One of the main tasks of Content Control and Operations teams at distributors is to identify and prevent fraud. We can divide fraud into 2 main categories: Content Fraud and Consumption Fraud.

Consumption Fraud happens after a release has been approved and distributed, and consists of the artificial manipulation of metrics used by digital platforms (or DSPs) to inflate numbers and revenue, either through bots, playlists focused on playing the same songs over and over again, or streaming farms.

Artificial Streaming has grown so much in recent times, that platforms are taking new measures to combat it. For example, Spotify is fining 10 Euros per month for each song that has been detected with high levels of Artificial Streaming.

For this reason, we developed a tool in Backstage, in which White Label account managers can view and take action, which we call the Smart Artificial Stream Filter.

This tool uses filters to detect anomalies in the daily analytical reports that we receive from digital platforms.

To do this, we must go to the Analytics section within Backstage and activate the Artificial Stream smart filter.

There we can even filter by date ranges and by country.

By applying this filter, the system will automatically show us all the results in which we detected anomalies.

Each row of the results corresponds to: A track (ISRC), on one day, on one platform, in one country.

Example: If the same track has streams with anomalies on 3 different days, there will be 3 results. But if that happens in 2 different countries on the 3 different days, there will be 6 results, and if there are also 3 different platforms, there will be 18 results.

Detecting anomalies with the smart filter does not mean with 100% certainty that these streams are artificial, but they do represent a high-risk signal, and at the very least they must be manually verified by the distributor and consulted with the user who owns the song.

The smart filter is not perfect and is constantly being updated, so we cannot guarantee with 100% certainty that all artificial streams will be infallibly detected by the tool.