Apple Search Ads Validated Keyword Research for App Store and Play

A keyword set is worth exactly what its demand evidence is worth. Before a term reaches your recommended set we buy it as an exact-match keyword in Apple Search Ads and run it for fourteen days: normally 60 to 90 candidate terms against 1,500 to 2,500 euros of media that you fund on your own account. The number beside each keyword is therefore impressions Apple served to people who typed it, plus the tap-through rate your current icon and title earned against it. A term that cannot buy 300 impressions in a fortnight does not survive the shortlist, whatever popularity score a tool has assigned it. What you receive: Two keyword sets, one for the App Store and one for Google Play, built separately because the indexes are not comparable. Each term carries its measured impressions, its probe tap-through rate, the apps currently holding the top ten results, and our reading of whether you can take a position there inside two release cycles. Beneath that sits a field allocation: which terms go in Apple's 30-character title, which in the subtitle, which in the 100-character keyword field, which are conceded and why. For Play, the same allocation across the 30-character title, the 80-character short description and the body, marked against the terms Play is already indexing from your current text. German is not translated English: German search behaviour breaks most keyword tooling. Compounds are single tokens, so Kontowechsel and Konto wechseln are different queries with different competitive sets, and choosing between them decides which one you can win. Umlaut and ae spellings split demand and need separate probes. In the German storefront Apple indexes the English (UK) localisation alongside German, and we use it as a second keyword field rather than a duplicate of the first, which typically adds 60 to 80 usable characters of index. In banking and insurance there is a further gap: people search Depot and Kreditkarte, not the compliant product names in your marketing, and half the terms your product team suggests turn out to have no measurable demand at all. Branded, competitor, category: We report these three pools separately, because they behave differently and are worth different things to you. Branded demand tells you what your other channels are producing and is not an ASO result. Category demand is where the work is. Competitor demand we measure and then mostly hand back to Apple Search Ads: we will not place a competitor's trademark in your store metadata, because that is a takedown request waiting to be filed and we have seen listings pulled for it. Not included: Ongoing rank tracking. Rank is a diagnostic here, not a deliverable, and a monthly rank report is a way to bill for movement you did not cause. Metadata writing is a separate engagement, as is Apple Search Ads management beyond the fourteen-day probe. We do not forecast install volume from keyword data; anybody who does is selling you a spreadsheet function. Not for you if: Your app is pre-launch with no store history, or you cannot run Apple Search Ads in the target storefront, because then the method has nothing to measure. Also not for you if what you want is a large keyword list. We usually hand over 25 to 40 terms per store and a written argument for the ones we dropped, and clients who wanted volume find that thin. It is meant to be.