These notes reference programmes and modules taught here. They are not a star wall. One reservation sits next to the praise on purpose.
“I came for keyword folklore and left with a telemetry map. The ASO Measurement Studio module on page-view share versus impression share is now the first argument we have in sprint planning. Finance stopped asking for ‘ASO ROI’ as a single number.”
“Screenshot Experiment Clinic taught me to write a stop rule before the designer exports frames. We killed a Songkran variant on day nine instead of letting it run because someone liked the colour. Conversion was flat; we would have called it a win last year.”
■■■■■■■■□□ 8/10
Listing narrative workshop was tighter than expected. The preview-video session felt thin if you ship mainly on Play. Still, the store-promise versus onboarding checklist paid for the seat in one release.
“Keyword Collision Lab flagged that our English subtitle was eating the Thai query we actually wanted. Mild reservation: the stemming lecture assumed more NLP comfort than our intern had. We paired them with the checklist and it held.”
Catalogue Residency sat inside our freeze calendar rather than inventing a classroom title. The cannibalization note on two sibling games is still taped above the ops desk. I would not book it if you only have one live app; the desk time would feel oversized.
Case · fintech wallet, Thai iOS
Conversion rose; D1 told on us
After cohort 9, a wallet team shipped a screenshot set that pushed bill-pay into frame one. Product-page conversion moved from 3.1% to 3.8% over fourteen days. D1 retention slipped. The measurement spec from ASO Measurement Studio already had a dual gate: conversion and D1 both had to hold. They rolled back frame one, kept a quieter bill-pay frame in position three, and wrote the episode into the next board pack instead of burying it.
The useful part was not the lift. It was the pre-registered stop rule. Without it, the screenshot would still be live.
Case · two casual games, Play + iOS
Sibling titles, same English subtitle
A publisher in the Catalogue Residency discovered both games used a near-identical English subtitle chasing “puzzle offline.” Thai metadata diverged; English did not. Query share on the newer title rose while the older title’s page views fell in the same cluster. The collision lab checklist made the overlap visible in one sitting. They froze English subtitles for six weeks, differentiated the older title toward a mode it actually owned, and stopped treating the dip as a seasonality story.
No rank miracle followed. Impression quality simply stopped leaking between siblings. That was the brief.
