"Find out which keywords your competitors use" is standard App Store advice, and it skips over a problem: you can't. A competitor's keyword field lives in their App Store Connect account, and no public Apple API exposes it. Any tool that claims to show you a rival's keyword list is showing you a guess.
What you can see is the outcome. Apple's public search results show which apps come back for a given term and in what order. Check enough terms and a competitor's footprint shows up without anyone having to guess at their metadata: the terms they appear for, the ones where they're ahead of you, and the ones you hold that they don't.
That comparison is a keyword gap. Most of what it turns up isn't worth acting on, so the useful part is knowing which rows to ignore.
The short version
- Pick competitors by search overlap, not by category.
- Split the comparison into three lists: terms they appear for and you don't, terms where they're ahead, and terms where you're ahead.
- Ignore their weak positions. A rival buried at position 150 isn't evidence that a term is worth having.
- Sort what's left by demand, highest first.
- Close a gap only if your app genuinely does the thing the term describes.
1. Pick competitors by overlap, not category
The competitors you think of first are usually the biggest names in your category. They're often the wrong comparison. A meditation app and a sleep sounds app can sit in different categories and fight over the same twenty terms, while two apps in Health & Fitness might never appear for a single term in common.
A better definition is behavioural: a competitor is an app that keeps showing up next to yours in search results. That's how our report picks the rivals it suggests. It counts the terms where both apps appear in the top 50, and it ignores category entirely.
2. Split the comparison three ways
Comparing two apps produces three lists, and each one asks a different question.
They appear, you don't. The terms where the rival shows up and your app isn't in the first ~190 results Apple returns. These are candidate opportunities, and this list is usually why people run the comparison in the first place.
You both appear, they're ahead. You're already relevant enough to show up, just further down. These are often cheaper to improve than the missing terms, because Apple already connects your app to the query.
You appear, they don't. Easy to skip, but worth reading: these are positions you hold that your closest rival hasn't reached. If you're about to rewrite your metadata, this list is what you don't want to lose.
3. Ignore their weak positions
A competitor appearing for a term tells you something only if they appear near the top. Apple returns roughly 190 apps for a typical search, and position 150 on a board that long is closer to incidental text matching than to evidence that the term brings them users.
Our comparison only counts a competitor's term when they're in its top 50. The limit exists partly because deep positions don't mean much, and partly because they don't hold still. When we measured positions across repeated crawls a few days apart, apps in the top 10 of a term stayed in the top 10 94% of the time, and the median position moved by just 2. But 1 in 10 moved 13 places or more. Near the top of a board that's still informative. At position 140, a 13-place swing is the difference between one stretch of filler and another, and the exact number stops meaning much.
4. Sort by demand first
The obvious way to sort a gap is by how far ahead the competitor is. It's also the least useful one. A term where they rank #2 and you're nowhere doesn't matter if nobody searches it.
So sort by demand first and position second. Our demand figures come from Apple Search Ads, the only place Apple publishes how much a term is searched. There's one catch worth knowing: since October 2025, Apple no longer reports a popularity score for terms that score below 35 on its popularity scale. A term without a score isn't zero demand. It's unknown demand, somewhere below that line. Treat those as unconfirmed rather than dead, and put the terms with a confirmed score at the top of the list.
5. Close only the gaps you actually fit
This is the step most competitor analysis skips. A rival appears for a term because their app is relevant to it. If yours isn't, copying the word into your metadata won't make you relevant, and Apple's rules forbid keywords that aren't relevant to the app.
A few kinds of gap should almost always be left alone:
- Their brand. A rival often ranks for its own name and the names of its features. Apple doesn't allow competing app names in your keywords.
- A feature you don't have. If they appear for "widget" because they ship widgets and you don't, the gap is a product question, not a keyword one.
- A different audience. A calorie counter and a diabetes log can share vocabulary without sharing users.
What's left is usually a short list: terms that describe what your app already does, with real demand, where a direct competitor has shown it's possible to appear. That short list is the output.
Closing a gap
Once a term makes the list, the work is in your metadata. Its words need to appear somewhere across the app's name, subtitle or keyword field. Apple combines words across all three, so you don't need the exact phrase. Where characters are tight, the keyword field mistakes that waste characters are usually the cheapest source of space.
Then give it time. Metadata changes only take effect with a new version, and positions settle over the following crawls. Judge the change a week or two later, not the day it ships.
The honest caveats
- Public search results aren't what every user sees. Positions come from Apple's public Search API. The in-app search results a real person gets are personalised and undocumented, so treat positions as a consistent reference, not a copy of anyone's screen.
- We only compare terms we've crawled. A term neither app has been checked against doesn't appear in the comparison at all. That's missing data, not a missing gap. Running a report on either app queues its candidate terms for the next crawl.
- Not appearing is bounded, not absolute. Apple's public search returns roughly 190 apps and no second page. "You don't appear" means "not in the first ~190," which in practice amounts to the same thing for discovery.
- A gap isn't a forecast. Appearing for a rival's term won't necessarily bring you their users, and we don't publish download or revenue estimates to pretend otherwise.
What to do with this
Search for your app and open its keyword report. The Compare against a competitor section lists the apps that most often appear next to yours, and each one opens the three-way comparison, sorted by demand.
Related: why every app has keywords it can win covers the other half of the problem: finding terms that no competitor has claimed yet.