10 markets, scored by how much their users are suffering
Every number below comes from 6,328 real App Store reviews across 43 competitor apps. We do not republish review text — we publish aggregated complaint structure, recurring phrases with counts, and payment evidence mined from Hacker News.
Why these numbers differ from our live pages. This report line is a frozen snapshot — 10 markets scanned once, on 2026-09-14, and left as they were. Our category insight and methodology pages run off a different, live pool: 3 app groups, 46 apps, 18,145 reviews, recomputed every rebuild. So the two lines print different totals and neither is wrong — they are two corpora answering two questions. For a specific market's complaint structure, trust this page; for what is current this week, trust the live pages. We would rather explain the gap than hide it.
Two percentages, two different questions. The complaint rate on each card is the share of all App Store ratings that are 1–2 star, taken from Apple's published star distribution. The category breakdown inside each report comes from the reviews we scanned, which are drawn from Apple's own review ordering (it skews critical but not uniformly across apps) — so its internal proportions are far higher by construction and are not the complaint rate. We used to publish the sample figure as if it were the rate. It was wrong by roughly 10×, and every number here has been restated.
Read this before you read the list. A high complaint rate is necessary but not sufficient. Cleaning Services scores highest, but Auto Repair is the one we'd actively warn you off — not because pain is low there alone, but because pain is low and incumbents carry 25,783 ratings. The verdict inside each report combines all three dimensions.
How we score
Three dimensions, not one
01
Complaint rate
Share of all ratings on the App Store that are 1–2 star, taken from Apple's published star distribution — not from the reviews we scanned. High means real, sustained frustration — users stay and suffer instead of churning.
01b
Why not the scanned reviews
Our crawler pulls the most negative reviews first, so any percentage computed inside that sample is inflated by construction — on this data by roughly 10×. We used to publish that number. It was wrong, and every category page here has been restated.
02
Incumbent strength
Median total ratings across competitors. Under ~1,500 means almost no review moat. Over ~20,000 means you cannot win head-on.
03
Payment evidence
Documented first-hand revenue reports from public founder posts. If we can't verify anyone earns here, we say so rather than guessing.
Limits
What this data cannot tell you
Mobile consumer apps only — the most profitable B2B web products are invisible here. The rule-based classifier reaches roughly 55–70% coverage and ~75% accuracy: reliable for ranking which complaint dominates, not for treating any single percentage as exact. App review RSS caps at 500 reviews per app, so what you see is a recent-sample view, not a lifetime view. If you spot something wrong, we would rather hear it than have you act on it.
Go deeper
The free summary is about 30% of what we have
For one market, the full report adds every complaint category, all recurring phrases with counts, a differentiation checklist, complete payment evidence, and a build order.
All 87 complaint categories with counts
Every recurring phrase we extracted, counted
Differentiation checklist ordered by impact
Full payment-evidence library
Build-order recommendation, with the arithmetic shown