The way he thinks about things is it's not a threat to his business to have competitors in his space not just because he is prepared to undercut them on cost but because he knows that his true focus is on customer satisfaction. And competition, whether through providing customers with a choice or encouraging his own systems to provide better services is good for his business.
But in this case it's really just because Netflix is too insignificant to matter to Amazon overall. It's just a tiny fraction of their revenue and what they're trying to do.
And if you want to be Machiavellian of course it's good because then Amazon has leverage of Netflix.
Seriously how would it look if Amazon turned down a company because they were competing? It would be an admission of defeat and weakness for AZ. That's not how an "apex predator" thinks.
Keep your friends close, and your competitors....
It looked shitty when it happened with Amazon and Apple
The latter decision has been reversed now though https://techcrunch.com/2017/06/05/amazons-prime-video-app-is...
I found this to be really stupid behavior of Amazon. Seemed like they only wanted to push their competing product (Fire TV).
It's good they figured things out, though. A Prime Video app was announced at WWDC.
It would look bad: https://www.wired.com/2015/10/amazon-apple-tv-chromecast/
Amazon's selling. Netflix is buying. It'd be stupid to refuse the money. Netflix wouldn't lose much sleep over it–at their size, building their own infrastructure can't be much worse than renting.
Plus you'd get the idea out there that Amazon isn't the sort of reliable utility company their marketing suggests. They have their fingers in so many markets, a third of their customers would start to fear being cut off.
AWS is also a separate division within Amazon, and probably not lacking self-confidence. The people are judged by the numbers of AWS, and they'd probably resist attempts to get them to forgo revenue just for the sake of some other division's feeling.
"Netflix will most likely exist even if we don't provide them server services as there exist other cloud providers and / or Netflix is big enough to roll their own, so we might as well have a piece of our competitor's success by having their infra spend come to us."
Here is a blog post (2010) describing the decision to move out of our own data centers:
https://medium.com/netflix-techblog/four-reasons-we-choose-a...
So they already have their own system. As as some below have mentioned, they could simply move to Azure if they wanted without much trouble.
That's not to say it might not be a worthwhile approach over the long run but it would be a huge strategic investment in something that has essentially nothing to do with what they view as their core differentiating service to customers.
Would Netflix have done things differently if they had known they'd reach this level of usage? Perhaps. But disentangling now would be very difficult and almost certainly disruptive--including a likely reduction in quality of service.
And disentangling be disruptive? Stand up your new datacenter and shift your apps over with DNS when they're ready to serve production traffic. If you can move traffic between regions, you can move it between on prem and cloud.
A lot of big companies that compete with Amazon in some vertical have a no-Amazon policy for infrastructure. Some even require their business associates (vendors, etc) commit to contracts to do the same.
Amazon seems to want to be in every business sector there is, so almost any company could end up competing with them.
This post has two contradictory quotes in my opinion:
"It’s worth noting that this AWS price wouldn’t be what Facebook paid in this hypothetical situation. Much like Snapchat and Netflix, Facebook would be a heavy and influential user"
"Facebook’s years of specialization for running Facebook are in contrast to AWS, whose storage is designed for multi-purpose (albeit-heavy) use." - Amazon already has several tiers of storage (S3, S3 - Infrequent Access, Glacier, CloudFront)... who is to say that they wouldn't be motivated to introduce more classes by a large customer like FB.