We want you to trust the numbers, so here is exactly how we get them, and where they are weak.
We take a snapshot every day and keep the history. We are not affiliated with Apple.
Categories are Apple's 25 App Store categories. Apple has no official niches, so we define about 235 niches. Each niche has one or more search phrases (for example “habit tracker”). The apps that rank for those phrases, in the US store and in each local store, form the niche. We keep the 30 most relevant search results per phrase.
solid = bayesian_rating × log10(ratings + 10). The Bayesian rating pulls apps with few ratings toward 4.0 (as if each app had 300 extra ratings of 4.0). So an app needs both quality and scale to rank high.
The sum of rating counts of a niche's top 10 apps. Rating counts grow with downloads, so they are a stable proxy for demand. They are not download numbers.
New ratings in the last 30 days, from our daily snapshots. Until we have 30 days of history, we use the longest window we have and scale it to 30 days. Before that, the opportunity score uses each app's lifetime rating speed (ratings ÷ age) instead.
An app is a breakout when it is younger than 540 days, has at least 50 ratings and an average of 4.0★ or more, and collects at least 1 new rating per day.
For the leading apps of each niche we read the most recent reviews from Apple's review feed. Claude (an AI model by Anthropic) groups the 1-2 star reviews into recurring themes, lists missing features and writes a short opportunity note. The share of negative reviews also feeds the opportunity score. Refreshed about every 28 days.
Apple publishes no revenue. We estimate it from the top-grossing chart position with a power-law model:
daily revenue ≈ category_top × market_share × rank^-0.85
We show a range from 0.5× to 2× of that value. Apps outside the top 100 get no estimate. Use these ranges to compare niches by order of magnitude, not to forecast your own revenue.
Each part is the niche's percentile rank among all niches in the same country store, then weighted:
| Part | Weight | Higher score when… |
|---|---|---|
| Demand | 25% | many people use apps in the niche |
| Growth | 20% | the niche gains ratings fast relative to its size |
| New apps win | 20% | young apps are in the top 10 or break out |
| Unhappy users | 15% | a high share of recent reviews is negative |
| Weak quality | 10% | the top apps have lower ratings |
| Fragmented | 10% | no single app holds most ratings |
The score is a starting point for research, not a verdict. Always read the evidence behind it.