How App Store and Google Play search work
Both stores index a handful of text fields, then order the matches by strength. Here is what the App Store and Google Play read, and what they ignore.
Video transcript
Every store search works in two steps. First, which apps match the words you typed. Then, in what order they appear. Matching reads only a few text fields. On the App Store: the app name, the subtitle, and a keywords field nobody else can see. On Google Play: the title, the short description, and the full description. Everything else on your page sells the app. It doesn't rank it. Stores read words, not phrases, and Apple combines its three fields by itself. So a word you repeat wins nothing. Which field moves a ranking most? Applyra measured it on real listing changes. The title, by far. The second field is a real gain, a fraction of the title's. The third barely moves anything when a word goes in, but take it out, and you lose ground. So: title first, second field next, and the third for what didn't fit. And neither store announces a change. When one lands, a whole store can drop overnight, and it looks exactly like your listing broke. So before you rewrite anything, check whether the store moved first.
App Store and Google Play search both work in two steps. Each store indexes a small set of text fields from your listing and decides which apps match a query from the words in those fields, then orders the matches by how strong each app looks on that storefront. Everything Applyra measures hangs off those two steps: what you are indexed on, and where you land once you are.
What each store indexes
| Store | Indexed for search | Ignored by search |
|---|---|---|
| App Store | App name (30 characters), Subtitle (30 characters), Keywords field (100 characters) | Description, Promotional text, What's New |
| Google Play | Title (30 characters), Short description (80 characters), Full description (4,000 characters) | Release notes |
Those budgets are per localization on iOS and per language on Google Play, so every locale you add hands you a fresh set. The full table, with the counting rules that decide whether you actually fit, is in the metadata fields each store indexes.
How the words are read
Each store splits your fields into single words and builds the multi-word queries itself.
Apple combines all three of its indexed fields and permutes across them, which is why
photo editor,photo filter in the keywords field buys nothing that photo,editor,filter
does not, and costs six characters more.
Two consequences worth writing on a sticky note:
- A word is indexed once per listing. Repeating your title's term in your subtitle is not reinforcement, it is a field spent twice on the same thing.
- Filling a field is not the same as ranking. Length wins nothing on its own. What wins is carrying the searched term, and a longer field simply fits more of them.
Which field actually moves a ranking
This one is measured rather than inherited from ASO folklore. Applyra compared real listing changes with each app as its own control, so the app's relevance is held constant and only the metadata moves. What a tracked keyword gets when it enters a field:
| Field | Effect on entry |
|---|---|
| Title | The biggest move by far |
| Second field (subtitle / short description) | A real gain, a fraction of the title's |
| Third field (keywords field / long description) | Almost nothing |
The full study, with its figures and how long each change takes to land, is in how long ASO changes take to take effect.
The third field looks worthless from that column alone, and it is not. Take a keyword back out of the iOS keywords field and the app loses real ground on it. That field holds a position rather than winning one, which is exactly why you fill it and then leave it alone.
Same hierarchy on both stores, with a stronger effect on the App Store, which is convenient: one rule of thumb covers you whichever store you are writing for. Title first, second field next, third field for what would not fit.
The demand side: whether anybody searches the term
Matching gets you into the results. It says nothing about whether anybody types the query. The stores never publish search volumes, but they do publish their own autocomplete, and that is a demand signal they compute from real searches. Applyra reads it to build the traffic score, and pairs it with how strong the incumbents are to build the difficulty score.
Search is not the only surface
Both stores also rank apps in category and global charts, which is a discovery channel you do not target with words. It counts in the visibility score, behind your keywords, because most apps never chart at all and the keywords are the part you can act on.
The rules change without notice
Neither store announces a ranking change. They still happen, and they are visible in the data within a day. One can push a whole store's keywords down overnight while the other store does not move at all, and the mid-table positions, the ones a small app actually fights for, take the worst of it.
That is the kind of event that looks exactly like "my listing broke" from inside your own dashboard. Before you rewrite anything, check the algorithm change detector, and read why your App Store keyword rankings dropped overnight for the diagnostic that costs nothing: compare against a store you did not touch.
Frequently asked questions
Which fields does the App Store index for search?
Three: the app name (30 characters), the subtitle (30 characters) and the keywords field (100 characters). Apple does not index the description, the promotional text or the What's New notes, so keywords written there win you nothing on the App Store.
Which fields does Google Play index for search?
Three as well: the title (30 characters), the short description (80 characters) and the full description (4,000 characters). Unlike Apple, Google reads the long description, which is why a keyword disappearing from it costs you the ranking it was holding.
Does repeating a keyword in several fields help you rank?
No. Both stores index a word once per listing, and Apple combines its three indexed fields and builds the multi-word combinations itself. A term repeated from the title into the subtitle spends characters you cannot get back and buys nothing.
Which metadata field has the biggest effect on rankings?
The title, by a wide margin, on both stores. A keyword entering the title moves a ranking several times more than the same keyword entering the second field, while the third field mostly holds positions you already have rather than winning new ones.
Last updated on