What is Cohort Size?

Why It Matters

Cohort size matters because it's the denominator behind every cohort retention or cohort revenue percentage, and a percentage without knowing its underlying sample size can be dangerously misleading. It matters because in a small cohort, a single user churning or staying swings the retention percentage by a large amount, while the same single user is statistically invisible in a large cohort, so two cohorts with identical retention percentages can carry very different levels of confidence. Ignoring cohort size means a team might overreact to a dramatic-looking retention swing in a slow signup month that's really just noise from a handful of users, or underreact to a genuinely meaningful shift buried in a large cohort. It matters for planning too, since seasonal or channel-driven differences in cohort size affect how much weight leadership should put on any single month's retention story. Analysts typically report cohort size alongside retention percentage specifically so readers can judge how much statistical confidence to place in the number.

Example

A product's January signup cohort includes 4,500 users, while its typically quieter August cohort includes only 180 users. A single user churning or staying in the August cohort swings that cohort's retention rate by more than half a percentage point, while the same single user barely moves the January cohort's retention rate at all, which is why analysts treat retention percentages from very small cohorts with more caution than those from large ones.

Frequently Asked Questions

• How small is too small for a reliable cohort?

  There's no fixed universal cutoff, but cohorts in the low hundreds or fewer start producing retention percentages that swing significantly from just a few users, so results from them should be treated as directional rather than precise.

• How does cohort size affect cohort retention rate?

  Cohort size is the denominator in the cohort retention rate calculation, so a small cohort size means each individual user represents a larger share of the resulting percentage, making the metric noisier and more volatile.

• What typically causes cohort size to vary month to month?

  Seasonality, marketing campaign timing, and shifts in acquisition channel spend are the most common reasons a signup-based cohort is larger in some periods and smaller in others.

• How often is cohort size tracked?

  It's tracked whenever a new cohort is defined, typically on the same weekly or monthly cadence used for cohort retention reporting, since the two are always reported together.

• Should small cohorts be excluded from analysis entirely?

  Not necessarily excluded, but they're usually flagged or weighted differently, since combining several small cohorts into a larger grouped cohort can produce more statistically reliable retention trends than analyzing each small cohort alone.