A rising line of total customers is the most reassuring chart in a board pack and one of the least informative. It mixes customers who joined last month with customers who joined three years ago, and it hides how many are leaving. Strong acquisition can cover weak retention for a long time, until marketing becomes too expensive to keep filling a leaking bucket.

Read the business one generation at a time

A cohort is a group of customers who started in the same month. A cohort table follows each group over time and shows what share is still active after one, two, three months and beyond. Each row is a generation of customers; each column is their age. Read down a column and you see whether newer customers behave better than older ones.

Share of each monthly cohort still active, by months since it joined. Darker is better. Retention improves from the April cohort, after an onboarding change. Illustrative data.

Three patterns to look for

  • The first-month drop: the biggest loss usually happens right after the first purchase or the first invoice. It is a product and onboarding problem, not a marketing one.
  • The floor: healthy businesses see curves that flatten. A curve that keeps falling to zero means there is no lasting core of loyal customers.
  • Newer is better: if recent cohorts sit above older ones at the same age, the improvements are working. If they sit below, growth is being bought with lower-quality customers.

Making it a management routine

Ask for the cohort view every month, next to revenue. Agree on one definition of an active customer, keep it stable and segment the table by channel, product or region when a pattern needs explaining. Above all, tie every initiative, from a new onboarding flow to a price change, to the cohort where it started, so its effect can be seen rather than argued.

BS

Bartor S.

Founder

All insights