At Apsalar, we’re focused on providing the most powerful set of actionable analytics to increase user engagement for mobile app publishers. Our mission is to help our customers improve retention and increase monetization from engaged users. We recently launched an educational blog series to help developers understand the various types of analyses available to better understand user behavior and increase user engagement. Catch up by starting here.
In our most recent post, we introduced the concept of creating cohorts to better understand user behavior and the correlation with retention. Let’s quickly recap:
A cohort is a group of users that has completed a specific action within a specific period of time: for example, a set of new users that launched your app on Thursday, December 1. Once a cohort has been identified, it can be used to better understand three dimensions of user behavior:
- Retention: Are users coming back regularly to the app?
- Engagement: Are users “active” within the app?
- Monetization: Are users spending money?
We’re going to dig into why cohort analysis is essential to managing engagement. Engagement is the measure of a user’s activity levels within your app and the extent to which users take advantage of the features you’ve built and the goals that you’ve set for them to achieve. As such, each app publisher may define engagement differently:
- For a game like Angry Birds, it could be reaching higher levels.
- For a shopping app like eBay, it could be actively bidding on an item.
- For a social app like Instagram, it could be sharing photos with friends.
In the end, higher levels of engagement should lead to higher retention (measured in terms of Daily Active Users – “DAUs”) and more monetization (measured in terms of Average Revenue Per User – “ARPU”). That’s why it’s crucial to analyze and optimize your apps for user engagement.
Let’s run through a basic engagement cohort analysis example:
You have what you believe to be the greatest photo-sharing application in the app store. Your user base has grown steadily since launching a few weeks ago, but you’re not sure whether users are continuing to share photos in the days after downloading the app. Are your users engaged with the app or not? Cohort analysis can provide the answer.
Instead of relying on inconclusive top line information such as user reviews and vanity metrics, you want the ability to accurately track photo-sharing on two dimensions:
- Within a cohort: how did users in a specific cohort engage (share photos) over time?
- Comparing cohorts: how did users from different cohorts engage over time?
To demonstrate, we’ve created a cohort of users with the event defined as “Launched App” (the first time they launched the application, by day) and we’ve defined the engagement indicator as “Sharing Photos.” Here is the setup:
- Cohort event
- Event: “Launched App”
- Event: “Shared Photo”
This will help uncover what percentage of each cohort is sharing photos following their first use of the app.
Note that the ability to look at cohorts on a daily basis (i.e., each new day constitutes a new cohort) is important to see how many of your users are sharing photos in those critical first few days after they’ve launched the app for the first time.
If we look in column 1 in the graphic below (which represents the first day after the cohorting event), you’ll see that there are a high percentage of users—about 80%—sharing photos after first launch. This is fantastic. Unfortunately, that percentage begins to drop sharply in the days following. This is a clear indication that users are quickly becoming less engaged. By the second day, the percentage of engaged users is below 50%, and by days 5 & 6, less than 25% of users are sharing photos. These may be retained users in that they are showing up, but they are not engaged if your app was specifically designed to enable photo sharing with friends. Also, if you are only looking at top line trending data, photo sharing may appear as if it is on the upswing as new users flood in and initially share photos in the first day or two. But it’s the cohort analysis that reveals the true level of engagement.
Indeed, if you make a change to the app to improve the photo sharing experience, all you need to do is compare the cohorts of users that have downloaded updated versions of your app, to the cohorts relating to the prior version. In this way, you’ll be able to quickly see if the changes improved engagement or not, as the photos will be shown as being shared more or less frequently and for a longer or shorter time relative. These are data driven results that can’t be ignored.
With cohort analyses, you have the data to keep iterating and improving the user experience for greater engagement that can result in higher retention and monetization. In fact, with a fully flexible cohort analysis system, that allows you to track any event as a cohort event, and not just first use of the app, even more powerful data can be collected. Now you can track retention and monetization indicators using any cohort event. In the example given below, if the first sharing of a photo is the cohort event, then the cohort is the group of users who first shared a photo on a given day. With a retention indicator as launch of the app (after the cohort event), you can track the impact of engagement, defined as photo sharing in this case, directly on retention. In other words, did sharing photos keep users coming back to the app and drive up DAUs? The ability to answer questions such as this will be key as you strive to build and create a profitable and sustainable mobile business.
Check back in to our next post as we show how to use cohort analysis to better calculate the monetization of your users.
If you have any questions, feel free to drop me a line at firstname.lastname@example.org.