Cohort analysis use cases

This article discusses several typical use cases for which cohort tables are helpful to provide useful insights to take next actions.

App engagement

Suppose that you want to analyze how users who install your app engage with the app over time. Do users install the app and never use the app afterwards? Or do they use the app for a while, then stop using the app? Or do the users remain engaged over time?

You can create a six-month cohort analysis. Visitors do not count as engaged in subsequent months, unless these users have a session or at least launch the app. Cohort Analysis would then show you patterns in usage where App Install always occurs on Month 0. You might notice that usage dips in month 2, regardless of when users installed the app. This analysis allows you to send an email or a push message to all users during the second month after they install the app to remind them to use the app.

Subscription

You work at Adobe.com and offer a free Creative Cloud subscription. The goal is for users to upgrade from the free version to the 30-day trial version or, ultimately, the paid version.

Use Cohort Analysis to understand, for example, that anywhere between 8% and 10% of free Creative Cloud users upgrade in the first month after installation, regardless of when they installed. Then 12-15% upgrade in the second month of use. After that, upgrading significantly drops off: 4-5% in month three, 3-4% in month four, and 1-2% in month five.

Recognizing that you do not want to lose potential customers in month three, you set up an email campaign designed to go out in the middle of month three to a sample of users. In that campaign, you offer a $50 coupon to users who have not yet upgraded.

Check back with your cohort analysis a few months later. For cohorts formed after the launch of the campaign, conversion to paid Creative Cloud subscriptions in month three has risen from 4-5% to 13-14%. The conversion results in hundreds of thousands of dollars per cohort, for every monthly cohort that hits month three from that point forward.

Complex cohort segments

You do analysis for a major hotel chain that targets multiple customer groups for promotions and track the customer groups against performance. To identify the best groups of user cohorts to target, you want to create very specific cohort groups. Use the augmented Inclusion and Return Criteria within Cohort tables to define just the right cohort groupings with multiple metrics and segments. This analysis helps you to identify customer groups that underperform so you can target them with promotions and deals to increase bookings.

App version adoption

You are the analyst for a large insurance company that drives customer engagement through the use of its mobile app. As new features are added to the app, customers should upgrade to the latest app version. You can analyze and compare app versions side-by-side using Custom Dimension Cohort to see which customers on which app version to target. Additionally, you can track retention and churn to see if specific app versions are driving customers away from using the app over time. Through mobile messaging efforts, you can re-engage with these users so the users upgrade to the latest version to take advantage of the latest features.

Campaign stickiness

You are the analyst of a multinational media company that uses targeted campaigns to drive users to their various platforms to drive engagement. Ad spend per platform is based off customer engagement and retention. Successful campaigns are critical to the success of the business. You use the new Custom Dimension Cohort feature in Cohort Tables to compare various campaigns side-by-side to identify which campaigns are most effective at acquiring and retaining users to increase engagement. You can then identify which aspects make a campaign successful and apply that knowledge to other campaigns to increase engagement across various platforms.

Product launch

You are the analyst for a large apparel retailer that has many specific customer segments that drive large portions of revenue for their business. Each segment has specific products designed and created with the segment in mind. With each product launch, you want to know how the new product has boosted sales to various cohorts over time. Using the new Latency Table setting in Cohort Analysis, you are able to analyze a given customer segment’s pre-launch and post-launch behavior and revenue. Using this information, you can identify which products are driving new revenue and which are not gaining traction with customers.

Individual stickiness - most loyal users

You are the analyst of a major airline that derives the majority of their success and revenue from repeat and loyal customers. In many cases, loyal travelers comprise the majority of the revenue and retaining those customers is critical to the long-term success. Identifying the most loyal and consistent customers can often be difficult. However, using the new Rolling Calculation setting in Cohort Analysis, you are able to analyze loyal customer segments and find out which travelers were repeat purchasers month-over-month. You are then able to target these travelers with rewards and perks for their loyalty. Additionally, by switching the cohort type from retention to churn, you are also able to identify which customers were not repeat purchasers month-over-month. Then you can target those segments with promotions to re-engage with these customers so these customers remain loyal in the future.