We need to understand what users are doing in today's onboarding — where they spend time, where they drop off — before we can know if a new flow is actually better.
Today we don't have a clear picture of what's tracked. The work is to map what we have, identify gaps, fill them, and propose a small metric set we can use for any future A/B comparison. AARRR activation framing applies.
Scope
- Inventory what's tracked today. Walk every step of the current onboarding (signup → org → project → feature → first SDK eval) and produce a short map of what's emitted, where it lands, and what's queryable from it. We don't have this map today; building it is part of the ticket.
- Identify the gaps. From that inventory, decide what's missing to compute activation rate and stage-by-stage drop-off end-to-end.
- Fill the gaps. Emit new events for the missing steps. Confirm everything lands somewhere queryable.
Once events are flowing, propose a small metric set (2–3 metrics, anchored on AARRR activation) and document it.
Out of scope
- Instrumenting the new onboarding flow — separate ticket once this lands.
- Computing baselines from the data — separate ticket once the metric set is agreed.
Done when
- A short audit doc lists what's tracked today and which events needed filling.
- Every step of the current onboarding emits a tracked event, queryable in the sink.
- A proposed 2–3 metric set is documented.
We need to understand what users are doing in today's onboarding — where they spend time, where they drop off — before we can know if a new flow is actually better.
Today we don't have a clear picture of what's tracked. The work is to map what we have, identify gaps, fill them, and propose a small metric set we can use for any future A/B comparison. AARRR activation framing applies.
Scope
Once events are flowing, propose a small metric set (2–3 metrics, anchored on AARRR activation) and document it.
Out of scope
Done when