How a study runs
How a retention study runs
The weeks are boring on purpose. Access, tables, interviews, then a workshop — so the brief is about people who stopped coming back, not about a new set of charts.
A study only starts when a counterpart can answer “what counts as a return?” in a sentence. If that sentence does not exist, week 0 is spent writing it down, not drawing conclusions.
Files stay on an agreed share. We do not copy event exports onto personal laptops for convenience. Interview recordings are deleted after the notes are agreed, unless your legal team requires a longer hold.
The workshop is the point at which the product team is allowed to disagree. We would rather lose an argument in the room than watch a brief sit unread. If you already know you only want a marked-up table, book a cohort reading session instead of the full study.
Access and scope
We agree which product, which countries, and which event names we may read. You name a single counterpart who can answer data questions within a day.
Cohort tables and first cuts
We rebuild weekly and monthly return tables from the export, mark holidays that distort Malaysia traffic, and list the drop-offs that look structural rather than seasonal.
Conversations with people who left
We recruit from your lapsed list (with your consent) and hold short interviews. Notes stay anonymised in the brief.
Findings workshop
A half-day session in Kuala Lumpur or on your floor. We walk through three to five retention problems, not a catalogue of every chart.