Averages can conceal the groups most affected by a campaign. Before comparing results, ask whether the cohorts represent similar opportunities to respond.
Five questions
- Were customers eligible for the same reason and at a comparable lifecycle moment?
- Did each cohort have a similar history of activity before the message?
- Could seasonality, promotions or channel changes explain the difference?
- Is the observation window long enough for the expected behaviour?
- Are short-term gains offset by opt-outs, fatigue or weaker later retention?
Use the answers to shape the claim
If groups differ materially, segment the analysis or adjust the comparison. When adjustment is not credible, present the pattern as directional evidence and propose a stronger test.
Need a view of your own lifecycle evidence?
We can help map the journey and identify the measurements most worth improving.
Talk through your measurement map