A study to objectify the debate with comparable data
With the evolution of music listening habits, online music services have experienced very strong growth in recent years, reaching 59% of the recorded music sales market in 2019 (SNEP study, The economics of music production in 2019, 2020 edition). Their growing weight in the music
economy has raised questions about how these companies allocate revenues from subscriptions to their services.
➔ Two distribution models
Revenues generated by streaming are now distributed to the rights holders in proportion to their market share (number of streams generated for the rights holder’s catalogue compared to all the streams generated on the platform), according to the Market Centric Payment System (MCPS) model. By definition, this model of distribution favours artistic proposals with the most committed audience and contributes to boosting revenues from streaming to titles listened to by heavy users
of online music services.
For several years, music professionals have been campaigning for an alternative model, known as the User Centric Payment System (UCPS), which consists of distributing the tax-free amount of each subscription according to the user’s actual listening: the user’s
subscription is distributed only to the rights holders of the tracks he or she listens to.
➔ An objective and independent study
Several studies1 have been carried out in France and Europe on the consequences of switching from one model to another for streaming revenues and diversity of supply. However, they have been carried out on the basis of different methodological choices, data perimeters and temporalities.
Therefore, their results, which are often contradictory, cannot be compared and do not allow this reflection to be objectified.
In this context, the Centre National de la Musique, as an observatory of the music sector in France, has carried out a study piloted jointly with the Ministry of Culture. Launched in April 2020, this study aims to carry out a comparative and objective analysis of the implications of the two systems, particularly with regard to their impact on diversity. Conducted with the support of Deloitte, the study was carried out in a concerted framework with professionals, thanks to the contribution of representatives of different professions in the sector: artists, producers, labels, distributors, collective management organisations and online music platforms.
Some online music services have, from the outset of the study, indicated that they do not wish to participate in it, or do not have the resources to process their data for the implementation of a common methodology. Only Deezer, Spotify and Sacem agreed to participate in the provision of
data, the development of the common methodology and, for both platforms, its execution2.
Main findings of the study
Beyond a few methodological issues deliberately left outside the scope of the analysis (see presentation of the method in our file), the study leads to the following conclusions:
- The switch to UCPS would make it possible to bring the distribution of income into line with the respective weight of the different types of consumers (according to the number of streams) and would limit the effect of income being channeled towards streams by heavy users. However, in an economy of access and not of possession, the question of whether one model would be more efficient or fairer than another remains a matter of principle or commercial strategy, which these figures can neither validate nor invalidate;
- The transition to the UCPS could favour a redistribution of income for the artists, titles and musical genres with the lowest audiences, but, if the percentages of evolution seem, not negligible, the amounts in value actually remain limited, in the state of the market: thus, beyond the 10,000th most streamed artists, all musical genres combined, the impact of the transition to the UCPS would be a maximum of a few euros per year on average per artist;
- The change to the UCPS could encourage a significant redistribution between musical genres with large audiences, to the detriment of rap and hip hop and in favour of rock and pop;
- The switch to UCPS could help to increase the market share of back catalogue (tracks posted on a platform more than 18 months ago);
- The question of the costs and complexity of the transition to UCPS remains to be clarified. Assuming that the data exchange interfaces remain unchanged, the development of the UCPS model would be the responsibility of the platforms. The associated costs estimated by the two partner platforms of the study are very disparate and should be estimated more precisely. Moreover, these costs may not be absorbable for smaller platforms and could be passed on to the entire value chain. Rights holders (distributors, producers, collective management organisations) could also bear the costs of verifying the reports submitted by the platforms (complex operations linked to the weightings applied at the user level for UCPS calculations);
- The impact of recommendation tools on the distribution of income in a UCPS model must also be assessed: fears have been expressed by some representatives of rights holders regarding recommendation algorithms, the operation of which is considered opaque. The quantitative analysis of the distribution of value between recommended and autonomous streams is complex and requires a common and shared definition, which is currently lacking;
- The fight against fraud: the UCPS model would make it possible to reduce the impact of one of the existing frauds, which consists of deploying a large number of streams for targeted tracks and artists, for example by “click farms”. With the introduction of UCPS, fraud could evolve towards targeting low-intensity or inactive users or even hacking into sub-accounts within bundled offers. The fight against fraud is one of the main challenges of music streaming, and continuous vigilance of platforms to detect fraudulent streams is essential and greater transparency is again necessary.
The study objectifies a number of points thanks to the application of a transparent and clear methodology. This is an important first step that all actors must take up in a logic of solidarity and
The CNM intends to continue and amplify this work in close consultation with all the players and for the benefit of the entire music industry in order to:
- make progress on the data that can be mobilised and ensuring greater transparency:
some stakeholders did not wish to communicate certain information that they had, or even
did not wish to participate in the study, which made it difficult to carry out, but it was also
noted that some data were not available or could not be used in a satisfactory manner;
- develop a better observation of practices and usages on online music services and a good
understanding of recommendation tools, favoring the implementation of good practices in
the service of diversity;
- work jointly on the development of the streaming market in France, which is lagging behind its European neighbours, by ensuring the sustainability for all players of their economic models and by promoting the diversification of audiences and uses.
The CNM has undertaken an extension of the observatory of musical diversity produced and
exhibited, which will be implemented during the course of the year 2021 and which will allow it to
make progress on the provision of data that is currently lacking. It also proposes to initiate a
prospective reflection with all the players in the sector on the sustainable conditions for the development of streaming, on the transparency of recommendation methods and on the fight
against fraud, for which it could play the role of a trusted third party.
 “User-centric settlement for music streaming”, Clouds and Concerts, March 2014.
“Music streaming in Denmark: An analysis of listening patterns and the consequences of a “per user” settlement model based on streaming data from WiMP”, Roskilde University, 2014.
“Pro Rata and User Centric Distribution Models: A Comparative Study” Digital Media Finland, November 2017.
Unpublished internal studies: Deezer, Spotify, Sacem, Merlin
 This execution could not be carried out homogeneously. Deezer carried out the common methodology on the basis of a complete
sample (year 2019 and completeness of the parameters) and within the defined parameters. S
Spotify, for its part, initially implemented its own methodology, with a scope that included all users, paying and non-paying, as well as
different axes and analysis parameters.Nevertheless, in order to be able to compare the results with those of Deezer, the common
methodology was carried out on a random sample of 100,000 users provided by Spotify on French consumption in the first half of