Data Study · Original research

The Skew Factor

Some categories run on a handful of juggernauts while others spread success evenly — the percentile data reveals which.

Original research by AppWanderer, computed from a sample of 517,612 public App Store listings drawn from our index — every figure here is calculated, never estimated. See our editorial standards and glossary.

Summary

Food & Drink turns out to be the most lopsided category in this whole dataset: the top 1% of its apps haul in 34,736.7 times the rating count of the median Food & Drink app. Stickers lands on the other extreme, with a p99/p50 ratio of only 215. That contrast is basically the point of this whole study. Popularity on the App Store doesn't distribute evenly within categories, and the unevenness swings by a factor so large that differences in average quality scores start to look almost beside the point.

Average that skew ratio across every category and you get 10,619.1, which is only 30.6% of the ceiling value of 34,736.7 that Food & Drink sets. Stickers' floor of 215 works out to barely 2% of that average. Even a so-called "typical" category sits far more skewed than normal statistical assumptions would predict, and the extreme cases still dwarf the typical one by a wide margin. Quality, when you measure it through the spread in average ratings instead, never gets anywhere near this kind of range. That's really the thing worth sitting with: popularity piles up hard in a handful of categories, while quality variance stays comparatively tame across the board.

34,736.7
Highest popularity skew (Food & Drink, p99/p50)
215
Lowest popularity skew (Stickers, p99/p50)
10,619.1
Store-wide average skew ratio
517,612
Apps in the study population

Food & Drink, Travel, and Shopping are where a tiny handful of apps own the whole category

Food & Drink comes out with a p99/p50 ratio of 34,736.7, which means the apps sitting at the very top of the rating-count distribution have racked up review totals tens of thousands of times bigger than whatever the median app in that category can claim. Travel isn't far off at 29,866.6, and Shopping comes in at 27,514. These are the categories that tend to be built around one or two utility-grade winners, the delivery app or booking app already sitting on your home screen, with thousands of smaller apps scattered around it that nobody has bothered to rate in any meaningful number. Finance (17,215), Lifestyle (15,089.6), and Entertainment (14,415.1) show the same shape at a smaller scale: a handful of apps define what the category even is, then a long, quiet tail stretches out behind them. Productivity (13,369.1), Business (13,290.8), and Sports (12,773.5) belong in this same group, though the dynamic changes a bit here, shifting from a single runaway winner to a cluster of entrenched incumbents splitting most of the attention between them. Social Networking (10,091.8), News (9,783.9), and Book (9,188.3) land at the gentler end of this tier, though even those numbers leave the median app looking almost invisible next to the leaders. It's tempting to read a big category-level rating count as proof of broad, even engagement across the board. In these categories, though, that aggregate figure is really just describing a handful of apps, not the category as a whole.

Figure 01 · Food & Drink, Travel, and Shopping are where a tiny handful of apps own the whole categoryComputed from the live index

Stickers and Magazines & Newspapers barely have a winner-take-most problem at all

Stickers comes in with a p99/p50 ratio of 215, the lowest mark anywhere in the dataset, and Magazines & Newspapers isn't far off at 801. Put either number next to Food & Drink's 34,736.7 and the difference is obvious right away: in these categories, the top of the distribution and the middle basically live in the same neighborhood. Developer Tools (1,465.7) and Graphics & Design (2,446.3) follow the same pattern, which makes sense for categories aimed at smaller, more technical audiences, where it's rare for a single app to run away from the pack. Reference (4,602.1), Music (4,963.8), and Weather (5,123.2) climb a bit higher but still sit in the quieter half of the rankings. Navigation (6,134.3), Games (6,184.2), Education (6,814.6), Medical (6,987.1), and Health & Fitness (7,292.1) round out the group, clustered well below the top-tier categories. None of this is mysterious: categories with lower ceiling effects, scattered use cases, or smaller user bases just don't tend to produce that one runaway app that drags the p99 figure upward. For analysts, it means the category-wide rating totals here give a more honest read on the whole population, rather than reflecting the pull of one or two standout apps.

Figure 02 · Stickers and Magazines & Newspapers barely have a winner-take-most problem at allComputed from the live index

The average category skew ratio is already extreme before you even look at the worst offenders

Averaged across every category, the skew ratio lands at 10,619.1. That's 30.6% of the single highest value in the dataset, Food & Drink's 34,736.7. Worth sitting with for a second: the "typical" category isn't some watered-down version of the worst offender, it's sitting at nearly a third of it. Stickers, at the bottom with a skew ratio of just 215, only accounts for 2% of that same average. So the low end isn't just lower, it's playing by different rules entirely. Put those two numbers side by side, 2% and 30.6% against the same baseline, and you can see how a single average glosses over real differences between categories rather than summarizing them. Collapse Food & Drink and Stickers into one store-wide figure and the thing that actually matters disappears: which category you're looking at tells you far more about concentration than any blended average could.

Shopping's p90/p10 gap of 44,773 accounts for 40.5% of the entire spread total across these twelve categories

Add up every p90/p10 gap in this cut and you get 110,583. Shopping alone eats 44,773 of that, which works out to 40.5%, and there's no real competition for second place. One category is carrying almost the whole comparative measure on its back. Games comes in a distant second at 15,441 (14%), with Social Networking further back still at 9,229 (8.3%), and even so, both of those clear everyone else below them by a wide margin. Finance (7,520, 6.8%) and Photo & Video (5,260, 4.8%) are roughly where things start tightening into single-digit percentage territory. Weather (4,789, 4.3%), Health & Fitness (4,708, 4.3%), and Graphics & Design (4,507, 4.1%) basically land on top of one another. It looks like these categories share something close to the same moderate distance between their well-reviewed apps and the ones nobody bothers reviewing. Entertainment (4,170, 3.8%), Lifestyle (3,542, 3.2%), Food & Drink (3,450, 3.1%), and News (3,194, 2.9%) close out the list, each holding the smallest slivers of this spread measure. Shopping's p90/p10 gap would stand out even by itself, but stacked against everything else, it just dominates. Nothing else here is even in the same conversation.

Figure 03 · Shopping's p90/p10 gap of 44,773 accounts for 40.5% of the entire spread total across these twelve categoriesComputed from the live index

Shopping and Games squeeze their middle tier harder than any category squeezes its top tier

Shopping comes in with a p75/p25 ratio of 327.4, and Games isn't far off at 319.6, the two highest figures in this study. That's a different picture from the p99/p50 list, where neither of these two even made the top. In Shopping and Games, the gap runs through the middle of the pack, not just among the rare top 1%. An app sitting at the 75th percentile in Shopping already pulls in a rating count 327.4 times that of one at the 25th percentile, long before you get anywhere near the extreme top. Food & Drink plays out differently. It topped the most-skewed list at the p99/p50 level with 34,736.7, but here it falls all the way to 72.3. Its concentration sits almost entirely at the very top, not spread across the middle. Social Networking (122.5), Finance (115.5), and Health & Fitness (111.3) land in a middle band, where even upper-middle and lower-middle performers are still clearly pulling apart from each other. Photo & Video (97), Graphics & Design (83.8), Weather (81), Lifestyle (76.3), News (73.7), and Travel (70.7) sit at calmer levels. If you're competing in Shopping or Games, you don't need to be a category heavyweight to feel the squeeze. Just slipping below the 75th percentile is enough to feel crowded out.

Figure 04 · Shopping and Games squeeze their middle tier harder than any category squeezes its top tierComputed from the live index

Quality variance doesn't track popularity concentration, and Business proves it

Business has the widest quality spread in the dataset, averaging 2.4 points of rating range, but it's not the category with the worst popularity skew. It lands mid-pack there, at 13,290.8. Entertainment comes in at 2.0 and Medical at 1.9, and neither one ranks near the top for popularity skew either. Stickers and Utilities tie at 1.8, and this is where any tidy narrative falls apart. Stickers has the lowest popularity skew in the entire study, 215, yet its quality spread is almost as wide as Business's, the very category with the most concentrated popularity outside the top tier. Lifestyle and Sports both sit at 1.7, Social Networking, Navigation, Finance, and Travel land at 1.6, and Weather closes out the list at 1.5. None of these quality-spread rankings line up with the popularity-skew rankings covered earlier. A category can carry wildly concentrated popularity alongside fairly modest quality variance, the way Food & Drink does, or carry almost no popularity concentration alongside real quality variance, the way Stickers does. Average rating and rating count concentration are simply measuring two different things.

Figure 05 · Quality variance doesn't track popularity concentration, and Business proves itComputed from the live index

The bottom line

Popularity concentration and quality variance turn out to be two different stories, and the data backs that up. In Food & Drink, Travel, and Shopping, a handful of apps soak up almost all the rating activity in the category, with p99/p50 ratios running from 27,514 up to 34,736.7. Stickers and Magazines & Newspapers sit at the calm end, where the top apps and the median ones aren't far apart. Shopping and Games carry that unevenness into the middle of the distribution too, not just the extreme top, while Business and Entertainment top the charts for quality spread without showing much popularity skew at all. Read category-level rating totals against that category's own skew figure, not against a store-wide assumption.

Method & limitations

How this was made: this study is computed across 517,612 apps carrying the public App Store metadata it needs, drawn from our live index of 1,063,379 tracked apps, and is recomputed on a schedule. It is a large, representative sample of the store rather than every app in existence, and the figures shift over time as the store changes.

  • Public metadata only. These figures come from what apps publicly list in the App Store. Downloads, revenue, retention and in-app subscription pricing are not visible to us — we name them, and we never estimate them.
  • A large sample, not a census. The study runs across a representative slice of the store rather than every app in existence, so figures move as the store and our index change.
  • Recomputed, not frozen. Numbers are recalculated on a schedule; a figure quoted elsewhere may differ from the one shown here after the next rebuild.
  • No estimates or third-party numbers are used.

Common questions

Does a high popularity skew mean the category has bad quality control?

No. The categories with the widest quality spread (Business at 2.4, Entertainment at 2.0) are not the same categories with the worst popularity skew (Food & Drink at 34,736.7, Travel at 29,866.6). Popularity concentration and rating-quality variance move independently of each other in this data.

What does a p99/p50 ratio of 34,736.7 actually mean for Food & Drink?

It means the apps at the top 1% of the rating-count distribution in Food & Drink have rating counts 34,736.7 times larger than the median app in that same category. The category total is effectively being carried by a very small number of apps.

Which category shows the most even distribution of popularity?

Stickers, with a p99/p50 ratio of 215, the lowest figure in the entire dataset. Its top apps and its median apps are far closer together than in any other category measured here.

Why do Shopping and Games top the midrange skew list but not the most-skewed top-tier list?

Because their unevenness shows up earlier in the distribution. Shopping's p75/p25 ratio of 327.4 and Games' 319.6 mean the squeeze starts well before the extreme top of the chart, unlike Food & Drink, whose skew (34,736.7 at p99/p50) is concentrated almost entirely in the very top performers and drops to just 72.3 at the p75/p25 level.

Data & citation

Original data from AppWanderer. The figures are licensed CC BY 4.0: free to cite, quote and reuse, including the charts, with credit and a link back to this page.

AppWanderer. “The Skew Factor.” https://appwanderer.com/studies/category-skew-index (data updated 27 Oct 2026).

Press & data enquiries: [email protected]

Dor Bass
Author
Dor Bass

Builds and maintains the index behind AppWanderer. Writes the studies and sets the qualifying rules behind every ranking board.