Original production test ledger · 10 minute read
What happened when PlaylistXfer tested 439 Apple Music matches.
These are dated production observations, not a universal success-rate claim. Every source is public, every scope is stated, and every limitation stays next to the number it affects.
Three production runs analyzed 439 Spotify rows against the US Apple Music storefront. The reports classified 433 as confident matches and six as unmatched. One run intentionally sampled only the first 50 rows of a 494-track source, so its result does not describe the other 444 rows. No Apple Music account was connected and no destination playlist was created during these tests.
The dated production test ledger
Each row below came from the live playlistxfer.com transfer path on August 14, 2026. The preview read the public Spotify source, and the analysis searched the Apple Music catalog. Testing stopped before the write step. This separation matters: catalog search can be exercised without asking for a listener's Apple Music library permission.
| Public source | Source scope | Analyzed | Confident | Unmatched |
|---|---|---|---|---|
| FIFA Songs | Full readable source | 339 | 333 | 6 |
| Daily Mix test | Full readable source | 50 | 50 | 0 |
| Rockabye Baby: Lullaby Covers | First 50 of 494 | 50 | 50 | 0 |
The combined 433-of-439 total is a description of these three scopes on this date. It should not be read as “PlaylistXfer always matches 98.6%.” Catalog availability, playlist edits, storefront, metadata, duplicate removal, and the selected review threshold can change a future run.
Test 1: the full FIFA playlist report
This source was chosen because it had already exposed a completeness bug. On August 2, the original Cloudflare implementation returned 46 readable rows while the ingestion backend returned 340. That discrepancy led to a production fix and a separate incident report.
On August 14, a fresh production read returned 339 unique readable Spotify rows. The live playlist can change, and PlaylistXfer removes duplicate Spotify track identifiers before creation, so the current count is a snapshot rather than a permanent property of the source. The important check was internal consistency: all 339 preview rows entered the full analysis, and the completed report also contained 339 rows.
Apple Music matching classified 333 rows as confident and six as unmatched, for a 98.23% confident rate within this specific report. The unmatched source rows were:
- Colors - Coca-Cola® Anthem, 2018 FIFA World CupTM by Jason Derulo
- Genius (feat. Sia, Diplo, and Labrinth) by Sia, Diplo, Labrinth, and LSD
- Brócolis by TAMEC, Lizinmec, Novato Mc, Blanko, and Gusttap
- Espresso (Remix) by AVLS
- CASSANDRA // K&L Session #001 by K&L and Cassandra Norlander
- Falling Down by Sustain
The safe behavior was to leave those six rows out. A title-only substitute might have produced a higher displayed percentage while selecting a different recording. PlaylistXfer treats an honest omission as better than a confident-looking wrong song.
Test 2: a complete 50-track source
The public playlist named Daily Mix test returned 50 readable tracks, all 50 with the metadata required by the preview. The analysis scope was Full playlist, so the report was not truncated or sampled. All 50 rows received a confident Apple Music candidate in the US storefront, with no review or unmatched rows.
This test is useful because it exercises the normal product path without the scale of the FIFA source. It checks that a public link can move from preview to a complete match report without Spotify login, that preview and analysis counts remain equal, and that the service does not require Apple Music write permission merely to show candidates.
It does not prove that every 50-track playlist will match perfectly. This source contains commercially released recordings with strong identifiers. A playlist containing local files, unavailable regional editions, unofficial remixes, podcasts, or ambiguous live versions should be expected to produce a different result.
Test 3: an explicitly limited sample from a 494-track source
Rockabye Baby: Lullaby Covers returned 494 unique readable rows in the production preview. Instead of presenting a quick test as a full-playlist result, the analysis was deliberately configured for the first 50 rows. The report preserved both counts: 494 in the original source and 50 in the analyzed scope.
All 50 sampled lullaby-cover recordings received confident candidates. That establishes that the sample completed cleanly; it says nothing about the remaining 444 rows. The distinction is visible because PlaylistXfer labels partial analyses and limits playlist creation to the rows that were actually analyzed.
A declared sample demonstrates how the product behaves when someone chooses a faster test mode. Hiding the original count would inflate the evidence. Showing both the 494-row source and 50-row scope lets a reader understand exactly what was and was not checked.
How the tests were run
- Use the normal public Spotify playlist URL shown in the ledger.
- Send the URL to the production preview endpoint and record the playlist ID, source count, returned row count, and source limitations.
- Start the production Apple Music analysis using the US storefront and the stated Full playlist or 50-row scope.
- Wait for the saved job to report complete, then record its original source count, analyzed count, confident count, review count, and unmatched count.
- Confirm preview and analysis completeness for every run described as full.
- Stop before Apple Music authorization and playlist creation.
PlaylistXfer prioritizes recording identifiers when they are available and uses title, artist, album, duration, version words, and catalog context as supporting evidence. A candidate that does not clear the confidence threshold stays in review or unmatched instead of being silently counted as transferred.
The public Spotify ingestion surface can change. PlaylistXfer therefore records the source adapter and its limitations in the result. Duplicate Spotify track IDs are removed for safer destination creation, and rows without readable track identifiers or metadata may be excluded. Those rules explain why a current readable count can differ from a number displayed by Spotify or from an earlier dated snapshot.
What these numbers do not prove
No human listened to all 433 selected Apple Music recordings during this audit. “Confident” means the catalog evidence cleared the product threshold; it is not a guarantee that every edition, mix, clean version, or remaster will be the listener's preferred choice. The match report remains the review surface, not a certificate of musical equivalence.
The tests used the US storefront. A recording available there may be missing elsewhere, and the reverse is also possible. Results can change when Spotify edits the source, Apple changes catalog availability, labels replace releases, metadata is corrected, or PlaylistXfer improves its scoring.
Finally, the audit did not create a destination playlist. That was intentional. Reading a public source and searching a public catalog can be tested without changing a person's library. The final write step requires the user to connect Apple Music, review the report, and explicitly request creation.
PlaylistXfer is independent and is not affiliated with, endorsed by, or sponsored by Spotify or Apple. Public playlist names and track metadata are included only to make interoperability tests reproducible.