Cohort Analysis

Cohort analysis is an important tool to measure engagement over time. It takes the data from a given web application or eCommerce platform and rather than looking at the users as one unit, it breaks them into related groups called cohorts for analysis. It helps to know if the user engagement is actually improving over time or is it just creating an illusion of improvement because of growth. It proves to be valuable because it separates the engagement metric from the growth metric.

There are certain factors that impact the user behavior and you can analyze them with cohort analysis -

• Ad content
• Target audience
• Channels
• Campaigns
• Website redesigns
• New product and services
• Discounts, promotions, and sales campaigns
  

Types of cohorts

Acquisition cohorts - It divides users into groups based on when they first signed up on your website. You can track them based on a daily, weekly, or monthly basis.

Behavioral cohorts - It divides the users into groups based on the behavior they show on the website or app. This can include any action such as a transaction, app install, uninstall, etc.

Examples of what you can do with cohort analysis

• Session retention by channel and device - Take email and paid search as a marketing channel and compare the customer retention between all sessions. You can also look at the retention you get on mobile devices and desktops. You can improve your mobile experience for the users if you see a low retention.
  

• Retention for sessions with transactions - If sessions with transactions have higher retention, then you need to find out if the users are returning back for subsequent transactions. You can study the behavioral flow by turning these cohorts into custom segments.
  

• Revenue per user by segment - Cohort analysis helps you to capture the worth of the traffic by using revenue per user as a metric