Cohort Analysis is one of the hot topics around mobile app analytics and great way for marketers to keep a pulse on customer loyalty and app churn rate. Cohort Analysis is used to look at certain groups of users and following their behavior over time. For example, a new restaurant may want to see if the customers who attended the grand opening are coming back to eat every 1, 3 or 6 months. In mobile analytics, cohort analysis usually focuses on segments of customers that downloaded the app during a common time period, and follows that segment over time. This type of data could reveal the expected life-span of an app, how upgrades or external factors influence app engagement or even how early adopter vs. late adopter behavior differs. This type of information is vital for mobile marketers in order to make actionable decisions around their mobile apps.
Adobe’s AppMeasurement Libraries have the metrics already built in to be able to run successful Mobile Cohort Analysis. Make sure the default LifeCycle metrics have been mapped appropriately to custom variables in SiteCatalyst. To find out more on the LifeCycle metrics in detail go to the Help Section on AppMeasurement.
This post will focus on running Cohort Analysis based on the month of the App Install. To start the analysis, check that your site is collecting the App Install Date in an eVar. The eVar expiration should be set to “never”. (If your eVar value is not set to “never”, please follow up on about using segmentation for Cohort analysis in the next post, however, continue with the remaining directions in this post for getting things set up).

App Install Date eVar Settings

After determining that you have the App Install Date eVar, add a classification for “App Install Month” to that variable. If you need more background on classifications in general please see the Classification Guide on the SiteCatalyst Help site.

cohort_2

Export the SAINT classification file, and add the Month and Year to column B. An easy Excel formula to use that will do this for you is =TEXT(A6,”MMMM YYYY”). Make sure you have column B cell format set to “Text” so that Excel will not automatically convert the Month and Year back into a date. Save the file as a .tab delimited file and upload to SiteCatalyst.

cohort_3

After the file has finished processing you will now have App Install Month Cohorts available in SiteCatalyst. Now you can see how certain Cohorts perform against each other. For example, which App Install Cohort has more app views or app launches this month? Which Cohort has a better churn rate? You can also trend up to 5 Cohorts in SiteCatalyst to view metrics over time.

App Launches by Cohort

Trended Cohort Analysis

The next post will focus using Report Builder for more advanced Mobile Cohort Analysis.

Cohort Analysis was also a highlight of the Advanced Mobile Analytics session at Summit 2013, you can click through to view the session and more!

0 comments