App intelligence is a commodity in 2026 and a niche map takes years. Apple publishes one listing per app: ratings, stars, chart rank, price, in-app purchases, screens, description, release date. Downloads and revenue go to the developer alone, inside App Store Connect. So every estimate on the market models the same public signals plus a panel, and every tool pulls the same ads, screens and prices. That layer gets cheaper every year. What no scrape produces is a map of the store in named niches with every app read and tracked month after month. That takes years and it decides what you build.
What every tool reads
| Signal | Where it comes from | Who can have it |
|---|---|---|
| Ratings count and stars | The listing | Anyone |
| Chart rank | Apple's top charts, by country and category | Anyone |
| Price and in-app purchase prices | The listing | Anyone |
| Screens, description, release date, version history | The listing | Anyone |
| Ads, and how long each has run | Meta Ad Library, public | Anyone |
| Keyword rankings | App Store search results, sampled | Every ASO tool |
| Downloads and revenue | App Store Connect, one developer at a time | Apple and that developer |
Nobody outside Apple has the last row. Everybody models it from the rows above.
How vendors estimate revenue
The vendors describe the method themselves as of August 2026.
Sensor Tower credits "proprietary modeling and panel-based methodologies" and runs its own panel plus data.ai's since 2024.
Appfigures turns "information that can be observed in the store" into estimates with models trained on "hundreds of thousands of apps that share that data".
AppMagic reads chart position: "The app's position in the stores' top charts reflects its actual performance." Sensor Tower bought it in May 2026.
They work where panels are dense and rank moves daily: the top of the charts. Every app in Appaloma clears $5k a month and 94 in 100 sit outside the top 100 of every US category chart. Down there the model extrapolates from ratings and price. Two tools print two numbers and neither is a measurement.
So Appaloma publishes a band: $5–10k, $10–20k, $20–50k, $50–250k, $250k+ a month. Confirmed figures take their band as written. We estimate the rest from price, use, country and ads, and the longest ad run is the strongest signal we have. A band changes a build decision. An exact number on an indie app is false precision.
What a scrape does not produce
AppMagic has a free tier. Appfigures shows a monthly estimate for millions of apps free. Ads sit in a public library, screens and prices on the listing. Appaloma holds that layer too, 227,000 screens and prices on 31,000 apps, and it is the least of what it holds.
The unit
Apple files apps in 24 categories. A niche is the set of apps a user compares before paying: photo-scan calorie counters, alarm clocks with missions, pray-to-unlock blockers. Filing an app takes reading it. Appaloma did that for 4,000 niches three levels deep and the map keeps splitting. How to pick from the map sets out the next step.
Time
The Meta library shows what runs now and a commercial ad that stops leaves within weeks. Only pull after pull keeps the record, and the longest run is the signal. Rising, New and Take compare a niche's pace now with its pace before. A tool that starts today cannot pull the creatives that stopped last spring or the ratings count from a year ago.
The complaint that repeats
We pool 641,000 public reviews across the top apps of each niche, one star first. After Character AI added an age check in August 2026, nine of twelve leaders in AI companion chat carry "won't let me in" reviews. Eight of twelve mission alarm leaders say the alarm did not go off. A complaint in eight apps is a brief.
The angle
The angle is what each app leads with, and the count is how many apps in the niche run the same one. "Calculate peptide shot doses" runs in 17 apps and "Pray to unlock" runs in 11 of 12. An angle 17 apps run is proven and an angle two apps run is open.
Our read
Estimates converge on one number per app and keep getting cheaper. The map is the work and there is no API for it.
Which layer to buy
Appaloma is the niche and USP map: 4,000 named niches, 30,000+ USPs, 50,000 apps over $5k a month, for the next build or acquisition. Tools by job prices keyword tracking, your own numbers, and competitor estimates. Sensor Tower alternatives covers the enterprise case.
FAQ
How accurate are app revenue estimates?
Estimates hold at the top of the charts where vendor panels are dense and loosen below it. Apple publishes downloads and revenue for no app, so every estimate models public signals plus a panel. Read an indie estimate as a band.
Does Apple publish download numbers?
No. Downloads and revenue sit in App Store Connect and each developer sees only their own. Apple publishes the listing and the top charts.
Why does Appaloma show a band instead of a number?
The number would be a model, so the page prints a band: $5–10k, $10–20k, $20–50k, $50–250k, $250k+ a month. Confirmed figures take their band as written. We estimate the rest from price, use, country and the longest ad run.
Can I get ads, screens and prices without a paid tool?
Yes: the Meta Ad Library, the listing, and AppMagic's and Appfigures' free estimates all carry it. None of it tells you which niche the app sits in, who else is in it, or what changed in the last year.