App store logic works on a simple assumption: once a category stops piling up new ratings, people must be falling out of love with it. Look at the categories that lost the most rating momentum over the past eighteen months, though, and that assumption starts to fall apart. Some of them really are losing users. Others just ran out of first time raters for Apple to prompt, which produces the exact same flat chart while meaning almost the opposite thing.
Three things decide which side of that line a category ends up on: how habitual daily use is, how much of the install base Apple has already prompted, and how long ago the category's adoption wave actually peaked. Apple's own developer documentation for SKStoreReviewController spells out part of this, and it's a detail most coverage skips right over. Once you know it, a rating slowdown starts to read differently. Assume the worst just because a chart is falling, and you miss half the picture. The categories with the steepest rating cliffs aren't necessarily the ones bleeding the most users. Some of the steepest drops actually belong to categories that, by other measures, are doing just fine.
Why Rating Momentum Falls Off a Cliff
People misread rating counts constantly. A rating count isn't a proxy for downloads. It's a record of how many times Apple's system decided to surface a rating prompt to a given user, and how many of those prompts got a tap back. A category with a big daily audience, think social video, music streaming, messaging, keeps throwing off fresh prompt opportunities day after day. A category built around a one-time decision, like matching on a dating app or trying an AI chatbot for the first time, burns through its prompt opportunities fast, and that happens whether or not the underlying business is actually struggling. Most rating dashboards don't draw this distinction, so teams end up misreading the signal as a matter of course.
A category can look like it's losing rating momentum simply because its long-time users got prompted months back and there's nobody left to nudge. That's actually the milder scenario. The real question is whether the install base is still pulling in enough daily active users (DAU, in industry shorthand) to keep generating new prompts. It's only once that growth stalls that you're looking at a rating cliff worth worrying about.
A shrinking install base and one that's simply matured can look almost the same for a stretch before the two finally pull apart.
What the All-Time Leaderboard Actually Shows
AppWanderer tracks north of 252,000 apps, but its Most Rated of All Time board only counts the ones with a solid rating history behind them, and right now that's 70,613 apps. Those two numbers tell a story on their own: most apps in the catalog never build up a real rating history, and fewer still ever pick up any real momentum.
Five categories are pulling in ratings faster than everyone else right now. AppWanderer's latest count puts 抖音 at the top of the all-time list with 52.1 million ratings, with YouTube close behind at 48.1 million. Spotify: Music and Podcasts comes in third at 41 million, Instagram sits at 29.3 million, and Facebook rounds out the top five at 27.3 million.
None of these five are recent arrivals. Each one earned its place through years of daily use, not one viral moment. The math backs this up too: 70,613 out of a catalog above 252,000 works out to roughly 95 percent qualifying, maybe a touch lower once you account for the catalog's true size. Almost every app AppWanderer tracks eventually racks up enough history to land somewhere on the all-time board, but almost none of them build enough momentum to threaten the top five. The real divide isn't between having ratings and having none at all. It's between having some and never catching up to the leaders, and that gap isn't closing.
The Categories That Cooled Fastest
Flip the curve around and the categories that fell apart outright become obvious. AI chatbot and companion apps show this most clearly. They spiked through 2023 as novelty seekers grabbed them by the millions and left a flood of five star reviews, then stopped opening the app once Apple and Google folded comparable AI features straight into their own operating systems. That doesn't mean the category is dying. For most people, the free trial phase just ended, nothing more. Rating momentum tracks ongoing engagement, not downloads, so once curiosity faded, the prompts dried up fast.
Dating apps follow the same arc, just slower. Match Group and Bumble have both flagged softer engagement in their own public filings, and a category built around finding a match, then no longer needing the app, was never going to produce the daily prompt volume of a habitual use app. Industry guidance points to twelve to eighteen months after the trend crest as roughly when novelty driven categories peak and fall, though that's a pattern, not a rule. Social video looks nothing like that. People open it daily for years, and Apple's prompt system gets a fresh shot every time they do. Same store, same review system, wildly different trajectory. The difference comes down to how people use the app, not how much they love it. A rating cliff measures usage, not affection.
Usage decides everything here. Affection and quality barely factor in, only how often someone opens the app.
When a Rating Cliff Doesn't Mean What You Think
This doesn't hold up in every category, and I should say so before anyone thinks I'm overselling this. Niche categories, specialized business tools, government utility apps, single-purpose accessibility apps, just don't generate enough ratings for a percentage swing to mean much. A category that goes from 40 new ratings a month to 10 looks like a 75 percent collapse on paper. But at that volume you're staring at noise, not a trend, and treating it like real momentum is a mistake that belongs to small-catalog categories specifically, not the high-volume ones this piece is actually about.
If what you want is a plain popularity check, the App Store's own Top Charts or a download tracker like Sensor Tower will tell you who's winning this week's download race. That's a different question, though, and honestly it's probably the one most people mean to ask in the first place: download rank tells you about acquisition, rating velocity tells you whether the people who downloaded the app are still bothering to open it. This piece isn't trying to replace download-rank tracking, it won't function as a revenue forecast, and it won't tell you what app to build next.
Mix those two things up and you could end up cutting budget from a category that's actually healthy, just because its rating count leveled off, or you could keep pouring money into one whose ratings are climbing only because it's still coasting through its first, temporary adoption spike. Get that backward on a real team and the data hygiene mistake costs actual money.
Before you shift a single dollar of budget or a minute of attention, treat a rating slowdown as a usage pattern signal first and rule out prompt saturation before you read it as a popularity signal. Look at three things: how old the category's core user base is, whether daily engagement, not just downloads, is still climbing, and whether the category's win condition is habitual use or a one time decision. A category stalling out after years of daily habitual use, the kind that dominates AppWanderer' Most Rated of All Time leaderboard, deserves real concern. A category stalling out eighteen months after a novelty spike is usually just running out of new people to ask.
Skip that check and you'll misread a mature, still healthy category as dying, or mistake a temporary novelty bump for durable growth. Start with the usage pattern. Everything else follows from there.