How to read this, and what it can't tell you
Where the ideas come from. 10,291 app listings from the US iOS free, paid and top-grossing charts — Utilities, Productivity, Business, Sports, Travel, Weather, Social Networking and the cross-category Top Overall lists — were grouped into 1,105 distinct product ideas by the job each app does. 6,860 listings mapped onto an idea.
What the unmapped 3,431 listings are. Roughly half are single-brand apps — one airline, one bank, one football club — businesses rather than ideas you can build. The rest missed the name-matching, including some of the biggest names on the charts. Those unmapped listings hold about 52% of all charted revenue , and 199 of them earn over $100k a month. Every market-size figure here is a floor, not a ceiling.
Painkiller or vitamin. A painkiller is reached for at a specific blocked moment: the portal rejected the upload, the storage is full, the tide is coming in. A vitamin is real value the user can defer forever. Hybrids are genuine jobs people tolerate not doing. The call is about the moment of reach, not how useful the app is.
The $5k floor matters more than it looks. 80% of listings report revenue as "under $5k/mo" with no further detail, counted here at $2,500 each. That means 579 of the 1,105 ideas have a "market" figure that is just competitor count × $2,500 — no revenue signal at all. Those are labelled below-the-floor and shown as a range. Treat them as evidence of search demand and weak incumbents, never as evidence of money.
Opportunity blends provable money, audience reach, fragmentation, newcomer-friendliness, channel clarity, and build difficulty. Fit re-weights that for a solo developer who ships fast and prefers subscription or IAP. Both are judgment expressed as a number — use them to sort, not to decide.
Limits. One day's US-store snapshot, so seasonal categories are caught mid-season and non-US markets are invisible. Absence from the charts is not absence of demand. Clustering runs on app names, so a few apps sit under the wrong idea — the competitor tables are there so you can check rather than trust the grouping. And nothing here replaces talking to a user, which is still the only thing that settles it.
Dataset snapshot 16 Sep 2026 · interface redesigned for app.buildinstaller.com