Cohort analytics is a type of analytics that helps product teams compare cohorts of users within a digital product. Each cohort shares a common characteristic over time, such as acquisition date, purchase date, or a particular behavior. By analyzing cohorts, product teams can tweak features or run marketing campaigns to increase their engagement or retention.
Why do cohort analytics matter?
Cohort analytics allows product teams to track the many different users within their platform and cater to their specific needs. Each group is known as a cohort and may have very different characteristics. Common examples of cohorts:
- Paid users
- Power users
- Users who signed up during a particular week
Teams that ignore cohorts and only use metrics that reflect broad averages across all users can miss important details. For example, in a consumer app, there may be more free users than paid users. If the product team calculated the overall revenue per user and sent a push notification asking all users to pay slightly more than the average, they’d lose out with both audiences. Free users might be offended and churn because any payment would be a big increase from zero and the paid users would get a steep discount, leading to lower overall revenue. The app’s product team cannot treat them all the same. That’s why they need to sort them into cohorts. Cohorts analytics show product teams where advanced users need more features, where less advanced users need more guidance, and where every cohort typically runs into trouble along their user journey.
Learn how tracking cohort growth over time can help you grow your business.
What features do cohort analytics offer?
Cohort analytics track and analyze cohorts of users and have three main characteristics:
Cohort analytics ingest user data through integrations into users’ apps or websites or by allowing users to upload .csv files. Teams typically find that the fewer steps needed for data ingestion, the better.
Cohort analytics interfaces allows teams to manipulate user data, pull reports, and create charts. An attractive interface should be top of mind for teams when they evaluate cohort analytics vendors. The easier to use the platform is, the more likely it is to democratize and spread data insights throughout the organization.
A RETENTION GRAPH OF THREE COHORTS: NEW, POWER, AND DORMANT USERS.
Tracking and measurement
Cohort analytics allow teams to define cohorts with Boolean logic filters or in a drag-and-drop interface. They can name cohorts, save them, run reports, and set alerts for when user behavior deviates from the norm. Teams should seek out cohort analytics that save them time by automating routine tasks such as pulling reports or mining data for insights.
How do you use cohort analytics?
Cohort analytics platforms exist to automate the tedious parts of cohort analysis. Without them, teams must design their own system in-house or export data into a Microsoft Excel spreadsheet and pepper it full of complex equations. Each time the team tweaks or tests their product, they’ll need to export the data again manually. A cohort analytics platform eliminates steps in this process. It integrates into the product directly to track cohorts, alerts teams to meaningful changes in cohort behaviors, and makes algorithmic suggestions. There are 4 steps to performing cohort analysis:
1. Decide which question to answer
Teams with cohort analytics must begin by asking a question tied to a desirable outcome they want to achieve. For example, financial services product teams might ask why the some users churn after three weeks. SaaS teams might ask why some number of users never make it past onboarding. A gaming app team might ask why in-game purchases among power users were low one month relative to previous months.
2. Define the metrics
Each question needs one or more metrics that tie the analysis to a desirable outcome, like sign-ups, purchases, repeat purchases, behaviors taken, or behaviors not taken. It’s important to choose metrics tied to a demonstrably valuable outcome like revenue rather than what are known as vanity metrics such as views, likes, or engagement. They aren’t strong indicators of how to improve the product for each cohort.
3. Define the cohorts
Cohorts are most commonly defined by sign-up or purchase date. For example, a consumer app might define cohorts by week. Every new user would be part of the cohort tied to the week in which they signed up. SaaS companies might track cohorts by their first purchase date and e-commerce companies might track cohorts by either their first or second purchase date. Cohorts can be measured by:
- When users signed up
- When users purchased, once or repeatedly
- Acquisition source, such as from a sale or promotion
- Behaviors, both taken and not taken
- Characteristics such as demographics
- Revenue type or volume, such as cumulative spending over time
4. Perform the analysis
Product teams can run one-off or periodic reports against cohorts to see how they perform in relation to the overall user base or to other cohorts. Based off what they learn, they can implement feature changes or marketing campaigns to nudge users toward greater usage and retention.