Music Usage Monitoring Explained
How monitoring actually works, what confidence scores mean, and why no system can honestly claim to monitor everything everywhere.
Monitoring answers one question: where did this music appear? Everything else — matching, reconciliation, claims — depends on getting that answer with enough context to be useful.
How detection works
Audio fingerprinting reduces a recording to a compact signature that survives compression, EQ and broadcast processing. Monitored streams are fingerprinted continuously and compared against a reference database. A match produces a detection with a timestamp, a source and a confidence score.
What confidence actually means
A confidence score expresses how strongly the audio matched a reference, not how certain the commercial conclusion is. High confidence on a short segment under heavy dialogue still needs review. Any platform that reports 100% accuracy is describing a wish, not a measurement.
Coverage is the honest constraint
- Sources differ: broadcast, radio and digital services each expose different data.
- Territories differ: a broadcaster monitored in one market may not be monitored in another.
- Periods differ: historical coverage is usually thinner than current coverage.
Detection is not the deliverable
A detection only becomes valuable once it is matched to a work, an owner and a territory, then compared against what was reported. Monitoring on its own produces a longer report. Reconciliation produces a decision.