What cohorts tells you
Cohort analysis groups scanners by when they first scanned, such as the week of first scan, then follows each group over time. It reveals whether newer scanners behave differently from earlier ones, which a single total hides.Cohorts make change visible. If each new cohort retains worse than the last, something in the experience is getting worse.
How to compute cohorts
Cohort retention = members of a cohort who return within a window divided by the cohort size, multiplied by 100.Illustrative example numbers, not real-world data: a January cohort of 200 scanners has 60 return within 30 days, giving 60 / 200 x 100 = 30% cohort retention. A February cohort of 200 scanners with 40 returning gives 20%, so February scanners retained less well in this example.
How to collect the data
Define cohorts by first-scan week, fix the follow-up window, and compare cohorts at the same age, not at the same calendar date.Keep the raw counts, not just the final percentage. Raw counts let you recompute the metric if you later fix the definition, and they make it possible to spot a denominator problem, such as a week where exposures were estimated instead of counted. Store the counts in one place with the date range, the placements included, and the name of the person who pulled them. That small discipline turns a spreadsheet into an audit trail.
Reading the result
Never judge cohorts as a single number. Break it down by first-scan week, placement, and offer. A campaign that looks average overall may be strong in one segment and weak in another, and the fix is completely different in each case. Compare the same placement against its own past, and compare different placements only when their audiences and contexts are genuinely similar. A short trend over four or more comparable periods is worth far more than a single snapshot, because it separates a real shift from weekly noise.Also check the relationship between cohorts and the metrics around it. If scans rise while conversions fall, the new scanners may be less interested visitors, or the landing page may have changed. If cost rises while cohorts stays flat, you are paying more for the same outcome. Metrics make sense in groups, not in isolation, and the habit of reading two or three together prevents most misreadings.