When House of Cards premiered in 2013, it did more than introduce audiences to Kevin Spacey’s ruthless Frank Underwood. It became the symbol of a new Hollywood, one where conventional wisdom claimed algorithms had replaced executives and data had become the industry’s most powerful storyteller. The tale was irresistible. Netflix supposedly looked at its viewing data, discovered that subscribers loved political dramas, David Fincher, and Kevin Spacey, then built a blockbuster around those insights. More than a decade later, that story has become one of Hollywood’s most repeated legends.

But a new piece of research suggests the legend was never quite true.

The Hollywood myth that refused to die

Speaking on The Town with Matt Belloni, Paris 8 University sociology professor Violaine Roussel discussed her new book Data-Driven Hollywood: The New Data Professionals in the Age of Streaming, which draws on interviews with 75 data scientists working across major studios and streaming platforms. When Belloni brought up the famous House of Cards anecdote, Roussel laughed before answering plainly.

“Yes. It is a tale. It is a tale that people tell.”

She added that while Netflix knew Kevin Spacey performed well on the platform, “I do not think it got that granular at that point,” noting that the company’s data teams were still “very, very small.”

That distinction matters because the popular version of the story suggests Netflix’s algorithms effectively ordered a political drama starring Spacey. Roussel’s research paints a much more nuanced picture. Her book explains that early streaming-era data teams were limited in size and influence, serving more as strategic advisers than creative decision-makers.

Through interviews and observations inside companies including Netflix, Amazon Prime Video, and Disney+, she argues that modern Hollywood relies on a combination of analytics, executive judgment, market realities, and creative instinct rather than algorithmic certainty. Netflix’s well-known “taste clusters” help organize audiences with similar viewing habits, but they are one input among many when evaluating projects.

The real story is far more fascinating than the myth.

From House of Cards to taste clusters: How Netflix’s data evolved

Netflix certainly embraced data long before most traditional studios, but its analytical capabilities have expanded dramatically since the House of Cards era. During the early 2010s, the company primarily relied on subscriber behavior, recommendation systems, and viewing trends. Today, sophisticated machine learning models analyze everything from completion rates and rewatch behavior to audience segmentation, retention, marketing effectiveness, and content discovery. Yet even with far more advanced technology, executives have consistently maintained that data informs decisions rather than replacing creative leadership.

That evolution reflects a broader shift across Hollywood. Data departments have grown from relatively small technical teams into influential divisions that work alongside production, marketing, and programming executives. Roussel’s research argues that the industry’s biggest transformation is not that algorithms choose what gets made. Instead, data professionals have become part of the conversation, providing evidence that complements, rather than overrides, creative judgment. The enduring House of Cards story survived because it perfectly captured Silicon Valley’s disruption of Hollywood, even if the reality was always considerably more complicated.

What do you think? Share your thoughts on whether streaming platforms should lean more on data or continue trusting creative instincts.