Field notes

Think more clearly about repeat use.

Practical notes for product, growth and data teams making retention decisions in Malaysia’s app economy.

Business analytics dashboard on a laptop

Analytics practice · 12 August 2026

How to read a retention curve without fooling yourself

The shape matters more than the headline number. Learn what a flattening curve, delayed drop, and cohort gap can tell you.

A useful retention curve begins with a precise return behaviour. A login, purchase, completed workflow and active subscription tell different stories. Compare cohorts only after the event and observation window are stable. Then look at the curve’s shape: the early slope points to activation friction, while the eventual floor suggests whether a durable habit or recurring need exists.

Market insight · 29 July 2026

Building useful cohorts for Malaysia's mobile-first customers

Language, payment habits, acquisition channels, and festive periods can all distort an aggregate retention view.

National averages hide useful differences. Separate cohorts when language, platform, acquisition channel or seasonal periods plausibly change behaviour. In Malaysia, Ramadan and Hari Raya periods can shift both acquisition volume and normal usage rhythms. Treat that context as part of the analysis, not noise to smooth away.

Colleagues discussing notes on a glass wall

Growth decisions · 8 July 2026

Fix activation before you chase long-term retention

A practical way to distinguish an onboarding problem from a weak recurring-value problem before investing in experiments.

If most users leave before experiencing the product’s core value, long-horizon messaging will not repair the curve. First identify the earliest behaviour strongly associated with return, validate that it reflects value rather than simple activity, and shorten the path towards it. Only then test recurring engagement loops.