[MUSIC PLAYING] SUMIT CHANDEL: Hello there, and welcome to the Firebase Developer’s Guide to Google Analytics Why the Firebase Developer’s Guide to Google Analytics? Well, chances are, if you’re a longtime firebase developer and used analytics for your apps before, you’ve looked at your analytics data in the Firebase console And the console is a great place to get a quick, at-a-glance view of how your app is doing and dig into some conversion and event reporting breakdowns But did you know that if you created your Firebase project after August 2019, your analytics data is available in both the Firebase console and the Google Analytics console as well? That means that you can use a lot of the deeper analysis and reporting tools in Google Analytics via the new App + Web property for your Firebase application data What we found is that not a lot of developers were aware of this And even if they were, they either didn’t know what tools were available in GA or weren’t sure about how to use them effectively And thus, we present to you the Firebase Developer’s Guide to Google Analytics This is our attempt at helping Firebase developers get more familiar and more productive with the suite of Google Analytics tools to more deeply understand their Firebase app data In this video, we’ll be covering the following features in the GA console First, the Comparison Analysis tool, which allows you to make side by side comparisons of different user groups based on things like platform, language, or more audience criteria that you can customize We’ll then present the Segment Overlap Analysis tool, which provides a visual and tabularized form of data to more deeply understand your user segments and find out where they overlap We’re also going to cover the Funnel Analysis Technique available in GA For Firebase developers out there who have used funnels in the Firebase console before but we’re less than happy when they found out that they were open funnels, you’ll be glad to know that the console provides advanced funnel analyses that use closed funnels by defaults More on that a bit later For now, I think the most helpful thing would be do take a quick tour of the Google Analytics console to get a feel for the tools available there and where to find them Now, if only I had someone who is an expert at Google Analytics to give me this tour KEVIN LAM: Oh, hey, Sumit SUMIT CHANDEL: Oh, hey, Kevin For those of you who don’t know, Kevin is the product manager for Google Analytics for Firebase He’s literally the best person to give us a tour of Google Analytics Thanks for dropping in, Kevin KEVIN LAM: Yeah, no problem So some features in the Google Analytics console are exactly the same as what you’d find in the Firebase console So you may recognize a lot of similarities between the two interfaces But what’s unique about what you’d find in Google Analytics is that it offers a much more advanced set of analysis tools to help you get deeper insights into who your users are and how they’re using your app Now, let’s take a look This is Bingo Blast, a bingo game available on both the Play Store and the App Store And if you like playing bingo, definitely feel free to give it a try Now, I’m currently on the home page for my Bingo Blast Google Analytics property Here I can take a quick overview of how my app is doing by looking at some key metrics, like user engagement and retention, which countries my users are coming from, and which events they’re logging most frequently There are also cards that show the sources and mediums from which my users are finding my app, as well as a revenue breakdown below Now, this home page looks very similar to the Analytics dashboard in Firebase, except with a new interface and more information with easier to access filtering options and reporting sections But if you really miss the Firebase reporting view, you can actually switch to it by clicking on the dropdown at the top left of the console and selecting Firebase reporting But let’s stick to the Google Analytics reporting interface for now to show you some of the cleaner and easier to access features I mentioned For example, you can easily adjust the date ranges for the reporting data you want to see in each card and jump into more detailed explorer reports by clicking on the link within the card Going back up to the top, there’s also a unique feature here called Insights, which provides automatically generated insights into your data You can think of these as the headlines telling you what’s going on with your app It’s our way of surfacing up information so you don’t have to do it yourself Now, let’s take a look at some of the other sections of the Google Analytics console Right under the home screen, there’s the Realtime view, which is similar to the stream view in Firebase It provides a realtime view of how many users have been active in the last 30 minutes, and where in the world and which
online networks they’re coming from, and what events they’re triggering in your app Under that is the Report section This is the grouping of standard Google Analytics reports about your users and their demographics, like country, age, gender, and interests There is also a Behavior report that gives you more information about your users’ behaviors, like which screens they’re spending the most time on And below that is a Technology report, which provides a breakdown of your data by platform, app versions, and device models We also released a new set of reports just for game developers that you can access using the switcher, but for this video, let’s stick to the default. Moving on, we have the Events section, which provides a detailed breakdown of the conversions and events that your users are generating in your app These reports are almost identical to the ones in Firebase And below that is the Explore section of Google Analytics This is where some of the deep analysis and data exploration tools of Google Analytics live This will be the most interesting area to explore in this video, and we’ll be taking a deeper look at some of the features in a moment The final section of the console is the Configure section This is where you can configure your analytics audiences, check on your custom user properties, and access Debug view to see analytics events coming in live while developing and testing your application Now, each of these features is exactly the same as the one you’d find in the Firebase console, and can be configured here or within Firebase Now, before we go exploring some of the deep analysis tools in the Explore section, there’s a new feature to call out right here that isn’t available in Firebase And that’s the Comparisons tool But before I demonstrate how to use it in Google Analytics, it’d be really useful to get an overview of what it’s intended to do So hey, Sumit, mind giving us a breakdown of the Comparisons feature? SUMIT CHANDEL: Well certainly, Kevin So the Comparison tool lets you, well, compare different user segments to each other It does this by allowing you to select different dimensions you want to include, or exclude, for the user groups you want to select in your comparisons And what do we mean by dimensions? Well, it’s an intentionally general term, because it can include a lot of different kinds of selection criteria to define the user group you want to use in your comparison For example, one dimension you can select from is the automatically collected user properties in your analytics data, such as demographics, like age, country, and language You can also select from custom user properties you’ve set up in your app So let’s say you have a premium users in your app, or maybe you’re collecting the amount of virtual coins your users have, like we do in Bingo Blast You can set custom user properties to mark your premium users or gamers with more than 10,000 virtual coins, and include that dimension in your comparison as well There are also device dimensions to select device models, brands, or app versions And an acquisition dimension, where you can select campaign tracking dimension values, like campaign sources or mediums, or ad campaign dimensions, like the ad network your users came in from Hey, Kevin, can you show us how it’s done in the console? KEVIN LAM: For sure If you look up top on the home page, you’ll see that there’s a default applied labeled All Users And right next to that is the Add a Comparison button where I can start creating and applying new comparisons Now, let’s say I want to learn more about my Bingo Blast players and to see a breakdown of which ones are most active by different age groups, so that I can figure out where my biggest opportunities are and where my most active users are I’ll start by clicking on Add Comparison and select the Age dimension For my first bucket of users, I want to look at those who are between the ages of 18 to 34, so I’ll select these two ranges here Now, if I wanted to, I could also add more dimensions to this, like gender or language But for now, let’s keep it simple and stick to age So I’ll hit Apply, and then after a couple seconds, you can see that the comparison is showing up in the Users chart here, as well as all the additional charts and cards below But we still need to add a couple more comparisons so that we can actually gain meaningful insights from this data So let’s add another comparison for age I’ll click on Add and add the Age dimension once more This time, I’ll select users between the ages of 35 to 54 Then I’ll click on Apply And now this new comparison is also showing up and being applied to my reporting charts and data Now, let’s add one more comparison for the last age group I want to include
And that will be for users between the ages of 55 and above I’ll click on Add Comparison one last time, select the Age dimension, and pick the age range 55 to 64 and 65 and above Then when I click on OK and apply this comparison, we’ll now see that the reporting data has the final comparison added in, and I’m ready to dig into the data So I can see some useful data here For example, I can see that the number of users in the 35 to 54 age group is about 65% larger than that of my 18 to 34 group And my 54 and above group is almost 71% larger than that of my 18 to 34 As I scroll down to the breakdown by countries, I can see that the top seven for each comparison are different And especially so for my age 18 to 34 group And aside from that, the top three countries are almost the same across each comparison group Scrolling farther down to the Revenue section, I can see that the revenue distributions across platforms by percentage are similar across age groups, with slightly more of the revenue coming from iOS users for my age 55 and over user group Another interesting insight is if I hover over the Android portion of the revenue for each user group, I can see that although my age 55 and over group accounts for the most number of users, those who are in the 35 to 54 group account for a much larger portion of my total revenue on Android Using these insights in combination can help inform where I should target my growth campaigns From this analysis, I learned which groups are lagging and which groups are generating the most revenue From there, I can actually consider if I want to use a growth campaign to increase the number of users in my lagging group or to further engage the highest revenue generating groups instead Now, another useful feature about the Comparisons tool is that once you’ve applied it, you can see the comparisons throughout other parts of the product For example, if I go to the Users or the Demographics report, I can see the comparisons being applied there as well And that allows me to compare and contrast regardless of where I am in Google Analytics Now that we’ve finished this tour and checked out the Comparisons tool, it’s time to dig into the Explore section, and to look at the powerful techniques to help me uncover insights about my app There are quite a few in here, and rather than covering them all, let’s focus on a couple that we think are most important, or most relevant, to Firebase developers Now first, let’s talk about funnel analysis Hey, Sumit, I heard that this was particularly important to Firebase developers Can you tell us a bit more? SUMIT CHANDEL: Ah, funnels Yes, if there is one analytics feature in the Firebase console that I’ve heard plenty of feedback and feature requests about, it is funnels So here’s the thing, funnels actually exist in the Firebase console, but they’re what are called open funnels What’s an open funnel? Well, I’ll see your question and raise you another and ask, what’s a funnel in general? For those of you who don’t know, I thought it might be helpful to cover the concept of funnels, since not everyone has encountered them before So a funnel looks like this If you’ve ever taken a chemistry class before, it probably looks familiar to you The idea behind the funnel in Analytics is similar to how a funnel works in physical reality As you pour in a large volume of liquid at the top of the funnel, it starts to narrow down as it travels down, and gets smaller and smaller towards the bottom In Analytics, we use the concept of funnels is to answer questions like, out of all the users who have started a game in Bingo Blast, which ones leveled up to the next stage? And out of those, which ones moved on to make an in-app purchase? Similar to a physical funnel, the number of users typically narrows down as you travel further through the funnel That is, if we’re dealing with closed files There’s also something called open funnels, which present data bit differently Each part of an open funnel represents an event, just like with the closed funnel However, in an open funnel, the number of users counted in each part of the funnel are out of the total number of users in your app, and not just the number of users who completed the previous step of the funnel That means that in an open funnel, you might have a distribution like this for the same session start, level up, and in-app purchase events we just saw Notice how the level up events is greater than the number of session start events Isn’t that weird? How could there be more users who logged a level up event than the number of users who even opened the app in the first place?
The answer is, we’re looking at open funnels Since this is an open funnel, the count of users who logged level up events in the second part of the funnel is out of the total number of users in your app, not the number of users who started a session in the first step Combine that with the fact that players can log multiple level up events and you end up with a situation where the second part of the funnel is larger than the first, and you have a conversion rate higher than 100% Now by default, funnels in the Firebase console are open funnels And this is something that a lot of Firebase developers were sometimes confused about In fact, we receive plenty of feedback to provide closed funnels to analyze Analytics data, since developers typically expect funnels to be closed But while these aren’t available in the Firebase console, they are available in the GA console Also, in addition to that, funnel analysis reports in Google Analytics have a lot more features that can get into much more detailed breakdowns of your data Let’s take a look OK, so we’re back in the console, and we’re about to take a look at the first analysis tool in the Explore section So I’ll expand the Analysis menu here and then select the Funnel analysis technique from the selection below Now I’ll create a new funnel analysis by clicking on the plus button right up here, and then selecting Funnel analysis OK, I’m ready to start So let’s take a look at the previous example I was looking at Let’s say I want to perform a funnel analysis to answer the question, how many users who start a game session then go on to level up, and then go on to making a purchase in Bingo Blast? At the basic level, what this amounts to is creating a series of steps, with each step building a part of the funnel The first step will be a user starting a session The second step will be triggering a level up event And the last step will be completing an in-app purchase Now, at first glance, there’s a lot of stuff going on in this interface, and we’ll take a look at it in a moment, but for these basic steps to create our first funnel, all we need to look at is the Steps section highlighted here So let me edit the steps for this funnel analysis If you’ve used the audience builder in Analytics before, you’ll notice that this interface is very similar First thing up top is to name the first step of my funnel I’ll put the name, Session Start, and now just below that I can select the condition for this step For the condition, I just want to select the Session Start event that gets logged when a user starts playing Bingo Blast from the Events menu here So I’ll select Events, and scroll down to Session Start, and select it from the menu Now already, there are a couple of cool things that happened here First, see that Summary card at the right? This is actually updated with the total number of users and events counted in my funnel based on current conditions It will also automatically update as I add more This is a unique feature in the Google Analytics console that isn’t available in the Firebase console when creating a new funnel And there’s also another cool thing that happened here as well You might have noticed that when I selected my condition, there are way more than just the events to choose from I could also create conditions based on dimensions These include automatically collected user properties, like age, country, and device category, as well as custom user properties registered to my game So let’s say I wanted to take a look at a specific age group and apply that to the condition to the first step of my funnel, for example, users in the 18 to 34 age group Well, I can just add another condition, select Dimensions, and then my automatically collected user properties, and then select Age I’ll then select the 18 to 24 and 25 to 34 age groups as values to include, and it’s done My funnel is now including all users within the 18 to 34 age group who have logged a Session Start event Also note that the Summary card to the right has been updated with the user and events counts based on the newly added condition The ability to select more than just events when creating a funnel is something unique to Google Analytics, and can come in handy for more fine grained analysis of your Firebase app analytics data To continue with my example, though, I’ll remove the age condition and continue adding on my second step in my funnel, which I’ll name, Leveled Up Also, you’ll notice that just above that I have an option to choose how this second step should be counted I can choose if I want to count it if it occurs anytime after the first step or if it occurs immediately after the first step I can also select the time window for how long after the first step
I want the second step to occur In this case, I want to include users who leveled up anytime after their session started, so I’ll set step 2 to count if it indirectly follows step 1, and won’t specify a specific time window for when step 2 occurred Now I’ll add the condition, which is basically when a user has logged a level up event, indicating that they have indeed leveled up So I’ll click on Add new condition, select Events, scroll down to the level up event, and then select it Notice how after I’ve selected it, the Summary card updates once again at the right of the screen Now I’m ready to add the final step of my funnel, which is the action of making a purchase in Bingo Blast So I’ll click on Add step once again and name the step Made Purchase This step should be counted no matter when it happens after the level up event in step 2, so I’ll leave this set as is indirectly followed by, and leave the time window unspecified Now I’ll add the condition, which is when a user has made a purchase and logged an in-app purchase event So I’ll click on Add new condition, select the Events category, and then scroll down and select the in-app purchase event OK, my funnel is complete, and the Summary card has been updated to show that I have about 100 users who have taken all the steps defined in my funnel Now that my funnel is ready, I’ll hit Apply at the top right, and I’m taken to the Funnel Analysis view where I can start digging into the data First, let’s take a look at the funnel graph analyses So I can see that I have about 6,600 active users who started a game session Among those, about 3,200 have leveled up in the game That’s about a 47.7% completion rate, which I can see up top up here And it’s also a drop of about 3,500 users between starting a session and leveling up For the 3,200 users who did level up, about 100 of them actually made a purchase, leading to a 3.1% completion rate, and a drop of about 3,000 users between leveling up and making a purchase Just below the funnel graph is a tabular breakdown of my data I can see the number of users included in each step, the completion rate going from one step to the next, and the number of abandonments and the abandonment rate between each step Now let’s look at some of the additional tools in the funnel analysis feature that we didn’t look at before back in the left-hand side So first we have the Tab Settings pane Here we have a Segment Comparison section, along with a prompt to drop or select a segment Then we have the Steps section, which we just used And below that, a Breakdown section, with another prompt to drop or select dimension The Segment Comparisons and Breakdown features that get applied to our funnel are quite powerful, and are unique to Google Analytics Let’s look at the Segment Comparison feature first So using this feature, we can segment the users in each step of the funnel to figure out things like, say, the number of users from different countries, and how the funnel graphs for these users compare The full list of available segments is in the Variables pane to the left of the Tab Settings pane You can also create any custom segment you’d like by clicking on the plus sign There are also ready-made suggested segments and template segments you can choose from as well But let’s say in my case, what I want to know is what the funnel breakdown looks like for users in my top Asian countries, and how the data compares These countries would be Japan, India, and Thailand So let me create those segments now So I’ll create my first segment and select User Segment, and add a new condition Here it will be the Country ID user property, which I’ll set to JP for Japan Now I’ll do the same for India, except this time the Country ID value will be IN And do the same for Thailand, setting the Country ID to TH this time And now that I’ve got my funnel data segmented by country, I can compare them in both the funnel graphs and the tabular data below them Now let’s look at the breakdown section in the Tab Settings pane So here I can break down all the tabular data shown for my funnels by additional dimensions to get even more insights into my data The full list of dimensions I can choose from are here in the Variables pane I can add new dimensions from a number of different categories as well, like Acquisition Channels to see which sources and mediums my user
came from; Advertising to see a breakdown of users from different ad conversions; User Behavior, like which other events they triggered while playing the game, or which screens they spent the most time on; E-commerce events; Generic events; and more In this case, let’s say I want to get a breakdown of users in my top Asian countries by platform, and see how many Android users and how many iOS users I have in each step of the funnel So I’ll navigate to the User section, scroll down to Platform, and then select it and hit Apply Now I’ve got Platform listed as an additional dimension in my Variables pane So now I can drag and drop the Platform dimension into my breakdown in the Tab Settings pane to have it applied to my data So I’ll do that And voila, I can now view the platform breakdown for each of my users in each country and in each part of the funnel in my tabular data I can also view the completion rates and abandonment rates for each of these users across the steps of the funnel too So the funnel analysis is a very powerful technique you can use on your Firebase app data through Google Analytics And the best part is that, yes, they are closed funnels, which is a big deal and big ask from Firebase developers But if ever you miss open funnels, you can always use this toggle here to turn this into an open funnel, or just use funnels in the Firebase console Hey, Kevin I think there was another technique you wanted to show us in Google Analytics Is that right? KEVIN LAM: Yeah, Segment Overlap While the funnel analysis technique is great for figuring out how many users move on or drop off from critical steps of my flow, sometimes you may just want to get a better understanding of your user base as a whole You may, for example, want to know which traffic sources your users are coming from, what age groups they belong to, and which platforms they’re using, as well as maybe which countries they’re coming from Knowing all this information can tell you what kind of audiences you should be targeting for your next campaign to get even more users into your app But now let’s say you only have a limited budget to spend Should you spend the money on users in the 18 to 34 age group, users who are on iOS only, the intersection of the two, and how do you factor in countries? To answer questions like this and to figure out how you should target your next campaign, you need to figure out the right combination to get you the largest and most opportunistic user group To do that, you can use the Segment Overlap technique This technique allows you to create and choose segments you want to use to identify overlaps between them, both visually and in a tabular form You can also add breakdowns across different dimensions and metrics, just as you had done with the Funnel Analysis technique Now, let’s use this technique to solve a real problem In Bingo Blast, we haven’t invested a lot of money in pay traffic before, but I want to change that I want to start by figuring out a good audience I can target that will provide me with both additional users coming into Bingo Blast and insights into whether my campaign strategy is working So let’s go into the Google Analytics console and see how using segment overlaps can help So now I’m back in the Google Analytics console and looking at a new Segment Overlap technique You’ll notice that the Variables and Tab Settings panes to the left include Segments, Dimensions, and Metrics, just as you had seen when Sumit was looking at the Funnel Analysis technique But in the Analysis section, we see a visual representation for our very first segment, which is all users in my app What I want to do is to start figuring out the characteristics that make up the majority of my user base so that I know how to target my marketing campaign First, let’s see what interests are the most common within my user base I’ll create a new segment to apply the overlap analysis by clicking on the plus button, selecting User Segment, and then adding a condition for the segment In this case, the condition I want apply is the User Interest category, so I’ll go into the Dimensions, Automatic, and scroll down to Interests, then select that As for which one to pick, I’m pretty sure that Media and Entertainment will encompass the majority of my users, but there are multiple categories here Would it be the one with Gamers, Casual and Social, or the Comics and Animation Fans? My intuition is that it’d be the Gamers interest group, but let’s validate that with our tool I’ll start by selecting the Gamers, Casual group
I can see the Summary card update and tell me that there are about 4,100 users included in the segment before I even apply it Now, let’s see what it looks like if I switch to the other one Wow, this group actually has 6,000 users, which is even more than the last one, and definitely not what I initially thought would happen But it’s a good example of how a tool like Segment Overlap can help you validate or disprove your assumptions Now, let’s move on and apply this segment to my analysis I’ll name this Comics and Animation, and click Save and Apply OK, now I get a nice visual for how my users stand amongst my global base But before I go ahead and decide that these are users I want to target, there are some other important factors I want to fold in as well For example, I want to make sure that I’m targeting users who are most likely to spend in my game, so let me apply another segment for purchasers And let’s verify that with the overlap So I’ll create another segment by clicking on the plus button, select User Segment, and add another condition This time, the condition is going to be based on the in-app purchase event So I’ll select Events from the menu, scroll down to in-app purchase, and then select it Now, let’s name this one Purchasers and click Save and Apply Cool I’ve got another interesting insight here Despite the fact that the majority of my users are in the Comics and Animation group, there’s a significant amount of purchasers who are outside of that And looking at the Totals table below, I can see the breakdown of users for each segment Specifically, I see that the total purchasers in my app is 127, but the total users who are both purchasers and comics is only 91 So if I were just to target the Comics group, I would miss a significant portion of my potential purchasers I’ll have to make a slight adjustment here And for the purposes of this video, let’s say I’m OK with the fact that I might lose some purchasers, and let’s just dig even deeper into that overlap There’s a useful feature in this view that allows me to create a new segment from any combination in the table below In this case, I want to combine the users in the Purchaser segment and the Comics segment So I’ll right-click on that row and click on Create segment from selection This will open up the Segment Editor, which allows me to combine the conditions from the segments I previously selected Now let’s click on Save And ta-da, I have this new combined segment instantly created At this point, you may be wondering, why am I combining segments and not just keeping them separate and then moving forward? Well, part of the reason is that you can only have three segments at a time By combining the previous two, I can effectively discard them and add even more layers on top of my current analysis The other reason, and this ties back to why I’m doing this in the first place, is that I can create an audience out of the intersecting segment and use that for targeting purposes So now that I’ve combined my purchasers and comic enthusiasts, I can remove their individual segments And we’ll delete them from my list of variables So let’s go ahead and do that First, deleting my Purchaser segment and then deleting the Comics and Animation segment There, now I can add the combined segment to my analysis by dragging it over to the list of segment comparisons I can probably remove the All Users segment as well, so let’s clean that up too OK, now that I’m just looking at my comics purchasers, I can break it down even further Let’s see what the breakdown is between my Android and iOS users I already have these segments created, so it’s as simple as dragging and dropping them in Looking at the breakdown below, I can see that virtually all comics purchasers are on Android And that’s pretty powerful, because when I think about how I want to optimize ad campaign, it allows me to really keep focused To take it one step even further, I’ll clean up my workspace by removing the iOS users and consider, alas, one more factor, age This time, instead of adding a new segment, I can actually add a new dimension directly in the table To do that, it’s as simple as dragging and dropping the Age dimension from the Variables pane onto the breakdown And after letting Google Analytics process the data, the tabular view is updated so that I can analyze it even further And looking at this information, I can see that the majority of my users are within the 25 to 64 age groups
That’s great, because now I can actually refine my ad campaign As you’ve just seen, this was just one quick example of how you can use the Segment Overlap tool But there so many more use cases where this tool can come in handy So we invite you to check it out for yourself using your own Firebase data, and to perform your own analysis using the automatically collected events or the custom events that you think are most important to your business And with that, Sumit, are we ready to wrap up? SUMIT CHANDEL: Yes, we are, Kevin It’s about that time Well folks, we hope some of the tools we covered in the Firebase Developer’s Guide to Google Analytics were helpful for you Now, we only covered a handful of the tools available in Google Analytics, but there are plenty more For example, you may have seen the Path Analysis tool in the Analytics menu, which lets you get a closer look at how your users are navigating through the different sections of your app Or you might have seen the User Explorer tool, which allows you to perform analyses at the individual user level, rather than at the aggregate level, to get another perspective into your user data Please check out the docs linked in the description to learn more about these tools, as well as the tools we covered in this video And hey, if you’d like us to do another video on more of these tools in the future, let us know by liking the video and subscribing to the channel, and maybe we’ll be back with a part two In the meantime, have fun analyzing your Firebase app data with Google Analytics, and get out there and build something great for your users Bye for now KEVIN LAM: Bye [MUSIC PLAYING]