Slicing and dicing the streaming revenue cake

In 2016, digital revenues made up 50% of the total revenue in the international recorded music business.  The share of streaming revenue across all digital income was 59%, and it keeps growing fast. In some forerunning territories like the Nordic countries, the share of streaming has already exceeded 80% of the overall revenues in the recorded music business.

Therefore, it is evident why all parties in the music business want to study in detail the current distribution methods of the subscription-based streaming services, and their fairness for each stakeholder group. Currently all major audio streaming services use the pro rata model in their distribution, but various right holder groups around the world have expressed a need to study the outcomes of alternative methods as well, such as the so-called user centric model.

As Finland is one of the leading streaming countries, it is natural that our local music organizations have wanted to compare the different distribution systems. The aim of the study “Pro Rata and User Centric Distribution Models: A Comparative Study” (2018) was to compare the current pro rata system and the alternative user centric system, using real data from a major streaming service, in order to see how these two methods affect distribution.

The overall income from an international music audio streaming service is shared between three main groups. The streaming service gets around 30% of the income and the rest is divided between the various right holders of recordings (55-60%) and works (10-15%).

In the pro rata model, monthly revenues are combined and shared between individual tracks according to the total amount of streams. In the user centric model, the calculation is based on the listening habits of each individual user, whose subscription fee is distributed only to the right holders of the tracks he/she has actually listened to.

The data used in the study consisted of Spotify Premium usage information in Finland during the month of March in 2016. This was, as far as I know, the first time Spotify had released this kind of data for public research purposes. The streaming report consisted of more than eight million individual streams. The researcher took a random sample from 10,000 tracks to represent the entirety of the material. This sample included tracks from 4,493 individual artists and 22,496 streams, which were played by 8,051 anonymised user IDs.

The key findings of the study demonstrated the basic differences and correlations between the pro rata and the user centric distribution models. The general trend is that as the overall stream count decreases, the revenue difference between the user centric and the pro rata models increases.

The pro rata model favours those artists and tracks who get the largest amounts of played streams regardless of whether those streams are created by a large number of users with fewer plays, or by a smaller number of users who have played them repeatedly.

The user centric model favours artists with smaller amounts of streams, especially when the overall stream count is smaller. However, it should be emphasized that the positive financial effect is not automatic in all user centric cases: instead, the outcome may even be the opposite. The results depend on the cumulative effects of both individual and user groups’ listening habits. In any case, the user centric model gives more immediate power to users to direct the money they pay for the service towards the artists or tracks they favour, compared with the pro rata model, which is not transparent from the user’s point of view.

However, the key findings of the study only represent the tip of the iceberg, as there are numerous key issues that affect all right holder groups of recordings. The following issues are amongst the most important.  

Firstly, the analysis could also have been done based on works instead of artists. If that had been the case, there would have been different kinds of variations available, such as several different recordings of a work by several different artists, works created by several composers and lyricists, and so on.

Secondly, it could also be researched whether subscribers choose tracks they listen to based on an artist, a work or a combination of the two, or simply prefer to use ready- or self-made playlists. Subscribers can also rely on each service’s algorithms, which learn the tastes of each user and offer recommendations or playlists based on listening habits. This would be a useful tool for estimating the importance of the user centric method for consumers’ willingness to pay for audio streaming services.

Thirdly, the current pro rata distribution system does not take into account the big differences between the durations of the various kinds of musical works. For example, the distribution process treats a play of a three-minute pop song in the same way as a play of a 15-minute classical music work. This is clearly not fair, and therefore some kind of adjustments would be required.

Fourthly, in most countries, it would be interesting to compare the shares of domestic and international artists, tracks and works. In smaller countries like Finland this is significant both from commercial and cultural points of view.

Regardless of the results of this study and the aforementioned topics, we should remember that music business is, at the end of the day, a business. The current distribution system is a result of business negotiations between different leading parties, and negotiating also remains the only way to change the system. It cannot be done by a single party such as Spotify. The companies who have the strongest leverage will dominate the system in the future just as they have done until this day.

These parties’ decision makers can only be convinced by demonstrating either the direct or indirect financial benefits the change would bring. One should also bear in mind that the costs of changing the distribution system have to be covered somehow. Furthermore, a change of method or increasing fairness as such do not automatically increase the amount of shareable money. It simply splits up the revenues in a different way. For the time being, the most likely scenario for a change is for the user centric system to find a way to make the bonds between artists and their fans stronger, which in turn will bring more paying subscribers for streaming services. The prerequisite for this is a transparent system, which proves to consumers that the money they pay is distributed to those right holders whose music has been played. This would be one of the most efficient ways to make the revenue cake grow and justify the efforts and costs of changing the current pro rata distribution system.

Jari Muikku is a consultant and the CEO of Digital Media Finland, and the author of the distribution method report.