Cohort analysis
DEFINITION
Cohort analysis groups subscribers who share a common starting point, most often signup month, and tracks how that group's behavior changes over time rather than relying only on blended company-wide averages. It reveals retention trends and the effects of pricing, onboarding, or channel changes that aggregate metrics can hide.
TABLE OF CONTENTS
RELATED TERMS
Cohort analysis is the practice of grouping subscribers who share a common starting point, most often the month they signed up, and tracking how that group behaves over time rather than looking only at blended, company-wide averages. Instead of asking "how many customers do we have active right now," cohort analysis asks "of the customers who joined in a given period, what share are still active one month later, two months later, six months later," which makes it possible to see how retention actually changes as a group of customers ages.
Cohorts do not have to be defined by signup month alone. A subscription business can also group customers by acquisition channel, plan tier, pricing version, or the offer or coupon they signed up under, then compare how each group performs over time. A subscription platform such as Recurly sits on the billing and subscription data needed to build these groupings, since every subscriber record already carries a signup date, a plan, and often an acquisition source.
Why cohort analysis matters for subscription businesses
Aggregate metrics like a single blended churn rate or a single average revenue per account can hide what is actually happening underneath. A business can have a stable overall churn rate while its newest cohorts are retaining far worse than older ones, or its oldest cohorts are aging out and being replaced by less loyal new customers. Cohort analysis surfaces that detail. It lets a team see whether retention is improving or declining over time, whether a specific pricing or onboarding change helped or hurt, and whether one acquisition channel brings in customers who stick around longer than another. Because subscription revenue compounds over the life of a customer, small differences in early-month retention between cohorts can translate into meaningfully different lifetime value, which is why retention and finance teams treat cohort views as a standard part of subscription reporting.
Common cohort dimensions
Signup period: customers grouped by the week, month, or quarter they became a paying subscriber. This is the most common cohort definition.
Acquisition channel: customers grouped by how they were acquired, such as paid search, organic, referral, or partner channel.
Plan or pricing tier: customers grouped by the plan or price point they originally signed up under, useful for comparing how different offers retain.
Campaign or offer: customers grouped by a specific promotion, discount, or coupon they redeemed at signup.
How cohort analysis works
Choose a cohort dimension, most commonly the signup month, and group all subscribers who share that starting point into a single cohort.
Define the metric to track over time, such as the percentage of the cohort still active, average revenue per account, or upgrade and downgrade activity.
Measure that metric at fixed intervals after the starting point, for example at month 0, month 1, month 2, and so on, for every cohort.
Arrange the results in a cohort view, with each row representing one cohort and each column representing a period since signup, so that reading across a row shows how a single cohort decays or grows over time and reading down a column compares different cohorts at the same lifecycle stage.
Plot a single row as a line to produce a retention curve, and compare curves across cohorts to spot whether retention is improving, worsening, or holding steady.
Benefits and examples
Cohort analysis gives a subscription business a way to isolate cause and effect that a single blended metric cannot provide.
Spotting retention trends early: comparing recent cohorts to older ones shows whether newly acquired customers are retaining better or worse, often months before that shift would show up in a company-wide churn number.
Evaluating changes before and after: comparing cohorts that signed up before and after a pricing change, a new onboarding flow, or a billing process change isolates the effect of that specific change from normal month-to-month noise.
Comparing acquisition channels: cohorts split by channel show which sources bring in customers who stay longer, which can redirect acquisition spend toward higher retaining channels.
Informing lifetime value estimates: because cohort views track the same group over its full lifespan, they give a more accurate basis for projecting lifetime value than a single point-in-time average.
The following is a simplified, hypothetical example to illustrate the mechanics, not a real-world benchmark. Imagine a subscription business signs up 200 new customers in January and tracks how many of those same 200 customers remain active in each month after signup:
Month 0: 200 customers active, 100% retention.
Month 1: 150 customers active, 75% retention.
Month 2: 120 customers active, 60% retention.
Month 3: 108 customers active, 54% retention.
Each retention rate is simply the customers still active divided by the original cohort size: 150 / 200 = 75%, 120 / 200 = 60%, and 108 / 200 = 54%. Plotted as a line, these four points form the January cohort's retention curve. Repeating this same exercise for the February, March, and later cohorts, then comparing the curves side by side, would show whether retention is trending better or worse over time.
Frequently asked questions
What is a cohort in subscription analytics? A cohort is a group of subscribers who share a common starting point, most often the period in which they signed up, though cohorts can also be defined by acquisition channel, plan tier, or the offer they redeemed.
How is cohort analysis different from a regular churn report? A regular churn report typically shows a single blended rate across the whole customer base in a given period, while cohort analysis follows the same group of customers over multiple periods, showing how that specific group's behavior changes as it ages.
What is a retention curve? A retention curve is a plot of the percentage of a single cohort that remains active at each period after signup, such as month 0, month 1, and month 2, and it visually shows how quickly a cohort typically loses customers and where that decline levels off.
Can cohorts be grouped by something other than signup date? Yes. Common alternative groupings include acquisition channel, plan or pricing tier, and the specific campaign or coupon a customer signed up under.
Does Recurly support cohort analysis directly? Yes, Recurly offers cohort analysis and retention curve reports within its analytics capabilities. Recurly's analytics include:
Cohort Analysis Analytics: This feature groups subscribers by their initial purchase date and tracks their activity or subscription status over subsequent periods. It allows you to visualize retention curves, helping to identify retention drivers, compare acquisition channels, and spot churn trends early. You can toggle between contract retention rates and raw subscription counts.
Subscriber Retention Reports: These reports allow you to analyze and evaluate the retention and churn rates of paying subscribers over a specified timeframe using cohort analysis. This helps in understanding when and why subscribers churn, enabling the development of effective retention strategies. The Subscriber Retention dashboard often includes a Subscriber Cohort Descent line chart and detailed cohort analysis.
These features are available in both Recurly Commerce and Recurly Subscriptions analytics.
