Join our mailing list. Every so often you'll get analytics inspiration, product updates, and our best content.


How do cohorts help you measure impact over time?

June 28, 2018 • SuiLin Yap

Welcome to Feature Spotlight, a series here on the Facebook Analytics blog that highlights exactly how specific analytics features answer the questions marketers are asking. Today we're taking on cohorts, the feature that answers: How can I measure the impact of product updates, marketing campaigns, and more over time?

Most of the time, marketers aren't necessarily looking for more data. But cutting through the noise and vanity metrics to find the answers you know are in your data is hard to do. Cohorts help you understand the impact of product updates, marketing campaigns, and more on your websites, apps, Facebook Pages, and Messenger bots over time on important metrics for your business like retention and lifetime value.

What are cohorts?

Cohorts, broadly speaking, are a group of people tracked over time. In Facebook Analytics, we define a cohort as a group of people who perform a certain action and then take another action during a specific time period. For example, a travel company might want to know how long it takes people who started tracking flights to Europe to actually buy a flight. In this case, the first event is adding Cologne to their list of flights to monitor, and the second is purchasing a flight to Cologne. A cohort analysis would be the best way for Jasper's Travels to answer that question.

What's the value of cohorts?

Cohorts help you cut through the noise: Use them to determine the impact of your marketing campaigns, product changes, app updates, and more on your customers' behavior. For example, if you're implementing a new onboarding experience, you could monitor retention over time in your cohort groups.

How? Creating a cohorts let you establish a timeline against which you can measure your KPIs. If you know how long it takes your audience to take an action (whether it's buy, subscribe, or another conversion event) on average, you can measure not only conversion rate or revenue, but see the impact on your customer's lifecycle.

Analyzing long-term value based on a conversion event helps you identify their most successful users and customers, impact of changes to products, app experience, etc.

What metrics do cohorts help you measure?

Cohorts give you a richer understanding metrics with different groups.

Depending on the parameters you assign to your cohorts, you can use them to see the impact on the metrics that matter to your business, like LTV, retention, and engagement.

Here's what cohorts look like in Facebook Analytics.

How do different types of businesses use cohorts in Facebook Analytics?

  • Ecommerce. How many people view content on your website, then purchase? Or, since Facebook Analytics can integrate other Facebook marketing tools like Messenger and Pages, you could create a cohort around clicking a CTA from your Messenger bot, and making a purchase.
  • Travel. How many people who track flights to a certain destination book within a month? Or, how many view a post on your Facebook Page, and then go on to track a flight?
  • Games. Create a cohort to track the rate of in-app purchases after the first app launch. When are you acquiring your most valuable users?
  • Publishers. Publishers often want to increase user retention. See how many people view content, and then return to your site or app in 3 days or less. Or, monitor how many people comment on a post, and return to view content in a week.

How do businesses use this data to optimize their strategy?

  • Check the predicted LTV of these groups, out of the people who ultimately convert, and use that number to determine if it's worth changing your website experience, or creating a retargeting campaign.
  • Evaluate (also through LTV data) whether it's more cost-effective to acquire new users, and retarget current customers.
  • Create a retargeting campaign around the items or content people viewed, or added to their carts.
  • Identify the best-performing cohorts, and dig deeper to discover what they have in common.

How are cohorts different than funnels?

The two biggest differences between cohorts and funnels are:

  • The two actions taken by the cohort group are taken within a defined period of time.
  • Funnels can measure more than two user actions.

Being able to build a cohort based on when a group of people first took a specific action is key. Cohorts and funnels are similar in that you're tracking a group's progress through a series of actions, but you're not able to easily measure the impact of timing in a funnel—the impact of a marketing campaign that was running on your homepage for a specific time, for instance. Funnels are handy when you need to measure progress through a sequence of more than 2 steps, and help you identify bottlenecks or opportunities for optimization.

How can I dig deeper into my data with cohorts in Facebook Analytics?

Apply segment parameters device, demographic, app version, traffic source, or your own custom segments. You'll be able to to pinpoint exactly who your most successful customers and users are. You can use the “show by” drop-down menu to compare segments, too: compare demographics, revenue by location, channel, UTM campaigns and more.

This graph uses cohorts segmented by channel to compare purchases.

How can I create cohorts in Facebook Analytics?

You'll see “Cohorts” listed in the lefthand menu when you log in.

Looking for more info on how to create your own? Check out the cohort section in our Help Center.