Cineuropa spoke to Nina Vindum Rasmussen, researcher at the London School of Economics and Political Science and author of the article Friction in the Netflix machine: How screen workers interact with streaming data, published in May. Her study examines how screen workers interact with streaming data despite widespread secrecy. We asked her to share her findings with Cineuropa. The journal article can be freely accessed here: https://doi.org/10.1177/14614448241250029

(El artículo continúa más abajo - Inf. publicitaria)

Cineuropa: Could you introduce the scope of your study? Why do you believe it’s important to delve into the ways screen workers deal with streaming giants' data?
Nina Vindum Rasmussen: Data secrecy is one of the central power struggles of the streaming era. Over the years, it has been widely reported how streaming giants keep their viewing data under wraps. It was a major battleground in the Hollywood strikes last year with writers and actors demanding increased transparency to secure better contracts and remuneration. The issue is equally contentious in Europe, which is what my research examines. 

My study focuses on how European screenwriters, producers and directors navigate an opaque data environment. To this end, I have interviewed 30 screen workers from 11 countries about their collaborations with streamers, especially Netflix and Amazon. There is much to gain from studying both the direct and indirect encounters with streaming data in everyday production practices. 

How did you pick your participants? Can you elaborate on your methodology?
Creatives sign strict non-disclosure agreements with streamers, and many are – understandably – reluctant to break these. In addition, I was an industry outsider with a tragically empty list of industry contacts. I therefore had to be flexible in my recruitment approach. I sent emails to more than 100 workers and 120 trade unions. In every interview, I would also “snowball” new contacts by asking interviewees to refer me to others in their network. 

The actual interviews drilled down into the creative process and working conditions on specific streaming productions. Some participants were incredibly generous with what they shared. One showrunner shared snippets of notes they had taken during a meeting with a Netflix executive. Another writer sent me a questionnaire they had been asked to fill out by Amazon as part of the pitching process. Such insights proved invaluable. I also attended more than 20 industry events, which allowed me to contextualise the interviews within broader industry debates. 

What are your main findings? I see you label screen workers’ interactions with data by the way they “access,” “sense,” “generate,” and “resist” data...
Like most academics, I love putting things into categories. Across my interviews, I observed how screen workers’ discussions around streaming data could be split up into different “interactions.” First of all, every single participant would touch on access to data – or lack thereof. This depended on their previous experience, achievements and commissioner relationships. Some were not invited to meetings in which key metrics were revealed. However, interacting with streaming data does not depend exclusively on having direct access to facts and figures. Screen workers sense streaming data in other ways too, especially in the “notes” given by streaming executives. For example, one showrunner concluded that a Netflix executive’s use of imperative words like “must,” “want” and “need” revealed an underlying data-driven approach. I also think it’s important to consider how screen workers respond to streamers’ data-driven tactics by both generating their own data and resisting data altogether. In fact, several workers said that they did not care about metrics whatsoever. They just wanted Netflix or Amazon to order another season. All these interactions tell us something about the way streaming data and production practices intertwine.

Would you be keen to expand your analysis? If so, how?
Absolutely. My study casts light on the algorithmic logics of streaming and how they shape labour, creativity and content. Like many others, I am following the industry response to the most recent AI hype cycle. I’m especially interested in what it reveals about power asymmetries in the sector: for instance, who has access to AI tools, and what the use of them means for copyright and labour conditions. At the same time, I find it important to place the current debates around AI in historical context. Technological innovations have always shaped production practices. This is where the painfully slow nature of academic studies can be beneficial: it forces us to focus on changes and continuities. 

On a personal note, would you be in favour of full transparency? What type of effects do you think it may have on the industry?
Streamers have sought to address concerns around data obscurity by releasing selected viewership metrics in dribs and drabs. This cherry-picked information reveals little about how viewership drives programming decisions, though. Simplified viewership metrics make up a tiny fraction of the enormous, fast-flowing stream of behavioural data possessed by the likes of Netflix and Amazon. They don’t just know the number of hours we spend watching something – but also how we engage with it. From a business perspective, this behavioural data is a goldmine. But it’s vast and hugely complex. Full transparency therefore seems unlikely, also on a purely technical and practical level. That being said, there should be a strong incentive to provide creatives with enough insight to boost their bargaining power. Performance metrics form a key part of the puzzle when it comes to creatives receiving a fair share of the pie and sustaining their livelihoods. As such, increased transparency would strengthen the sustainability of the sector in the long run. That should be in the interest of global streamers too.