[bayesian_horse]

Youtube is better at grabbing the users' attention... Not necessarily with better content.

Though I learn more from Youtube than from Netflix, in general...

[bilbo0s]

Again though, Netflix is not there to grab a user's attention, Netflix is there to grab a user's money. We shouldn't conflate these two base motivations.

If netflix gets you to pay the monthly fee, while at the same time you only use a quarter of the bandwidth of most other netflix users...

well, I'd imagine that's a win for netflix.

Again, we'd need more of Netflix' internal data, but I wouldn't be surprised if a full modeling would reveal that netflix wants to optimize viewing time to be something less than 100% of some break even viewing level. I have to believe that people using Netflix more than a certain amount in a month end up costing netflix instead of benefiing netflix.

But again, we really need access to their internal data to say that for sure.

[jsnell]

Bandwidth is cheap, when the users are actually paying customers. Churn, on the other hand is very expensive. I bet Netflix prefers heavy users over those who watch two movies a month: the latter people are much more likely to cancel the service.

[bilbo0s]

It's been said before, but we'd need the internal data regarding viewing patterns, license fees (which are per user in many cases), bandwidth costs, etc etc, but I'd take your proposed wager. Without hesitation. I can think of very few scenarios where the amalgamation of these per user costs is lower for a heavy user. (Not saying it can't happen, just saying based on what I know, I'd feel pretty confident betting that it doesn't happen.) Additionally, throwing in what I know about consumer behavioral patterns, I'd wager even their light users are extremely unlikely to cancel Netflix.