Anomaly Detection is a new feature in beta release
We're working on getting it ready for you. In the meantime, expect changes to this feature before its final release.
You can monitor the accuracy of your royalty data with Anomaly Detection in Royalty Reports > Explore. Use Anomaly Detection with Royalty Trends to catch data quality issues, like unusual changes in reporting volume or missing data, before you submit your royalty report to SoundExchange or other recipients.
Anomaly Detection is Advisory Only
Anomaly Detection finds only potential anomalies. It does not modify royalty reports. Not all anomalies indicate reporting issues, and some reporting issues may not be identified by Anomaly Detection.
Review Anomaly Detection Results
Create a report in Royalty Reporting > Explore. See Royalty Reports.
Wait for the anomaly analysis to complete, typically a few seconds.
If anomalies are detected, expand Anomaly Detection to review.
Understand Anomalies
Anomaly Detection highlights unusual patterns in your royalty reporting data. Use its results as a starting point for investigation before exporting or submitting reports.
Anomaly Detection uses a statistical model to compare your selected data to rolling averages and year-over-year trends. Because of this, it might also flag normal, legitimate variations:
Business events
Audience changes
Programming adjustments
Holidays and seasonal trends
Each anomaly gives this information:
Severity
Anomaly type
Description
Affected reporting period
Expected value
Actual value
Anomaly: Performance Decrease
Performance counts are significantly lower than expected when compared to historical trends, rolling averages, or similar reporting periods.
Common Causes
A decrease in audience size
Missing or incomplete reporting data
Changes in programming schedules
What to Look For
In Royalty Reports > Trends, compare Total Performances and Play Frequency charts for the same dates.
If Play Frequency also decreased during the same period, the lower value may be explained by fewer songs being played. For example, a station carrying a live sports broadcast or special programming may naturally generate fewer performances because fewer tracks were aired.
If Play Frequency remained stable while performances decreased, investigate potential audience or reporting changes.
Anomaly: Performance Increase
Performance counts are significantly higher than expected when compared to historical trends, rolling averages, or similar reporting periods.
Common Causes
An increase in listener audience size.
A special event, promotion, or programming change that attracted more listeners.
Expanded station coverage or distribution.
What to Look For
In Royalty Reports > Trends, if Play Frequency remains consistent but Total Performances increases significantly, the spike may be related to increased listenership rather than a change in programming.
Anomaly: Play Frequency Anomaly
Play frequency is unusually higher or lower than expected based on historical listening and playback patterns.
Common Causes
Holiday programming
Special programming events
Live sports broadcasts
Talk-radio segments
Schedule changes that alter music rotation patterns
What to Look For
In Royalty Reports > Trends, review the Play Frequency charts for the affected period.
A decrease in Play Frequency often leads to fewer performances, while an increase in Play Frequency may contribute to higher performance counts.
Determine whether the change reflects an intentional programming decision or an unexpected reporting issue.
Anomaly: Missing Data Gap
One or more reporting periods contain no reported data when activity would normally be expected.
Common Causes
Metadata delivery interruptions
Station automation or encoder issues
Data ingestion failures
Temporary station outages
What to Look For
Review both Realtime Listening and royalty reporting activity for the affected period.
If the station remained online and listeners were present, a this anomaly may indicate that metadata was not successfully delivered.
If the station was offline or unavailable during the affected period, the detected anomaly may reflect an expected interruption in service.
Anomaly: Coverage Decrease
A significant decrease in the number of stations, streams, or reporting sources contributing data for the selected reporting period.
Common Causes
One or more stations stopping data delivery.
Changes in station configurations.
Reporting interruptions affecting a subset of stations.
Temporary outages impacting reporting coverage.
What to Look For
Review a royalty report for specific stations or reporting sources to confirm the presence of contributed data during the affected period.
This anomaly may explain related Performance Decrease anomalies because fewer reporting sources can result in lower reported performance totals.
Anomaly: Level Shift
A sustained change in baseline behavior across the report period. This anomaly is triggered when at least 40% of the days in the analysis period, with a minimum of 5 flagged days, show the same anomaly.