Chapter 7 · Product Analyst
7. Segmentation & understanding your users
~6 min read
Aggregate metrics describe an average user who often doesn't exist. The real insight lives in the differences between groups. Segmentation turns 'engagement is down' into 'engagement is down for new Android users in Brazil since the 4.2 release', a problem someone can actually fix.
7.1 The dimensions that matter#
| Segment by… | Reveals | Example question |
|---|---|---|
| Acquisition channel | Which sources bring quality users | Do paid users retain like organic? |
| Platform / device | Platform-specific bugs & UX | Is checkout broken on Android? |
| Geography / locale | Market and localization issues | Why is APAC activation lower? |
| New vs. returning | Onboarding vs. habit problems | Is the drop in first-timers? |
| Behavioral / power users | Who your most valuable users are | What do retained users do early? |
7.2 The 'aha moment'#
The classic application of behavioral segmentation: find the early action that predicts long-term retention. Compare what retained users did early against churners and find the action with the sharpest separation.
7.3 RFM segmentation#
RFM (Recency, Frequency, Monetary) cleanly separates champions from at-risk users so retention efforts are targeted rather than sprayed.
7.4 Combine quant with qual#
Numbers tell you what is happening; they rarely tell you why. Pair behavioral data with session replays, support tickets, app-store reviews, and user interviews. A drop in a funnel step is a fact; a session replay showing users confused by a new button is the explanation that drives the fix.
Get the next chapter and weekly interview tips by email
One short email per week. Skim in a minute. Unsubscribe anytime.