Netflix are sending emails to users accusing their ISPs of having poor performance, or even insinuating that they are throttling user services. This is disingenuous.
It's also a big deal. Netflix regularly accounts for around 30% of all internet traffic on any network in the US and provides no revenue to them (although it has recently been entering into 'side' agreements with various networks).
But for Netflix to name and shame, for example, Verizon if it believes that its network is underperforming is disingenuous.
The problem lies at the core of internet technology, where not a single layer was designed with video in mind.
If my video is buffering it could be a problem with my ISP, in which case all video would buffer. But what if it's just Netflix' own issue?
N
It could be my ISP has poor peering with Netflix' network. This means that Netflix hasn't arranged to have enough capacity to deliver to its customers' networks. So they can deliver 1Gbps to ISP1, but there are 5000 viewers at 1Mps. That's a contention of five to one. When you deliver a third of the US' traffic this may be a particular bottleneck.
Of course, Netflix don't handle live traffic, so their delivery is spread over multiple files, still the peering is an absolute throttle.
But what is really interesting in the symbiotic relationship in content delivery that is evolving.
In the UK Virgin Media used to be a content provider, channel operator, TV service and ISP. Then a new CEO decided that being the best ISP would be its focus and it shed the other businesses.
Meanwhile Sky has been moving in the opposite direction, building channel and content capabilities and extending its services and offering around improved broadband.
BT has been furrowing a middle ground, using sports to build a viable content proposition, but failing to leverage its TV service due to a woeful inability to invest properly in a decent broadband delivery, and, for many of its customers, using a tokenised model that is based on very short term metrics that enable it to maintain and milk its monopoly.
Meanwhile TalkTalk is somehow building a serious TV customer base (I have to admit to not knowing how the heck they are doing this) and Freeview is suddenly catching up by offering a boxed service.
So, in the UK you now have four potential service providers. But the reality is that all that counts are two things: the rights you have and the bandwidth you can provide.
It's also a big deal. Netflix regularly accounts for around 30% of all internet traffic on any network in the US and provides no revenue to them (although it has recently been entering into 'side' agreements with various networks).
But for Netflix to name and shame, for example, Verizon if it believes that its network is underperforming is disingenuous.
The problem lies at the core of internet technology, where not a single layer was designed with video in mind.
If my video is buffering it could be a problem with my ISP, in which case all video would buffer. But what if it's just Netflix' own issue?
N
It could be my ISP has poor peering with Netflix' network. This means that Netflix hasn't arranged to have enough capacity to deliver to its customers' networks. So they can deliver 1Gbps to ISP1, but there are 5000 viewers at 1Mps. That's a contention of five to one. When you deliver a third of the US' traffic this may be a particular bottleneck.
Of course, Netflix don't handle live traffic, so their delivery is spread over multiple files, still the peering is an absolute throttle.
But what is really interesting in the symbiotic relationship in content delivery that is evolving.
In the UK Virgin Media used to be a content provider, channel operator, TV service and ISP. Then a new CEO decided that being the best ISP would be its focus and it shed the other businesses.
Meanwhile Sky has been moving in the opposite direction, building channel and content capabilities and extending its services and offering around improved broadband.
BT has been furrowing a middle ground, using sports to build a viable content proposition, but failing to leverage its TV service due to a woeful inability to invest properly in a decent broadband delivery, and, for many of its customers, using a tokenised model that is based on very short term metrics that enable it to maintain and milk its monopoly.
Meanwhile TalkTalk is somehow building a serious TV customer base (I have to admit to not knowing how the heck they are doing this) and Freeview is suddenly catching up by offering a boxed service.
So, in the UK you now have four potential service providers. But the reality is that all that counts are two things: the rights you have and the bandwidth you can provide.