LaunchLens is a Chrome extension that starts app store optimisation before the app exists. It reads competitors' store pages and reviews, helps you decide what to build and how to position it, and then writes the listing.
How it works
- Research. Add competitors straight from their App Store or Google Play pages: metadata, reviews and version history.
- Insights. Review themes, extracted features, patterns across competitors and keyword clusters. You mark what's a must-have.
- Product. An editable product definition: name, one-liner, problem, audience, core features with evidence, differentiators and anti-features.
- Screens. UI screens generated with Google Stitch.
- Listings. Three listing strategies for both stores, scored for keywords, readability, uniqueness, conversion, policy safety and similarity risk.
- Export. A PRD, a screens document and listings, as Markdown, JSON or CSV.
How the AI is used
- Bring your own key: OpenAI, Anthropic, Google or OpenRouter.
- Fourteen AI tasks, each set to a fast or a powerful model tier.
- Structured output everywhere, validated against schemas before it's saved.
- Every run is reproducible. Each call is logged with its prompt version, inputs, outputs and token usage.
Decisions
Why make a hidden step visible? #
Listings used to be generated through a positioning step the user never saw. Regenerate, and the positioning shifted underneath, so the listings drifted. I replaced the hidden step with the editable product record in step 3, and the drift stopped. It's one more thing to read, but it's the step that was deciding everything anyway.
Why no backend? #
Your research stays in your browser, in IndexedDB. Pages are read only when you ask.
The numbers
- 74 commits in five weeks, from 27 April to 30 May 2026.