Cohort Analysis

Which group are you in?
Contributed by

Val Yonchev

Published December 18, 2018

What Is Cohort Analysis?

Cohort analysis is a subset of behaviour analytics which studies the difference in behaviour between different groups of people (customers / users), aka cohorts.

Instead of treating all users as part of one population, cohort analysis is segmenting the population in cohorts which share a common characteristic and then search for patterns in their behaviours.

The discriminating characteristic can be derived from the main personas (e.g. skill level) or from the events and timeline relative to the product (new users, returning users, users from month x, etc.)

Why Do Cohort Analysis?

This is a feedback loop practice, which allows the team to get prompt feedback from real life use of their changes.

The Cohort Analysis aims to produce actionable insights, which can inform better product decisions instead of using gut feel.

This practice enables product decisions for improvement, enhancement, overhaul or discontinuing of products/features. It is key practice in the sequence of practices which start with Design of experiments and aims to improve the product performance in terms of business outcomes.

The cohort analysis helps the team understand the impact of product/feature changes by comparing performance across comparable data sets (e.g. newly signed up users). It can help drive improvements towards a achieving a concrete and measurable goal, the pursuit of which results in change of user behaviour.

How to do Cohort Analysis?

  • Use Annotations: Start using Google Analytics Annotations or a marketing calendar to track your marketing activities and campaigns. This will help you see the impact of your marketing initiatives in Google Analytics reports, including the Cohort Analysis report.

  • Test Your Marketing Activities: Start running experiments to test audiences, channels, and website design by week or by month (eg test one marketing channel per month). Schedule a half-hour one day every week to look at your Cohort Analysis report to check the results of these tests.

  • Drill Down on Unusual Cohorts: If you see significant increases or decreases in metrics like user retention for a certain cohort, check your marketing calendar to identify what changes could be driving these movements in your metrics. Turn these cohorts of interest into custom segments to dig deeper into their behavior.

Look at Cohort Analysis

Links we love

Check out these great links which can help you dive a little deeper into running the Cohort Analysis practice with your team, customers or stakeholders.

Except where noted, content on this site is licensed under a Creative Commons Attribution 4.0 International license. This site is graciously hosted by Netlify