How Does App Store Search Work? From Typing to Results

Apple states that people can use natural, everyday language when searching, and that search hints appear once someone inputs text into the search bar (Apple: App Store search). Hints inspire people before they search or select a result. For you, hints are a free source of phrases; see keyword research without paid tools.

Apple states an app is searchable by app name, subtitle, keywords and company name, and that the primary and optional secondary category are indexed. Text relevance means matches for the title, subtitle, keywords and primary category. What Apple does not document: how words are stemmed, how plurals match or how words from different fields combine. See ranking factors.

Search results are based on text relevance and on user behavior, including downloads and ratings and reviews. Apple gives no weights. A page with strong text relevance but poor ratings may sit below a page with weaker text and strong behavior; Apple does not say where the balance is. Treat any specific claim as a hypothesis to test.

Apple lists these among results:

• Apps and games with their rating and up to three screenshots or app previews, depending on platform and orientation
• In-App Events and developer cards
• Custom product pages, shown when keywords are assigned to them (otherwise the default product page)
• App tags generated by large language models from your metadata
• Categories and editorial content, and Apple Ads

So the "result" is wider than a list of names. Your screenshots and tags also work as part of the result. The screenshots glossary entry and custom product pages explain those parts.

Search works per storefront. Users can search with localized keywords in all countries or regions where the localization is shown. If no localization matches a user's language, the next most relevant localization is used. That is why keywords need to be written per language; see localization.

This guide describes only documented behavior. The stages between matching and ordering are not described by Apple, and tools that show "ranking factors" are showing observations. A rank you measure with a tool is the best evidence you have; see tracking keyword rankings.

Each stage maps to a task. Hints feed your idea list, matching defines what goes in your fields, ordering is why ratings and conversion matter, the results page is why your screenshots matter, and the country stage is why localization matters. Keyword research touches only the first two; the rest is the work around it.

LocalASO reads the public search results for each keyword in each country, which is how it reports your rank and estimates volume and difficulty from the rating counts of the top 10. Estimates are not Apple counts. See the keyword research docs.