About AppRank’s data

AppRank displays third-party app-market estimates and store metadata. New iOS metadata imports use Apple’s public lookup API; public top-free and top-paid charts use Apple’s RSS feeds. Legacy revenue estimates came from third-party providers including Sensor Tower. These are estimates, not audited company accounts or live payment records.

Paid-ad evidence and English focus

The advertising desk reads AdScan’s public discovery API once daily: up to 250 newest results for each of two store-domain searches, Apple and Google Play. We publish only returned records marked active, with a reported start date within the last 90 days, a direct store URL, and an exact store-ID match in our app catalog. Tracking links, fuzzy brand-name matches, inactive ads and unknown app identities are excluded. We do not access private AdManage customer campaigns, budgets or performance data.

English-copy selection is a conservative heuristic: at least five Latin words, at least three distinct common English markers, and predominantly Latin letters. This can miss English ads and misclassify some multilingual copy; it is not a verified language label. Ads without enough text are excluded. English copy does not establish audience location; country targeting is unavailable.

Counts are unique AdScan records in this bounded sample, not an advertiser’s total live ads, unique creative concepts, spend or results. AdScan may still mark an ad active after it stops. We show collection and reported start dates, never infer continuous running time or performance. Fresh advertising cards expire after three days without collection. Source errors retain the previous snapshot and back off at least a day. A daily database lease prevents duplicate runs. Only sanitized public creative fields are stored in AppRank; credentials remain server-side.

Revenue and downloads

Monthly figures reflect the provider’s available monthly estimate at collection time. They should not be interpreted as total company revenue, profit, or a guaranteed measure of advertising and web sales. Store and platform coverage can differ between apps. Small rounded values can reflect provider reporting thresholds; AppRank does not currently retain threshold qualifiers, so treat them as approximate.

Dates and freshness

Store metadata and revenue estimates have separate collection dates in UTC. Refreshing Apple metadata never makes an older revenue estimate look current. It is not the end of a financial reporting period. The iOS source fields describe the previous month’s worldwide estimates, but AppRank’s legacy snapshots are labelled with the collection month. Historical charts therefore show collected snapshots, not verified calendar-month accounts. Updates are periodic, not real-time.

Annual estimates and comparisons

Annualized revenue multiplies the available monthly estimate by twelve. It is a run rate, not measured annual revenue. Compare apps with similar platform coverage and collection dates. A missing value means unavailable, not zero. Store ratings are separate from revenue estimates.

Public sources and coverage

Apple lookup provides app names, descriptions, version and release dates, ratings, screenshots, languages and store prices. Apple charts provide ordered rankings for a specific storefront and free or paid chart. Neither source provides exact downloads, revenue, profit or advertising spend. Raw daily source snapshots are retained with their source URL and storefront. Public metadata updates do not overwrite financial estimates.

Collection uses a shared rate budget, daily caching, retry delays and a resumable queue. Apps unavailable in one storefront are not treated as removed worldwide. Android public-source ingestion is not yet enabled.

Daily discoveries

Our daily desk follows the US and UK free/paid top 100 charts. Daily movement compares snapshots 20–52 hours apart for the same chart and storefront. An app absent from the earlier collected chart is labelled new to the tracked chart, not a new app or a rise from an invented rank. Cross-market comparisons use collections no more than 12 hours apart. We show collection dates and hide chart discoveries after three days without a fresh snapshot.

Release cards use developer-provided version dates and notes for recently collected US app records. They cover our tracked set, not all releases in the App Store. A new version is not necessarily a new feature.

The builder feed

We collect Hacker News via Algolia and the GitHub repository search API once daily into dated database snapshots. The launch feed contains Show HN posts matching “app” from the last 30 days, ordered by discussion points. The open-source feed contains non-fork, non-archived repositories tagged mobile-app, with at least 50 stars and activity in the last 30 days, ordered by total stars. This is a discovery sample, not a comprehensive market ranking or an endorsement.

Points and stars are cumulative community metrics, not app users, downloads or growth. These projects are not automatically matched to App Store identities. Source failures retain the last successful collection with its date; snapshots older than three days are marked as delayed. A shared database lease avoids duplicate fetches, and failed collection backs off for a day.

Saved apps

Watchlists and comparison selections are stored in your browser. They do not currently sync across devices or send alerts.