ASO July 25, 2026 · 4 min read

ASO Keyword Volume Estimates Are Getting Less Reliable — Here's What To Do Instead

Third-party keyword search volume scores have always been estimates — but App Store personalization is eroding even their directional value. Here's a research workflow built for 2026.

By the AppsOps news desk · July 25, 2026

If you have a subscription to AppTweak, Sensor Tower, or Mobile Action, you've probably built a habit around their keyword volume scores: high number means competitive, low number means opportunity. That mental model is getting harder to justify. App Store personalization has been quietly undermining the assumptions behind third-party volume data — and mid-2026 is a good moment to rethink how you actually use these tools.

What ASO tools are actually measuring

It's worth being precise about what keyword volume scores represent. Apple does not share search volume data with third-party tool vendors. Instead, tools like Sensor Tower, AppTweak, and Mobile Action infer relative keyword popularity from a combination of download panel data, rank tracking across many apps, and proprietary modeling. The numbers are ordinal rather than cardinal — they tell you roughly whether keyword A is searched more often than keyword B, not that either one generates a specific number of searches per month.

That distinction mattered less a few years ago, when App Store search was a single, centralized function. A keyword that showed 80/100 popularity for a developer in Austin would generate roughly the same results as for a developer in Berlin evaluating the same US store — the numbers pointed at the same underlying reality.

That's no longer true, and it's only gotten less true as Apple Intelligence and App Store personalization have matured.

How personalization is eroding the signal

Apple has steadily increased the degree to which App Store search results depend on who is searching, not just what they typed. A user with a health and fitness history sees different top results for "workout" than someone whose purchase history skews productivity apps. Apple's on-device models, described at greater length in our iOS 26 AI coverage, now inform discovery surfaces well beyond Spotlight — including, to a degree that Apple hasn't fully documented, App Store search ranking.

The implication for keyword research is structural: aggregate volume estimates from panels may be measuring a search population that doesn't represent your app's potential audience at all. A keyword with a volume score of 75 and strong general demand might convert poorly for a niche utility, because the people searching that term in Apple's personalized results aren't surfacing your app at position 1 or 2 — they're seeing apps that better fit their history.

High keyword volume doesn't mean high opportunity. It means high average demand — distributed across users whose contexts vary. Your app's slice of that demand is what matters, and no third-party tool can measure it directly.

This doesn't make volume scores worthless. They still help you identify whether a term has any audience at all and roughly how competitive the space is. But treating a score of 60 vs 70 as a meaningful signal is likely over-fitting to noise.

A better keyword research workflow for 2026

The most reliable keyword signals you have are first-party, not third-party:

Apple Search Ads Search Term Reports. Running even a small Discovery campaign with broad match exposes you to the actual terms users type before tapping your ad. These are real search strings from real users who were shown your creative — far more actionable than modeled estimates. Mine these aggressively before renewing a third-party tool subscription.
App Analytics → App Store → Search Terms. For organic, App Analytics shows you the search terms that drove impressions and page views for your default product page. This is first-party data from Apple. It's incomplete (Apple anonymizes low-volume terms), but what it shows you is ground truth for your app's actual discovery.
Conversion rate per term, not just impressions. A keyword that drives 5,000 impressions and 50 downloads beats one that drives 15,000 impressions and 30 downloads. Volume-first thinking ignores this.

Where third-party tools still earn their keep

Use volume estimates for competitive landscape mapping and long-tail discovery — i.e., finding keyword clusters you haven't thought to test yet, and understanding roughly which ones your competitors rank for. That directional function is still valuable. The mistake is substituting their scores for actual conversion data when prioritizing which terms go in your title, subtitle, or keyword field.

One underrated move: localize your keyword research, not just your metadata. A keyword that scores 70 in the US App Store may have almost no equivalent in Korean or Brazilian Portuguese — or the high-volume equivalent is a different concept entirely. If you're running 39-language localization, native keyword research per territory outperforms translating English keywords every time. Volume scores from English-market tools say nothing useful about search behavior in Japan or Germany.

The practical takeaway

Third-party volume scores are a starting point, not a decision. Use them to build your initial keyword universe, then validate and rank-order with first-party data from ASA and App Analytics. Weight conversion data above impression counts. And if you're localizing for multiple markets, treat each territory's keyword strategy as a separate research project — not an extension of your English metadata.

ASO tools aren't going away and aren't useless. But the teams getting outsized returns in 2026 are the ones treating Apple's own signals as the primary source of truth and using third-party tools for discovery and gap analysis, not as volume oracles.