recipes / mine-app-store-reviews
Mine App Store reviews with AI — complaints, praise, keywords
Reviews are the one place users say, in their own words, what your app is for — and their words are search queries. This prompt mines recent reviews for complaint themes, praise themes, and the vocabulary your metadata doesn't cover yet.
the prompt
Read the recent US App Store reviews for <your app> and for <competitor>. 1. Top complaint themes for each, with a representative quote per theme. 2. What do people love — the job they actually hired the app for? 3. Which words do reviewers use naturally that our visible metadata doesn't contain — and which of those are worth targeting?
what your agent does
- 01
Fetches recent reviews per storefront — ratings, titles, bodies — for any public app, yours or a competitor's.
- 02
Pulls your live listing so "words we don't cover" is computed against the real visible pool, not memory.
- 03
Scores the extracted vocabulary — reviewer language is only worth targeting when people actually search it.
- 04
Clusters themes with quotes attached; run on a competitor, the complaint list doubles as your positioning brief.
what you get back
Three lists you can act on: complaints (your roadmap), praise (your positioning), and reviewer vocabulary with demand scores (your metadata candidates) — each with quotes attached to keep everyone honest.
reviewers keep saying in your pool? popularity
"streak saver" no 31 · proxy
"home screen widget" no 54 · apple
"morning routine" partly 27 · applemake it yours
- Run it monthly and diff the themes — reviews are the fastest-moving public signal an app has.
- Mine only the 1–2 star reviews of the category leader — that's a positioning brief.
- Have the agent draft replies grouped by theme (pasting them into App Store Connect stays your job).