Activation Method

From read-data-schema, list early actions a new user could take in their first session/week: core feature uses, setup steps, collaboration actions. Include count-based candidates (“did X ≥ 3 times”), not just “did X once”.

2. Retention lift

For each candidate, split new users into did it early vs didn’t, and compare their retention N weeks later (use modeling-product-usage-metrics). Lift = retention(did) − retention(didn’t). A good activation candidate has large, stable lift.

Beware confounds: an action can correlate with retention without being causal (e.g. “visited settings” might just mark already-engaged users). Prefer actions that plausibly deliver the product’s value.

3. Reach × predictive power

Score each candidate (or combination) on two axes:

• Reach — share of new users who hit it within the window. Too low and the metric describes a tiny elite.
• Predictive power — the retention lift above.

Pick the definition that keeps reach acceptable (rule of thumb: a large minority to a majority of good-fit signups can hit it) while maximizing lift. Combinations (“created a dashboard AND invited a teammate”) and thresholds (“≥3 queries”) usually beat any single one-time action.

4. Definition shape

The output of this method is a concrete, testable definition:

A user is activated if, within N days of their first event, they [criteria].

Examples: “within 7 days, ran ≥3 queries”; “within 14 days, created a project and invited ≥1 teammate”. For B2B, apply the same at the account grain: the account activates when any user meets the criteria.

Record N and the criteria explicitly — they are the model’s parameters, and they’ll be revisited as the product changes. If a governed activation metric already exists in the semantic layer (see foundations governance.md), reuse it instead of inventing a new one.