Methodology

Source and collection

The archive requests public App chart JSON from Apple Marketing Tools. The first phase records Overall Top Free and Overall Top Paid charts. It does not collect category charts, downloads, revenue, ratings, reviews, or search-keyword positions.

Each of the 28 market-and-chart feeds is checked every two hours with staggered requests. HTTP 429 and transient server errors use delayed retry. A response is accepted only when the market matches, the results array is non-empty, every row has a unique stable App ID and name, ranks are continuous, and the result does not exceed the requested Top 100.

Markets and capital reference time zones

A Market Day uses the IANA time zone of the country's legally operative national capital. Server and visitor time zones never change an archived date. If a capital relocation legally takes effect, only future snapshots adopt the new reference zone; previous dates remain unchanged.

Market	Capital	Reference time zone
United States	Washington, D.C.	America/New_York
United Kingdom	London	Europe/London
Japan	Tokyo	Asia/Tokyo
South Korea	Seoul	Asia/Seoul
Germany	Berlin	Europe/Berlin
France	Paris	Europe/Paris
Canada	Ottawa	America/Toronto
Australia	Canberra	Australia/Sydney
Brazil	Brasília	America/Sao_Paulo
India	New Delhi	Asia/Kolkata
Indonesia	Jakarta	Asia/Jakarta
Mexico	Mexico City	America/Mexico_City
Spain	Madrid	Europe/Madrid
Italy	Rome	Europe/Rome

Daily Snapshot selection

Several Collection Attempts may occur during an open Market Day, but the archive keeps one full chart payload. Complete Coverage takes priority over Partial Coverage. Among complete results the newest wins. If every result is partial, the deepest coverage wins and the newest breaks a tie. Results captured at different times are never merged into a synthetic chart.

The selected result is Provisional while its Market Day remains open and becomes an immutable Final Snapshot after that day closes. Every attempt retains compact audit metadata, including its timestamp, status, count, source update time, content hash, and selection outcome.

Coverage and daily states

Historical snapshots retain their own Coverage Limit. Older Top 50 evidence remains Top 50 after collection moves to Top 100. A valid shorter response is shown as Partial Coverage rather than discarded.

• Ranked: the App has an observed position in that Daily Snapshot.
• Not Observed: a valid snapshot exists, but the App is absent inside that snapshot's tracked range.
• Snapshot Missing: no valid selected snapshot exists for an expected Market Day.
• Not Tracked: the date is outside this market-and-chart archive.

Charts leave missing and unknown days as gaps. They do not connect a line through them or treat absence as rank 101.

Comparable movement, entries, exits, and churn

Two snapshots are compared only inside their shared range: Top 1 through the smaller Coverage Limit. An App visible only because tracking became deeper is Newly Tracked, not a New Entrant. An App below a shallower current limit is not called an Exit. Exact movement remains valid when the App is observed on both dates.

Comparable churn is calculated as (New Entrants + Exits) ÷ (2 × shared range) × 100. Snapshot Missing and Not Tracked dates do not contribute. Verified streaks require consecutive Final Market Days with explicit rank observations.

App History and publication threshold

An App series is research-ready after it covers at least 14 completed Market Days and contains rank observations on at least seven of them. App pages remain public before that point but stay outside search indexing and advertising. The default selected history window is 90 days, with 30-day, 365-day, and all-time alternatives.

The App-page identity uses stable reference metadata, preferring source records from the United States, United Kingdom, Canada, and Australia. Each Market's original names, artwork, and App Store URL remain separately versioned with its observations, so localization differences are preserved rather than silently overwritten.

Top Free discovery signals

• Popular Now: sums 101 − rank across latest observed Top Free snapshots.
• Rising in 7 Days: sums exact upward movement only where both dates have comparable Final evidence and the App is ranked on both.
• New on Charts: counts verified New Entrant events during the latest seven completed Market Days.
• Long-running: counts Top Free rank observations over the latest 30 completed Market Days; average observed rank breaks ties.

These are discovery signals, not App quality ratings or endorsements. No App type or category is suppressed.

Corrections

A structurally valid Apple result is never edited merely because it looks unusual. A confirmed site-side collection, date-assignment, migration, or storage error may be corrected with a replacement version. The original is retained as Superseded, and the reason, operator, time, and before-and-after hashes remain auditable.

Limitations

Charts are observations of Apple's public feed at particular times, not continuous monitoring. Apple can update charts between attempts. Rank does not reveal downloads, revenue, retention, quality, safety, or user satisfaction. Market availability and source metadata may change. Partial Coverage, missing dates, and Provisional status should be considered before drawing conclusions.