How do I find search volume for App Store keywords?

Apple never publishes absolute search counts. The closest real signal is the Apple Search Ads popularity score: a 5–100 index of relative search demand, viewable without running ads. Every 'monthly searches: 12,400' figure in a third-party tool is that tool's model translating the score, not Apple data. Use popularity for ranking keywords against each other, not for forecasting installs.

What is a good popularity score for an indie app keyword?

Practitioner rule of thumb: scores around 5 mean near-zero demand (5 is also the floor, so it hides everything from nothing to barely-something); the teens are thin; sustainable indie targets often sit in the 10–50 band where demand is real but the giants aren't fighting for it. The scale compresses the top: the jump from 60 to 70 represents far more searches than 20 to 30.

Is App Store autocomplete a good keyword research signal?

One of the best free ones: suggestions are drawn from what people actually type, ordered by Apple's own data, per storefront. If your phrase autocompletes after a few letters, demand exists; if you can type it in full without Apple offering it, treat any tool's volume estimate for it with suspicion.

FunnelHound / Guides / Keyword search volume

App Store keyword search volume: what the popularity score really means

By Emin, indie iOS developer building FunnelHound · Published 18 Sep 2026

Uncomfortable truth first: Apple publishes no search counts. Anywhere. For anyone. The only demand signal from inside the house is the Apple Search Ads popularity score, 5–100, and every "12,400 monthly searches" you've ever seen in a tool is a model's estimate of that score wearing a confident font.

Reading the 5–100 scale without getting fooled

Three properties matter. The floor is 5, not 0: a 5 spans everything from "nobody has ever typed this" to "a dozen people once did", so a 5 tells you almost nothing except don't bet the title on it. The scale compresses upward: behavior is logarithmic-ish, so 60→70 represents far more additional searches than 20→30, and the 90s are a different universe occupied by "instagram" and "whatsapp". It's relative, not absolute: perfect for ranking two candidate keywords against each other, useless for forecasting installs. The practitioner band worth knowing: sustainable indie targets usually live around 10–50 (real demand, beneath the giants' radar), the same popularity-versus-difficulty math the keyword tools automate.

Where the "monthly searches" numbers come from

Third-party tools take Apple's score, blend their own panels, ad data and category models, and print an absolute number. Some models are decent; none are measurements. The tell: different tools confidently disagree about the same keyword by 5×, which is fine once you treat the numbers as the score, re-costumed, and compare only within one tool. What you never do is arbitrage between tools ("tool A says 3,000, tool B says 15,000, so B's keyword is better"). That's comparing two costumes and calling it research.

The free demand checks that don't lie

Autocomplete: type three letters in the App Store. Apple's suggestions come from real queries, per storefront. Phrase completes early = demand exists. Doesn't complete at all = every estimate for it is fan fiction.
The popularity score itself, free in the Apple Search Ads console, no campaign needed, and it's per country, which matters enormously for localization since the same concept scores wildly differently across storefronts.
Your own impressions: the ground truth. Rank on a term, then watch search impressions. Rank #8 with impressions flat? The demand was never there, whatever the estimate said. This is the check that retires debates: volume estimates predict; impressions confirm.

Demand is a hypothesis until the funnel agrees

The whole game in one sentence: popularity scores and volume estimates choose which keywords to try; your own funnel decides which ones were real. A keyword that looked like a 40 and moved nothing gets replaced at week six; a modest 15 that quietly feeds installs earns its slot forever. FunnelHound sits on the confirm side (search impressions, installs and purchases from App Store Connect, milestone on the day the new keywords shipped), so the estimates stay accountable to the only counter that pays.

Estimates propose. Impressions dispose.

FunnelHound shows whether the keyword actually delivered: search impressions to installs to money, from App Store Connect.

Get FunnelHound

Share this on X

Data notes: popularity score behavior per Apple Search Ads as of 2026; the logarithmic-ish reading and the 10–50 indie band are practitioner consensus, not Apple documentation. Third-party volume models change methodology without notice. Compare within one tool only.