Engagement
This is the sequence we use for parser event validation. Taxonomy cleanup and tracking plan reconstruction add extra weeks; they do not skip the sample rules.
01 · Intake
Farah Idris collects sources, date range, and the five reports you cannot afford to misread. We ask whether debug builds live in the same project as production. If they do, the export must be split before the pass starts. You do not need a finished tracking plan; you do need to say which conversion the board currently believes.
02 · Sample
CSV or JSON lines, timestamps, event name, properties, and a field that distinguishes iOS, Android, web, and any OEM build. Fourteen to thirty days is enough for most catalogues. Mixed files delay the calendar. We refuse samples widened with personal identifiers beyond what the analytics events already carry.
03 · Validation pass
Harith Osman reads high-traffic events for type, required properties, allowed values, and whether the name still matches the screen. Wei Liang Tan flags duplicates that will matter if you later ask for cleanup. Amina Rashid marks which mismatches actually move the reports named at intake. The specimen on the home page is the tone of that list: short, dated, sourced.
04 · Findings workshop
You receive the findings the working day before the call. The hour is for ranking the first ten fixes and deciding whether taxonomy cleanup should follow. We do not use the hour to walk through every unused event in the appendix. If both squads need to hear the same recap, book the handover session as a separate half-day.
Reading happens in Seremban. Calls are online unless you prefer the Tingkat 1 room. We do not join your release train and we do not keep a standing seat in the vendor console. If the catalogue moves after the date range, that is a new sample, not an informal extension of the same week.
Bring the export, the screens, and the five reports. Leave the wish for a single score at the door.