What is cohort analysis in email marketing?

Cohort analysis groups subscribers by something they have in common, then tracks how each group behaves over time. It is one of the most useful tools for understanding real list health, as opposed to the snapshot view your average campaign report gives you.

The most common cohort: acquisition date. Take everyone who signed up in January 2024, track their open and click rates month by month over the next year. Do the same for February, March, and so on. Now you can see whether January signups are more engaged at 12 months than August signups were. You can spot acquisition quality changes that aggregate stats would bury.

Other useful cohort types:

• Acquisition source: did subscribers from your lead magnet retain engagement longer than subscribers from a giveaway?
• First-purchase cohort: do subscribers who bought within 30 days of signup stay more engaged than those who did not?
• Re-engagement cohort: did the subscribers who responded to your win-back campaign stay active, or did they go cold again within 60 days?

Cohort analysis requires per-recipient event data, not just campaign-level aggregates. You need to know what each individual subscriber did over time, not just what your average open rate was last month. This is why webhook event feeds matter: if your ESP only gives you 30 days of raw data, long-term cohort analysis is not possible without capturing events yourself.

The practical starting point: most ESPs let you filter campaign results by signup date. That is your simplest cohort. If January signups have a 35% click rate at 6 months and December signups have 12%, something changed between those acquisition periods worth investigating.

For how AI can help predict which cohorts are at churn risk, that guide covers predictive scoring approaches you can layer on top of cohort segmentation.