Keyword Research for App Store and Play, With a Ranking Call

Start with the one first-party signal: Apple publishes a search popularity score from 5 to 100 inside Apple Search Ads, and it is the only volume figure for the App Store that comes from Apple. Every third-party estimate is modelled from it. Across medtech and utility apps, a term scoring under 25 has almost never produced install volume I could measure. The useful working band sits between 30 and 55, where demand exists and the incumbents holding the top slots are not global brands. Everything I recommend is anchored to that score and to a judgement about whether you can actually take the position, which no tool outputs. The two stores are not one problem: The App Store indexes the app name, the subtitle, the hundred-character keyword field and your in-app purchase names, and does not index the description at all. Google Play indexes the title, the short description and the entire long description, and weights repetition differently. A single keyword list ported across both is the most common thing I am hired to undo. The App Store side is a packing exercise inside roughly 160 characters of indexable text with no word repeated between fields and no character spent on a space after a comma. The Play side is an editorial exercise, because the long description has to read as prose to a human while carrying the terms in the forms people type. Polish, and why it is not a translation of the English list: Play's tokenisation handles Polish morphology unevenly. Aplikacja, aplikacji and aplikacją do not reliably collapse to one stem, so a Polish long description carrying only nominative forms loses the queries people actually run. The map I deliver for Polish carries the inflected variants that matter and marks the ones not worth the character cost, which is most of them. What gets ruled out before ranking is considered: In medtech a term can be the highest-volume phrase in the category and still be unusable. Indication names, condition terms and anything reading as a therapeutic claim will either fail review or create an exposure larger than the traffic is worth. I mark those explicitly rather than quietly dropping them, so your regulatory lead can see what was considered and why it was refused. For Polish energy apps the equivalent list covers tariff, switching and savings language. What you receive: A keyword map per storefront and locale. Each term carries its Apple popularity score or its Play demand read, the apps currently holding the top positions, a difficulty judgement, and a call: target now, target after the next release, or do not target. Then the field placement, which term belongs in the title, which in the subtitle, which in the keyword field, and what that choice costs the terms it displaces. The trade-offs are written down because that is the part you will argue about internally. Not included: Writing the final metadata, creative work, paid search management, and ongoing rank tracking. Each is scoped separately. Who should not buy this: Anyone who wants a list of a thousand keywords. The deliverable is a set of decisions across roughly forty to sixty terms per locale, most of which say no, and a client who measures this work by row count will be disappointed by it. Also anyone without Apple Search Ads access, because without it there is no first-party volume signal and I would be guessing at your expense.