Simplifying Cohort Analysis with the Tapjoy Platform

by • October 19, 2015


For many app developers, the concept of cohort analysis can seem a bit intimidating. But cohort analysis is critical, and quite simply, it is the best way to analyze your users’ behaviors and truly understand what’s going on within your app. It strips away all of the background noise that comes with aggregate reports and lets you hone in on what matters most: app engagement, retention and monetization.

Tapjoy has made cohort analysis extremely simple for developers to understand. Using the Tapjoy dashboard, building a new cohort is as easy as choosing from among the many pre-configured cohort variables or identifying any of your own characteristics, then clicking a button to get just about any report imaginable for that particular group. Or click and drag any two cohorts to compare one against the other, then easily take action based on the insights you see.

Below are a few helpful hints to help you get the most of Tapjoy’s cohort analysis capabilities.

Build custom cohorts for your app

Whether you want to track a simple cohort such as users who installed the app within the last week or a more complex cohort using multiple variables, Tapjoy makes it easy to build custom cohorts for your app.

From the main Overview tab in the Tapjoy dashboard, click the “plus” button in the upper left to start building a new cohort. We provide seven different pre-set variables for you to choose from, including operating system, app store, country, referrer, game version, device type and paying status. You may choose any combination of these seven variables, or you can choose to create your own custom cohort variables based on your app’s features and events, such as user level, friends count, purchase group, etc.

For the example below, we created a sample cohort that consists of Paying users in the United States on the Android platform that came from an advertising source known as “AD1.” We’ll select these variables and click “Save.”


Compare your cohort to the total user base

Once a cohort is defined, you can easily compare the behaviors of its users to your total user base so that you can identify where the cohort might be under- or over-performing for certain metrics.

In our test scenario, by clicking on the IAP Revenue tab we learn that our Test Cohort has a higher percentage of paying users than the total user base (1.69% to 1.37%) and a higher ARPDAU ($0.19 to $0.10), but a lower overall ARPPDAU ($5.55 to $7.01). This tells us that even though more users are paying for our in-app content, on average they aren’t paying as much, which we can confirm by looking at the “Revenue per Purchase” chart. We might therefore want to try promoting higher-priced items to these users in order to increase their average revenue per purchase.

Compare two cohorts side-by-side

Sometimes you’ll want to compare a cohort directly to another cohort instead of your entire user base in order to get a clear understanding of how specific variable affects performance. For instance, if we want to understand how referrer “AD1” is driving high-value users in relation to other ad sources, we can create another cohort selecting “AD2” as the referrer and the do a side-by-side comparison.

By comparing these two cohorts next to each other, we can easily see that the “AD1” referrer drives users with a higher ARPDAU. Therefore, we might choose to invest more of our advertising budget into this referral source than “AD2.”


Compare side-by-side for retention

Similarly, you can compare the performance of two distinct cohorts side-by-side to see how they stack up for another key metric: retention. Similar retention and revenue reports can be run for any cohort variable, from OS and app store to app version, device type and more. These insights are not only helpful for optimizing ad spend but also for deciding how and where to deploy certain engagement or monetization campaigns, which we’ll get into next.


Target specific offers and promotions by cohort

Once you have a cohort set up in the Tapjoy dashboard, you can target that cohort with any type of in-app content you want, whether a monetization campaign such as a product promotion, display ad or video ad, or an engagement message such as a reward or announcement.

In the example cohort we created earlier, in which we noticed that users had a lower revenue per purchase than the rest of our users, we can create an in-app product promotion that encourages them to purchase one of our higher-end virtual goods or virtual currency packages. All we have to do is go into the “Monetize” tab, click the “Create Content” button and then the “In App Product Promotion” button and select “Create.”

When choosing which user segment the promotion should be delivered to, you can create a new target user segment or simply select “Import User Segment” to select the cohort we’ve already identified. Once the content is created and the placement identified, the content will be delivered automatically to the appropriate cohort.

Ready to get started?

There is an almost limitless number of cohorts that can be created in the Tapjoy dashboard using any of the pre-set or custom variables available. Creating cohorts is as simple as checking a few boxes, and our side-by-side comparison gives you very clear insights to help monetize, engage and acquire users more effectively.


Source: Tapjoy

Comments are closed.