Got a struggling app? Save it with cohort analysis
One of the most important events in measuring a customer’s lifecycle of engagement with an application is the moment when they install or first monetize.
Understanding those critical days, weeks or even months afterward are what can determine either the success or failure of an application.
How does one continue to monitor these essential metrics when customers are in a constant state of flux you ask? The simple answer is cohort analysis. While statistical in nature (and a tongue-twister to say), cohort analysis is an extremely powerful tool that can benefit nearly every role: product managers, marketers, executives and analysts alike.
What is Cohort Analysis?
At a basic level, cohort analysis is a type of segmentation. The difference is that it’s based on a common characteristic or experience shared by a group of customers within a defined period of time.
Think of it in terms of comparing the post-graduation performance of two different classes of students—one that graduated in 1950 and another in 2010. At what point did they get their first job? Six months post-graduation? Three months?
This same concept applies to groups of customers with regards to mobile and social apps. What campaigns have driven the highest amount of revenue per user seven days after installation? After the first monetization event, what segment of customers showed the highest engagement in terms of number of sessions in the first month?
Cohort Analysis Lets You Make the Right Changes
Beyond knowing what app features are the most effective, cohort analysis lets you also know when they may have the most impact.
For instance, does offering a free shipping promotion immediately after installation motivate customers to make their first purchase? Does pushing out an in-app notification about leveling up three days after a user first monetizes increase their engagement in a game?
Separately, cohort analysis is a great way to benchmark performance between applications for groups of customers. How does the post-install behaviour of your new app differ compared to a successful legacy app?
Being able to look at this comparative behavior can immediately uncover potential opportunities or obstacles in the days and weeks after launch.
Most importantly, at the heart of customer lifetime value is cohort analysis. You have to know when a customer first engages with your app in order to track their lifetime value. Looking at your app’s overall revenue performance will only give you part of the story.
You need to see how each of your cohorts of customers are monetizing over their lifetime, as well. Without understanding that revenue metric, you may never know if you’re generating positive ROI. This is especially important in cases where a customer’s lifetime value is less than what you’re spending to acquire them.
Which Cohort Analysis Tool is Right for You?
These cohort analyses exercises may seem simple, but it can be an extremely complex process to first capture and then “normalize” groups of customers based an event in order to compare behaviours over time. So what is one to do if they lack the know-how or back-end systems to perform these types of analyses?
Today, most Web and mobile data intelligence platforms offer some form of cohort analysis. However, the granularity and flexibility provided by each tool can vary greatly.
When evaluating solutions, make sure you understand the use cases for cohort analysis (like the ones mentioned above). And whichever one you pick, make sure it’s capable of answering all your questions. It’s the best way to get the insight you need.
And if you’re going to be in the San Francisco Bay Area on May 20-21, don’t forget to join us at the Kontagent Konnect User Conference 2013 in San Francisco. We’ll be showcasing cohort analysis and other best practices for mobile and beyond!
About the author: Kylee Hall is a product marketing director at Kontagent. She has a penchant for data-driven marketing and holds an MBA in econometrics and statistics and marketing management. Previously, she worked at The Nielsen Company and Coremetrics (now IBM). When not geeking out on the latest data and trends, she spends her time as an avid crossword-puzzler and on the search for the perfect margarita. You can reach Kylee at @kyleehallb.