Blog
Notes from experiment reviews.
Sample-size myths that waste mobile sprints
Why “two weeks of traffic” is not a power calculation — and what to write instead.
Reading confidence intervals with product partners
A plain-language script for when the interval straddles zero and marketing wants a headline.
Experimentation habits for Thailand app growth teams
Store lag, bilingual UX tests, and evening critique windows that match Bangkok rhythms.
Guardrail metrics that actually save releases
How to pick health metrics that can veto a “winning” variant without becoming noise.