These 62 releases aren't missing from a spreadsheet — they're missing from the client's field of view. Artist+title search returns noise on generic compilation titles (“Summer Hits,” “BBQ Classics,” “Various Artists”); the only reliable path is ISRC-level relationship mapping. And each work isn't just one recording: on average, 2,121 related recordings — covers, live versions, samples, remasters by other artists — sit under the same work in MLC. Every one of them is a potential source of royalties tied back to the same rights.

The official channel isn't where the audience actually is. YouTube's Topic channel — auto-generated, unmanaged, with barely 5,000 subscribers — converts views 28 times more efficiently than the official one, and just over half of the artist's top-performing videos sit entirely outside her own channels. The pattern even shifts by content type: Feat content drives the majority of views on owned channels (27.8%), while lyric videos overperform on external ones (14.3% vs 9.9% for Feat) — acquisition and retention are happening in different places, through different content. None of this is visible from an artist's own channel dashboard, and with 54 publishers spread across just 39 works, knowing exactly where value is created matters even more.

Clean-looking metadata doesn't mean fully collected rights. This is the gap between the data existing and the money actually reaching the rights holder — invisible until someone cross-references collection societies against DSP-level matching, track by track.