The demand the app store can't see

Before freezing the prompt sets I wanted to know which questions carry the most real demand, so I pulled app-store search volume for the head queries across every category. The result was almost comically flat. On AppTweak's popularity scale, which runs 5 to 100, nearly everything sat at 5, the floor, meaning below measurable. "Best VPN" cleared it at 54. A handful of bare category heads, "best workout app", "best video editing app", "best password manager", registered in single or low double digits. Everything else, including every "best X for Y" and every "X vs Y", read as noise.

These aren't low-demand questions. They're questions people don't type into an app store. They type them into Google and into chatbots, exactly where this index listens.

There are two different populations here. App-store search is short and navigational: someone who already knows they want a VPN types "vpn" into the store box. The decision happens earlier, somewhere else, in the messy questions: "best VPN for streaming", "Duolingo alternative", "CapCut vs VN", the kind an app store never sees. Published research on how people query AI agrees: people use keywords with search engines and full, conversational questions with assistants. Those questions don't exist in a keyword database.

Why this matters

This is the clearest argument for why an AI-visibility index is not just a re-skin of app-store optimisation. The entire layer of demand where recommendations get made, where an assistant decides which three apps to name, is invisible to the tools built around the store. You can be flawless at app-store keywords and absent from the answer a user actually acts on.

And absence there is expensive in a way that compounds. People increasingly explore in the chatbot and only then verify on Google or in the store. If the assistant never names you, the user never starts the search that would have found you. You're cut from the decision before the store is ever opened. That is the demand this index measures, and the reason it measures it where it does.

What this is

A note from building the method, not a leaderboard. App-store volumes are AppTweak's 5–100 popularity score, pulled 3 June 2026; the contrast with web/AI demand is directional, not a precise multiplier. The fuller write-up, with sources on AI query behaviour, lives in the project's findings notes. See any index for what the demand actually names.