If you compare the same App Store keyword across multiple ASO tools, the numbers will not match. One tool may call the keyword easy. Another may call it competitive. One may show strong demand. Another may show a modest opportunity.
This is not always because one tool is careless. It is usually because each metric is built from a different mix of public signals, Apple Ads data, store snapshots, proprietary panels, and estimation logic.
Apple does not publish every metric founders want
Apple gives the user strong data for their own app through App Store Connect: impressions, product page views, downloads, conversion, sales, and territory breakdowns. Apple does not publish competitor downloads, competitor revenue, private keyword fields, or exact organic search volume.
That gap is where ASO tooling lives. Tools observe what they can, model what they cannot, and present the result as a decision aid.
Popularity is usually a demand proxy
Popularity may use Apple Search Ads popularity when credentials or data access exist. Some tools combine that with autocomplete, result behavior, app rankings, and historical observations. The output is useful as a demand index, but it should not be read as exact traffic.
The right question is not “is this number exactly true?” The right question is “does this keyword have enough demand to be worth testing if the result page is reachable?”
Difficulty is a model of the current result page
Difficulty usually summarizes the strength of the apps already ranking. Inputs can include rating count, average rating, review velocity, title match, subtitle match, category rank, publisher diversity, app age, and whether the top results are dominated by a few brands.
Two tools can disagree because they weight those inputs differently. One may punish massive rating counts. Another may emphasize title relevance. Another may reward publisher diversity because it means the market is less monopolized.
Download and revenue estimates are not Apple facts
Competitor downloads and revenue are estimates. Vendors may use chart rank, category rank, ratings/reviews, IAP pricing, subscription surfaces, country availability, proprietary panels, and historic calibration. Exact formulas are rarely public.
These estimates can be useful for direction: which apps seem stronger, which categories have money, which countries look active. They should not be used as courtroom evidence.
Rankings are the most inspectable signal
Live search rankings are easier to inspect than popularity or revenue estimates. If a competitor ranks #3 for a keyword in the US today, that is observable. It may vary by device, personalization, or storefront behavior, but it is still a concrete market snapshot.
This is why ranking evidence should stay close to keyword decisions: rank history, top result changes, and country comparison are more actionable than one isolated score.
How founders should read ASO metrics
Use metrics to narrow decisions, not to outsource judgment. A good keyword has relevant intent, meaningful demand, reachable competition, and a product page that can convert that searcher.
When scores disagree, inspect the result page. If the top apps are weak, stale, poorly localized, or only loosely relevant, the opportunity may be real even when one tool’s difficulty score looks scary.