How a review works
The engagement measurement review is the shape most visitors need. Other sessions borrow pieces of it. Nothing here is a hosted product; it is a sequence of conversations and pages.
1. Intake
We ask which app, which stores, who owns the event list, and which question about staying you actually need settled. If the question is “make the numbers go up,” we will send you back to product work. If the question is “does completed_onboarding still mean a person finished the last step,” we can begin.
2. Inventory
Aisha prints or spreads the list. Every event that claims to describe engagement gets a line: trigger, screen, properties, last known owner. Names that fire on a preview as well as a confirm are marked in the margin. This week is quiet on purpose.
3. Session sample
Wei Jie watches a bounded sample you already have permission to share. Thirty sessions, or three days of traces, is enough to see whether the inventory’s story survives contact with a thumb. We pause on screens where the event name and the visible task disagree.
4. Findings
The note is written for the people in the readout, not for a general audience. It quotes screen titles, event names, and one or two anonymised paths. We state what the current set can support (for example, “we can say how many people opened Settings”) and what it cannot (for example, “we cannot say they compared two plans”).
5. Readout
Ninety minutes. Priya joins when naming is the fight. You leave with the marked inventory, the note, and a short list of recording changes — not a backlog of product ideas unless you explicitly ask us to stay in that conversation.
If this sequence matches the work you need, request a measurement review and name the last release that changed navigation.