the second part will be with ah, slightly more data. We have a very valuable data. So So that first tip number 44 please check World are your most popular products when it comes to either quantity or or revenue or any metric you have there. Ah, if you don't have implemented intensity commerce, you will see only data. Ah, to average press, you won't see sort of to average quantity you won't see product refund amount or Carter decelerate. And by two, Nikola. I don't worry about them right now. We'll explain them in a couple of minutes. Anyway, uh, what to do with it? Ah, if you find out that for some reason you have in your top five no. 10 2050 products, depending on how many of them you have some products that you don't promote that much or you didn't even expect them to be s sold Much very easy thing to do. Just try to promote them, or maybe on your home page in your on as one of the first product in your categories, Or try to put them in newsletter if you find out for some reason that people don't buy them just occasionally by the regularly and they are your top products. So this is something the dealer again, users will tell you. Hey, this is the products will love. I should promote it because we like it so very easy thing to do. Just check if if you don't have some of the products that are on your I know 27th page in category and people are buying it anyway so very easy thing to do. So please check what are your most popular product on we have here to metrics that might seem similar. Any purchases and quantity about the island. As you can see, the quantity ISS more than three times higher than unit purchases. And the difference is very simple. If you buy, for example, three NASDA Hello Door Bell USA. In one transaction, it will be counted three times in quantity, but only once in Burgesses on what basically put the unique purchases mean is the number of transactions in which product was salt. So it really doesn't matter whether he bought one or 10 of them will be still counted. One. So it tells you how many people pretty much molded. So that's the difference between it. And again. If we who were on a question mark there is an explanation before today according exactly, doesn't mean so do This is easy. Think, then we're going to tip number 45 on. It's about average quantity, which is this metric on. Please check it and do again. Very simple thing. If you find out that for some reason people are buying, let's say two or three pieces off, some off some product naturally tried Teoh on your product detail off this of this product , Tricon naturally have their pre selected Ah, the number off off off pieces as you find in here. So if, for example, we know that perfect small CEO Wide World Alarm s sold 2.5 times, let's try to offer them by default two or three pieces and then just one. And you might be surprised of how many people will start to bite in higher and higher quantity. This is things added multiple times that almost every time it started to work. So people started to buy naturally more, more pieces off some product because they just buy it. I I didn't know the reason it I'm not thinking work every time, but this is very easy thing to do, and it'll increase your average numbers quantity without doing something extra. So this is something you can do with your currently users base, and it will not cost you anything. So these are two tips if you only have the basic data in Google Analytics and now we're going to show what ah, huge game changer enhanced e commerce module s. So, uh, let's have a first look on. Ah, bit off theory behind it. And then we're going to see what metrics we have there. Okay, A little bit off theory before we'll go back to interface. What in his e commerce is about is pretty much sad, and its name will enhance our view on how people behave when they are actually buying something. Let's start maybe with a purchase. This is often something we only care about when it comes to product performance. We want to know how many times particular product Vosper chased, maybe in quantity. And this is it. And I will say it's it's very similar to gold set up with Shaun Wright. We only see the final touch point off on, and if we start to think about it slightly more differently, we'll find out that there are many more touch points before per chase can happen on. This is what enhanced e commerce will show us something before purchase with product on. We're starting with product least few thesis. This is the first place where you can interact with product. It means that you can see it in ah category list when there are many of them. But this is the first place where you can interact with product. Then you can see a product detail, which is second that point. Theun. If you want to buy it, you have to add it to cart, right? This is this is third stuff want. Then there is another one. You can remove it from cart. Uh, then you have to go to check out with it, which is another type point, and then you can purchases. So as you can see, there are five more touch points before a purchase can occur on. We shouldn't ignore them. And this is exactly what in his e commerce is about telling us, the whole product funnel not only the last touch point, and now we're going back to to the interface to see ah, where all these metrics are. So we're making the interface and let's have a look on in his e commerce metrics. We already know what's the logic behind all of them. Ah, but maybe let's, um, at first go back to merchandise store and repeat the stages are the status is that product can happen. So we're starting with a product list feel. So this is this is exactly it. Every time a product is viewed it any an analyst, it can be either category or or it can be one of the product in the recommended products under detailed is quite common. Ah, this is actually what product beetle of you mean. So this is the first touch point. 2nd 1 ISS, once we click on it and we are here on a product detail view. So So this is the moment when we sand information that the product was fueled on the name of this such point is probably told you. So this is it. Next touch point in a product funnel is adding it to card. So, um, this is it. We can see in couple of seconds. Yeah, that it was added, as you can see here and then we can remove it from card, which is here by clicking on this on this I come So let me do that. Yeah, we are informed of the protocols removed, So this is another touch point on. And if we go to basket like licking here and from Boston going to check out, this is the moment from this now on, we would send information that the product was also in a check out. So these were the five steps prior to purchase, So it was just a quick, quick repetition. And now we're going to data. So we're begging Google Analytics on, and we are in the same record as's previous, which are ISS convergence, e commerce and product performance. And the only thing we will change right now is the vertical tap. Eso we're going from summary to helping behavior will minimize the tap on Dhere. We are thistles, exactly the place where we have this amazing metrics. And this is good Step number 46. And if you're running an e shop or you're selling anything on the web side, this is really a must have think if you do not have it, you're being kind of Ah, they would say blind product analytics because you're only watching the last step and ignoring the previous ones, which is five of them. So let me show you what they are. We have for every product, we can have a separated funnel from the very first product this view to the unique branch A So, uh, this Is it productive use than product detail views, product ads to cards removed from cards, Perak checkouts and unique purchases. Thes 1st 6 metrics are absolute, and then we have two calculated ones. Carter. These elaborate and bite Anita rate the car to the threat is actually ratio between adds to cart and productive views. So this is why it can also be called Ah, it's to CART rate. And then there is a bite obliterate, which is the ratio between unique purchases and product Italian. So this can also be called US product conversion rate. So amazing metrics, we should try to have these metrics this tumor tricks as close as possible to to each other a Z we can see in case off merchandise store on Lee. What? It tells us that it's only 1/3 off. Products that are added to card are also bought, which I will didn't say it is a huge number. So Ah, we're nice. Were nice starting point. Anyway, let's have a look on the data and in ah, on the top funnel eso we're starting between product list on product details when I would suggest you to do is to explore these data and create a couple more calculating metrics. I already knew how exported right clicking, scrolling on the very top, clicking on export and then exporting either directly to Google sheets or Axler C s. Well, it's entirely up to you on the first couple. It in metric I would create is the ratio between Douglas views and practicalities. Uh, this will tell me how the truck they particular product is from a list you. So? So how often do people click on its detail? S O. This is the first calculated metrics then the 2nd 1 is product detail views and it's two cards. We already have it because this is a cartoon reiterate. So we should try to have this number as a huge is or as high as possible, then product removed from card. What I do often is that I compare removes from cars to add to cart because once there is a removed there has to be also at. So I compare those numbers on. Once I see something significantly Hi, I'm trying to find out why it can be there. They're probably thousands of reasons, but quite common are that there's some extra shipping that users have to pay for it, or some some extra costs in general. Ah, tied to a product in, ah, a cart, then for luck checkouts on unique purchases. So So instead of only seeing unit purchases, we now have a set off five grade metrics. Ah, another good thing to check. If you want to find out whether you have it implemented, you should see some numbers different from zero Here. Here, if you only see zeros in cartoon iterated invited is the red You do not have this mogul implemented in your analytics set up. So if you don't have it, blessed, ask your implementation guide to do that. It's It's as I said, it's a must if you're running an e shop and now back to analytics. Uh, what is great thing to do is, uh, to first start with ratios between car to reiterate and bite Anita, right? Ah, as we can see, the something that bobs up to To my eye is line number six, Google T White, where I can see that there's their squad. Nice car to declare it 46% almost, but only 7.6 is biting either right, Which is how much? 18? Yeah. 18 of the products. Only 18 of the products that are added to card are also bought. So this is definitely the first product I would I would inspect. And if I look on the very left, I can see that from the top 10 products we have here. This one is mostly viewed in a product list. So I assume that Google is a heavily promoting in somewhere. But I can see that is the only how much is it? Ah, 3000 divided by 60,000 yet 10%. So only 5% ctr bit from from Frankly, you two probably tell you, I don't know if it's a lot or not comparing to others. But definitely, if I compare those two numbers, just there's something something wrong with the product. Next thing worth sticking. Even if you haven't implemented, it can happen that you don't have it implemented correctly, and unfortunately it happens quite quite often if I will show couple more lines year. The first healthy check is that by to detail, raid should always be lower number than a cartoon illiterate, right? So if we scroll down, maybe. Ah, yeah. Number line number 24. You go. Clear. Bena for Beck can see that it has higher by to decorate and cartoon eatery. Right? Which is which is Ah, nonsense. It The product has to be added to card before it can be purchased. Right, So there's probably something broken. So before you start with analytics, please do this healthy check whether the numbers you see actually makes sense. Ah, so, uh, yeah, this is This is something that gives you amazing opportunity for for every every product you can find whether there is a problem and it doesn't only have to be about the problems with your products, it can also happen that you will find that you have some product, Bigler Nice conversion rate, which is by two D threat and ngh terminology. But it will look to the left and they will find out that it's not promoted in terms of that , it's not frequent. Reviewed in Prague lists. It doesn't have a very primitive use, but maybe it makes sense to start it promoted. Once you see that it converse for nicely. Ah, or wise words. I will find out which is exactly aligned. Number number six. The product does not convert where well, and you're heavily promoted. So So you're just wasting your your promo places on your website by with Barack. That does not work for some reason, right? And, ah, this this set of my tricks won't tell you what exactly the problem with the product or opportunity, but it will tell you exactly where, because, as we know this are the product. Dutch points, which are basically that it's is exactly the product funnel so we can find a bottleneck for every product. Or it doesn't only have to be about one product. It can be about whole product category from click here in one of the vertical tips or product brand. So yeah, please do that. A Zay said. It's amazing set of metrics on if you if you are serious about product and let things. This is a master. Please implement it. If you do not have it, eso this waas this waas I would say Verne eyes beginning into product analytics. It will require a significant amount of time to play with it, but it's definitely worth to do that 20. Segmentation: we have been looking on Lee on aggregated data so far, it means that we had all traffic sources, device types, purchasers and non purchasers, new and returning users in one group for a lot off analysis. It's completely okay to use this view, but slightly advanced analysis will require such an approach. There's plenty of pretty fine segments in G A, but we're going to show how to build a segment from the scratch. G A allows us to create segment based on almost any dimension and metric and their combination, and this is a very powerful technique when it comes to advanced analytics. So we're going to tell something about the very powerful technique called Segmentation has mentioned s so far. We've been looking on all our users in one bucket, so there was no differentiation between the two countries or users who already bought something versus the ones who didn't and so on insulin. And now we're going to show how to do segmentation where a two first, we're going to show how to create segment from scratch, and this is tip number 47. If you want to find out what segment of users you have applied on your later. It's always shown here. As we can see, we have all users, which is 100% of our users on. If we want to create a sick segment, we have to click on this grey area telling us at segments. So let's click on it. And the Google analytics have by default decent amount off pretty fine segments. So, uh, feel free to scroll down how many of them you have there on. There is also a huge gallery off segments created by by Google contributor. So if you click on import from Gallery, you'll see Google Analytics Solution Gallery with a lot of segments, so you can then important without building them from scratch. So feel free to explore. Explore this option. Anyway, we're going to show I'm going to close it now how to build segment from scratch. So if you want to do that, we have to click on Ah, read about new segment. If I this one on, we should see something like this on, and this is the window where we can create it. There are many possibilities of doing that just by clicking somewhere. So if we want to create a segment on Lee off, for example, males, we click on it. And if we wait for ah for a couple of seconds, it'll show us how big these user group it's so we can see is 36% if you will also click on , the females will be probably something hiree affect 1% and the rest of our users are not recognized. Or the Google analytics doesn't know what gender you have. Because yes, we know it's based on, ah, behavioral data. And so we can, either going that by tap and trying to trying Teoh create a segment or there isn't advance section, which I wouldn't say advanced because actually easier Do create a segment there. So let's go into conditions on this is where we can create it. So we're going to show how to create a very simple segment off our mobile sessions or more while users. So we we can either filter users or efficiencies, which in this case, is the same. So let's assume we have sessions. We can either include or exclude something from segment, and now we have to choose from all the dimensions and metrics we have so as I said there could be literally tens off thousands off segments you can create here, which is a one hand grating on the other. It's now that grid, and we'll get to that in a couple of moments. So, uh, well, do you want to create is our mobile users, which we know that I mentioned for it. It's called Devise Category, so we want Theo. Include only sessions where device category and there are many 20 fires or options. We have device category that contains and it will click there and wait for a second. It will show us what options we have, so we want only more while and if we click outside this box, we will see how huge this group is. It it's, ah, slightly more than 1/4 of all our users on this is it. If we know will name the segment s well while Titian's and we click Save, That's it. Pretty straightforward. We have to wait now for a second until the segment is created. It applied, and as we can see right now on the very top, we can see that our segments changed from all users to Onley mobile sessions So if we will explore now any report we have in Google analytics, it will be on Lee four mobile sessions, which is script. We have just smaller a group of users based in some condition, which can be either dimension or metric. And now we can see an export data important thing to remember if you will compare more segments, which you can do up to five, I think, uh, please, always. Let's say we will have also a desktop here so you can see now that we have two segments here on, I will click on Apply and we are comparing segments. Please have a rule that if you are creating the segments and you comparing two segments, try to always have it contradictory in terms off that one user or one session can be Onley in one segment, not in both of them, because otherwise, ah, it would make sense to compare such a thing. So in case of these, we can see at that desk. Abusers and mobile sessions are contradictory, right, because if you're in desktop, you can be a mobile user at the same time. So a simple rule nothing. Nothing extra advanced but But please keep that in mind so that if you would, for example, compare I will remove desk abusers, mobile sessions and all users. For example, it wouldn't make much sense, right, Because all mobile sessions are also included in all users. So it doesn't mean my sense, right? This this comparison because these data are distorted. Ah, let me show. You will remove this segment and try to creating new one. Ah, that I'm clicking here. And if we are going back to conditions, we can use multiple conditions here. Not only that, we only filter or including sessions that mm Y scategory complains more while. But we can also use more conditions either or or and by which we can either widen or narrow the second, depending on what we want to find out and what segment we want to create. So it can be Onley users coming from mobile devices. And, uh, let's say they're Sydney iss will be half San Francisco. I'm just know clicking on the first option 1/2 right on. I can see that it just 0.92% of users. Anyway, we're going to tip number 48 which is which has two parts. I would say, first of all, feel free to play with it, play with it a lot. I use segmentation every day because because if you want to do slightly more advanced analytics, segmentation is a must. Without it, it's just do amount to do that Do a huge amount of data have when a lot off off things are hidden from you, So I will cancel it now. And what is important to remember you will find in the descriptions my my favorite segments I use, which, which makes sense to use for outside almost any business. But keep in mind that a ZAY said, um, the possibility that it can create a segment based on any metric or time engine is both great and bad. Thing on the bad is for one reason that you might create meaningless segments that were telling you nothing. So try to work in a way that you will always hate, have High professes that you want to find out something, and based on Andy will build segment. This is something that will were guys, and you won't spend tense off hours just by creating segments telling you nothing So let's assume you. You you have a scenario that you want to find out. Ah, what? Traffic sources are popular between your purchasers. So you you have to create a segments off producers versus non purchasers, right? Remember the contradictory segment. So this is the hypothesis, is you you you have to have or, for example, you want to find out whether we are content consumers. If you have some block also in your in your website if they convert mawr or not. So we have to create the group of users who viewed a page from lock and compare it to the one who didn't. And then you will see what is the conversion rate off those two groups. So again, I want to find out whether my content guns, inner slumbered more or not. Another thing you want to, For example, find out that what are the traffic sources that your users come from and then they use side search so we will create a segment offside surge users, and then you will look what are the traffic sources and you might may find that they are only branded users or lawyer users have come from either brand search or or direct Perfect . So you you can, for example, exclude them from any campaigns you have, because you know that it doesn't make sense to target them somehow. Or you may find that you have some loyal customers, which would be, for example, the segment off users who bought at least at least three times, which lets you help to define this segment. It would be conditions, and now we want to include only users when number, connections or user is more than I don't know to let CL make. This group is against the only 23 users. Anyway, I have a very short window here, which might be the reason for that small number. Anyway, you will want to find out whether, if they're loyal, it also means that they are coming than directly to our website. So they typed, You are all dressed, directing browser and press enter. If you find out so that they are coming mostly by a director, I figure again can exclude them from any campaign to have, because it just doesn't work on them. You don't have to pay for every any impression off your do have because they are coming directly Ah, or another. Another segment can be Ah, If you find out that, for example, or your hypothesis will be that your your ah mobile users if they come for at least three times from from a mobile device Theun, they just don't don't use sites or, for example, right. Or maybe they just If they, for example, come three times from from a mobile device, uh, then they do not convert. So we only have do sessions due to during which you have a chance to to persuade in, to buy something or to convert. Eso always just just did this exercise. Every time you're going to create a segment, just don't create segment and then trying to find something interesting at first have the hypothesis or a question and then create a segment. This is the way it will. It will work and also compare it, for example, your direct and desktop traffic or story mobile and desktop traffic to see what are the popular sources for for mobile and for and for desktop. Or you will have the hypothesis that your Internet Explorer users do not convert as good as a as a Google chrome, but you'll find out from the reports of browsers. But if we will compare those two segments, you can see with our, for example, top landing pages for for these two users groups. And you will find out the differences between them, right and so on and so on so we can create thousands off thousands off off segment that it has to be meaningful signals, right? Otherwise, you will waste the time. So it's a set. It's very, very public, powerful technique. It just has to be used very piously and only would it make sense? So play with it, have good hype offices, and you can apply it on any report you have in Google analytics so enjoyed. It's it's really great technique and possibility of Google Analytics. 21. Google analytics access rights: G A is a blood for with multiple levels of permissions. There are four of them from bottom to top its reading and lies collaborate, Add it and many juicers for analysis. It's completely enough to have only read and analyzed permissions. Having these gives it access to all reports. Detailed explanation off what every level means is in the interface. When adding a new user, just make sure they do have at least two people who have many users writes. It can happen that one of you will forget the password or split company, and no one else will have permissions to add. Remove new users. It's okay to use the common sense here and take it as your home. You don't want everyone to have access to everything. And now we're going to show Final Two tips, which is about the construct chur and and how to how to use it. We have to go to Admin Section, where we should see something like this and has mentioned Tip number 49 is try to always have at least two users who have m ean writes in every property, or maybe maybe even account level because, as I mentioned. There are many reserves, reason, reasons why it can break somehow or we will forget the password. And then it's now that easy to contact Google that you forgotten your password. Please, can you send me some email where I can reset it into one and so on? So it's just sometimes quite painful if you want to do that. So please remember that it's it's, Ah, three structure. So we have one account, which is, let's say we are one company. Eso we want to have all our websites under one accounts off. If I would, for example, have another website, not just celebrating, that sees that I would see the year and for every property which is a website, we can have more multiple views on and we can add users on every every level on either account, which means that then we would have access to all of the properties here, or we can added on a property level which would automatically give access to all views we have here. Ah Buis set of data on which we are looking and for every website that can be up. Teoh, 25 reviews. I think on it could be. You can have for some reason, separated view for for, ah, mobile and desktop traffic. For some reason, it doesn't mean what sense to me. But anyway, you might find some some very type offices that you want. So four you can create separated views. Eso Let's go back to two user management. Eso think of it as s mentioned as your home. You don't want everyone to have access everywhere. Eso just follow this rule and you will make a mistake depending on which which, uh, level off off a county want to get access? If I, for example, wanted to give someone access to my property, you just have to click on at and New users Important thing. If you want to add someone to Google Analytics, uh, the user has to have a Google account. It doesn't necessarily mean having a Gmail account. But Google account. You can link your address, too. Google, which will create the separated Google account so it could be also your business address. So if you will type for example, some user at gmail dot com Ah, and you will select what kind of permissions you want to have so they're read, analyze, collaborate at it and then manage users, which is the top up, a level of permissions. So this is about about permissions. And then, ah, final tip number 50. Ah, please have a rule that one website has one property. Do not try to measure two separated properties into on account. Sometimes it makes sense to do so, but it's technically, um, slightly more advanced techniques. So once you want to do that, ask someone who's experience in that it requires, Ah, a bit of changing your implementation. It's not that straightforward. And if you will do that just by placing one Trucking Co two to websites because it will break your data. So don't do that just by by the technique I just described. So sometimes it makes sense to have two Properties member into one account. I also did a couple of times, but but please have a good reason for that and have a rule that one website equals one property. It will help you to have ah, clean data that don't need, um, extra filtering or segmentation if you want to do even a very simple analysis. So this were to final dips regarding account structure 22. OUTRO: Congratulations. I hope you found the course useful and full of interesting tips and information. Keep in mind that analytics is the way you think about the later. Verify the hypothesis for multiple views. And don't be hasty when making decisions. I would really appreciate your feedback. Please write me a couple of lines. Would you Liked what you didn't like that much or what should be improved. Thanks and happy analyzing.