Case study · App review analysis for product teams

ReviewStack

Paste a Play Store, App Store or Chrome Web Store listing and get every review scored for sentiment, sorted into categories, and summarised by AI into what users love, what is broken, and what to build next. Free to use, with an SEO program built into the product from day one.

ReviewStack homepage
Role
Product design, engineering, SEO and content
Status
Live · free tier
Stack
  • Next.js
  • React
  • TypeScript
  • Hono
  • Cloudflare Workers
  • Cloudflare KV
  • Groq
  • Resend
  • Cloudinary
  • Vercel
The problem

Why it needed to exist.

Thousands of reviews sit under every app, and almost nobody reads past the first page. The signal is in there: the crash that started with the last release, the feature request that keeps recurring, the competitor people keep mentioning.

Existing tools are priced for enterprise teams. We wanted the analysis to take seconds, cost nothing to try, and work for a solo developer as well as a product team.

What we built

The product, feature by feature

  • Three stores, one search

    Search by name or paste a URL for Play Store, App Store (per country storefront) or Chrome extensions; Google Business is next.

  • Sentiment and categorisation

    Every review scored positive, negative or neutral and auto-sorted into bugs, crashes, feature requests and six more categories.

  • AI insights

    An executive summary, top issues and clear next steps generated by Groq on top of the classified reviews.

  • Compare and watch

    Side-by-side comparison of up to four apps with a shareable URL, plus a watchlist and history on the dashboard.

  • Roadmap with voting

    A public roadmap where signed-in users vote and comment on what gets built next.

  • Search-first content

    Dedicated landing pages per store and a fourteen-article blog on review analysis, planned from a keyword map before launch.

ReviewStack homepage
ReviewStack on a phone
How it's built

Engineering decisions worth mentioning

  • Scrape and analyse at the edge

    Review fetching and analysis run as a Hono API on Cloudflare Workers with KV caching and a cron, so repeat lookups are instant.

  • Rate limits on every expensive route

    Per-IP limits on scraping, analysis and AI routes keep the free tier free without a credit card wall.

  • Passwordless sign-in

    OTP by email through Resend, with redirects back to the page the user was trying to reach.

  • SEO as a build requirement

    Store-specific landing pages, structured data and a content plan shipped with the product, not after it.

  • 3

    review sources analysed: Play Store, App Store, Chrome Web Store

  • 9

    review categories, from crashes to feature requests

  • 4

    apps compared side by side from one shareable link

  • 14

    long-form articles published as part of the launch SEO program

Building something like this?

This is the same team, stack and cadence you'd get. Tell us what you're making and we'll scope it honestly.