Frequently asked questions
Everything you need to know about AppVerdict. See pricing →
How it works
What is AppVerdict?
AppVerdict is an AI-powered app research tool. You search any app, and we scrape hundreds of real user reviews from both the App Store and Google Play, then use Claude AI to analyse them and generate a structured report — including what users love, what frustrates them, what they're asking for, and a business opportunity brief for anyone considering building a competitor.
How long does a report take?
Typically 30–60 seconds. We scrape up to 500 reviews from each store simultaneously, then send them to Claude AI for analysis. The first time you run a report for an app it takes the full time. Repeat searches within 24 hours are served instantly from cache.
How many reviews are analysed?
We collect up to 500 reviews from the App Store (sourced from international stores that serve reviews reliably) and up to 500 from Google Play — up to 1,000 reviews total per report. Claude AI analyses up to 300 of the most relevant reviews to generate the insights.
How fresh is the data?
Reviews are fetched live when you run a report. Reports are cached for 24 hours — if someone else has already run a report for the same app today, you'll get the cached version instantly. After 24 hours, the next request fetches fresh reviews.
Scoring methodology
How is the Sentiment Score (1–10) calculated?
The Sentiment Score is generated by Claude AI after reading all analysed reviews. It reflects the overall tone and balance of user opinion — not just the average star rating. A score of 1–3 is Negative, 4–5 is Mixed, 6–7 is Positive, and 8–10 is very Positive. Claude weighs the severity of complaints against the frequency of praise to arrive at this number.
How is the Opportunity Score (1–10) calculated?
The Opportunity Score measures how attractive a gap in the market is for an indie developer. It's a weighted score across five factors:
What do the five Opportunity Score factors mean?
• Gap Clarity (25%) — How clearly defined and consistent the unmet need is across reviews. High = many users mention the same pain point. • Achievability (25%) — Can a small team + AI tools realistically build this in under 6 months? High = realistic for an indie developer. • Market Timing (20%) — Is the competitor losing users right now? Low sentiment on the original app = better timing for a challenger. • Competition Size (15%) — Market size vs. difficulty of winning. Medium-large competitors score best — big enough market, but beatable. • Risk Level (15%, inverted) — High here means low risk: few technical, legal, or monetization obstacles. Score guide: 1–3 = Weak, 4–5 = Moderate, 6–7 = Strong, 8–10 = Exceptional.
Are the mention counts exact?
No — mention counts (e.g. '~45 mentions') are Claude's best estimates based on pattern recognition across the reviews, not exact keyword counts. They're directional signals to help you prioritise, not precise analytics.
Plans & billing
What's included in the free plan?
The free plan gives you one lifetime report with the full review summary — Verdict, What Users Love, Top Feature Requests, and Biggest Concerns. The Build Opportunity section is always locked for free users.
What's the difference between Pay-per-report and Pro?
Pay-per-report (€3.99) unlocks the full report including Build Opportunity for a single app — no subscription, no history saved. Pro (€29.99/month) gives you 20 reports per month, PDF export, and report history. If you research apps regularly, Pro pays for itself after 8 reports.
Can I cancel my subscription?
Yes, anytime. Go to your account settings (click your avatar in the top right) and manage your subscription. You keep access until the end of your billing period.
Is my data stored?
Reports are cached for 24 hours to avoid repeat API costs. We don't store your personal browsing history or report history (unless you're on Pro/Agency, where history is saved to your account). We never sell your data.
Technical
Where do the reviews come from?
App Store reviews are fetched directly via Apple's RSS feed from multiple international stores. Google Play reviews are fetched via the google-play-scraper library which hits Google Play directly. No third-party paid APIs are used for review collection.
Which AI model powers the analysis?
AppVerdict uses Claude Sonnet by Anthropic — one of the most capable AI models available. We chose Claude specifically for its ability to follow complex structured output instructions reliably and its strong performance on analytical tasks.
Why do iOS review counts vary between reports?
Apple's RSS feed is inconsistent — some pages return 50 reviews, others return 0, unpredictably. We compensate by fetching from multiple country stores simultaneously, but the total count can vary between 50 and 500 depending on Apple's API availability at that moment.
Still have questions? hello@appverdict.io