In plain English
Usually people who first arrived in the same week or month, tracked forward to see how many return. It is the standard way of measuring retention.
Cohort analysis answers a question no other report does: whether the people you acquired recently behave better or worse than those you acquired last year.
What to know
Grouped by start point
Usually acquisition date.
Tracked forward
Return behaviour over weeks.
Reveals retention trends
Improving or deteriorating.
Comparable across cohorts
Which is the point.
Why it matters
If recent cohorts retain worse than older ones, something changed in the product, the onboarding or the traffic source. No other report surfaces that.
Common mistakes
×Never running cohort analysis.
×Cohorts too small to read.
×Comparing cohorts of different lengths.
×Looking only at aggregate retention.
FAQs
What does cohort analysis tell me?
Whether retention is improving or deteriorating over time.
How large should a cohort be?
Large enough that the return percentages are not noise.
WRITTEN BY TARIQ SALLAM
Marketing Consultant. Entrepreneur. Content Creator.
I'm a marketing consultant, entrepreneur and content creator. I help businesses grow through practical marketing, websites, SEO, content and AI.
More About Tariq →