Back from the Forum Entreprendre dans la Culture

Last week we were invited to the Forum Entreprendre dans la Culture, which takes place every year at the Paris-Belleville School of Architecture. This event, organised by the French Ministry of Culture, brings together dozens of artists, creators, researchers, entrepreneurs, experts and decision-makers at various round tables every year.

Arvester was honoured to be invited to take part in a round table entitled "Are there still humans behind the curation of cultural works?” Moderated by Emmanuel Vergès, Director of the Observatoire des politiques culturelles in Grenoble, this conference gave us the opportunity to speak alongside researchers Samuel Gantier (Université Polytechnique Hauts-de-France) and Heritiana Ranaivoson (imec-SMIT, Vrije Universiteit Brussel), as well as Manuel Moussalam, Director of Research at Deezer.

The aim of this round table was to examine the role played by humans in the curation of cultural works, and in particular services that use algorithmic personalisation to recommend works to their users. As well as going back over our background and what led us to analyse discoverability, we were able to share some more general thoughts on the sector.

• Extending the scope of our analysis to the music industry gives us the opportunity to adapt our tools and methods to the specificities and uses of discoverability in this field. This analysis differs greatly from that of the audiovisual sector, where the homepage is the main point of contact between content and its audience, making it easier to measure the visibility of works in a relatively small space (a grid of a few hundred thumbnails). Music, on the other hand, is used differently: 70% of listening is organic and results from personal searches (via playlists or the search engine), so any measurement becomes more complex and difficult to objectify methodologically.
• In the audiovisual sector, two trends are emerging: firstly, hyper-aggregation, as demonstrated by the recent partnerships between Netflix and TF1, and between Prime Video and France Télévisions; secondly, the expansion of content into other formats, notably video games. According to our data, the latter category is the most popular on Netflix, which illustrates that human curation is still very much present, even in ultra-personalised environments.
• Finally, one of the questions raised concerned the value of this data, both for public decision-makers and for artists. In terms of cultural public policy, it is important to objectify practices in order to inform public decision-making before considering regulation. For rights holders, knowing the visibility of their works could, for example, open the way to better remuneration.

I would like to thank Philippe Tilly for his invitation to the Forum and Emmanuel Vergès for preparing this round table.