by gizzlon 7 hours ago
> ranking content based on clicks and minutes watched.
I suspect they just push what they want you to watch, like their own content. Seems that way to me at least, based on their quite shitty "recommendations"
> I suspect they just push what they want you to watch, like their own content.
Having worked close to the recsys folks at Netflix, I can tell you that this statement couldn't be further from the truth.
Why do they care what you watch? I expect they pay a flat fee to license content (if not, how is that policed?) so the marginal cost to them is the same no matter what you watch.
I'd guess they push you to their content for the same reason they make that content in the first place: they believe you'll like it and keep watching it.
Ad placement is one wrinkle that would incentivize promoting their own content, but I don't get the impression that's big enough to make the difference at the margins.
>I expect they pay a flat fee to license content
I wouldn't tbh, though I'll admit I'm speculating solely on public information. During the 2023 strikes, SAG-AFTRA and the WGA negotiated additional residuals based upon whether 20% of the streaming services subscriber base viewed the content within 90 days of release.[1] So, streaming platforms are evidently willing to share subscriber viewership data with 3rd parties if it's a contractual requirement.
I would be surprised if content licensors haven't negotiated an as good or better deal for themselves.
[1] https://variety.com/2023/biz/news/sag-aftra-streaming-bonus-...
If people are watching their content, they can rely less on licensed content and drive those costs down. It's a similar value prop to any vertical integration.
If you mostly use Netflix to watch licensed content, you're more likely to cancel when all the licensed content is removed from the catalog.
If they successfully steer you towards Netflix produced content, you're less sensitive to what happens to the licensed content.
The story I heard about most Netflix content going for very long is that after two seasons a show's cast unionizes and they didn't want to pay up and they'd rather cancel shows, which seems awful penny-wise pound foolish of them.
I think they also used their metrics to figure out people liked kevin spacey (whoops) - and created house of cards - which catapulted netflix's production side.
https://medium.com/@danial.a/how-netflix-used-data-to-create...