Motivation Segmentation

Behavioural telemetry tells you what changed. It cannot tell you why, because "why" is not a behaviour — it is the reason behind one. Two users with identical usage curves can be here for opposite reasons and need opposite products. This skill is about collecting a declared motive, joining it to everything, and staying honest about the ways both self-report and instrumentation mislead.

Duolingo is the case study throughout: it asks every new learner for a primary reason and publishes the cross-tabs. Your product is the target.

Collecting the field

• [[references/collect-a-declared-why-at-signup]] — one question at signup turns a correlation into an explanation; plus the two known failure modes of self-report.
• [[references/size-of-other-is-a-taxonomy-diagnostic]] — a catch-all that spikes for one segment is measuring how wrong your category list is there.

Reading it without fooling yourself

• [[references/largest-segment-is-not-most-engaged]] — the modal reason routinely spends less time in product than a smaller one, so headcount is a bad build order.
• [[references/split-by-cohort-before-concluding-what-users-want]] — the aggregate top row is consensus theatre; the inversions beneath it are where differentiated decisions live.
• [[references/motive-decides-the-lever-and-who-to-exclude]] — match the intervention to the motive, and exclude segments whose causal story is structurally different rather than pooling them.

When the mix changes