Netflix spent $17 billion on content in 2023. It spent an estimated $3 billion on technology infrastructure. Two years later, the ratio is shifting — not because content budgets are shrinking, but because the competitive advantage has moved upstream.
The new battleground is the moment before a viewer presses play. Personalisation, thumbnail selection, content tagging, and watch-time prediction have become as strategically important as any original series.
For traditional broadcasters entering this world late, the gap is not just technical. It is cultural. Building a recommendation engine requires a tolerance for failure and iteration that is alien to organisations built around the logic of a weekly schedule.