Monday, February 24, 2014

The really strange Comcast-Netflix deal

Advanced warning: the first comment here linked to an article which clearly understood this better than me. I am still perplexed - but this post should be taken with a grain of salt.



Verizon took the FCC to court on net neutrality and won.

The big potential loser was Netflix. Under the old rules [the ones that the court overruled] tier one internet providers like Comcast or Verizon had to deliver Netflix packets the same way as they delivered any other packets (on landlines anyway) - and that meant that they could not selectively slow Netflix packets. If they were not allowed to selectively slow Netflix packets then they couldn't charge Netflix for the privilege of having their information delivered.

The argument for slowing them is that Netflix uses so much - and - so say Verizon and others - they should "pay their share". [Of course the counter argument is that when I buy my capacity I am paying for data to be delivered to me - and I have paid for all data, not all data except Netflix...]

Winning the court case had to be a positive for Verizon (that is why they took it to court). And it had to be a negative for Netflix (the most likely party to be charged). You wouldn't notice that in the stocks though - Netflix is up hard, Verizon not so much.

Comcast had more reason to dislike Netflix than Verizon though. After all Netflix competes with Comcast's core service (video delivery) and Reed Hastings of Netflix used to complain endlessly about Comcast's breach of net neutrality rules.

However Comcast - as part of its deal in purchasing NBC - agreed to follow net neutrality rules until at least January 2018 even if the FCC lost the net neutrality issue in court. After the FCC lost in court there was only one company that could not legally selectively slow Netflix content - there was thus only one party which Netflix did not have to deal with.

That party was Comcast.

And so it is really strange that the first deal Netflix did was with Comcast.

Netflix has just given Comcast a few million dollars - and seemingly for no reason.

Now Reed Hastings is a clever man - and he is not in the business of giving away millions of dollars.

So I presume he did it for a reason.

And I think it is this. The reason is to set a price. A low price. The lower the better - because after this they have the serious negotiation with Verizon. [And Verizon took the government to court and won - which leads you to think that Verizon is likely to be aggressive.]

Comcast knew they were not entitled to a brass razoo - and Netflix was giving them a few million dollars. From Comcast's perspective that was easy [take the gift]. The amount looks low. Comcast manages to look reasonable (and I don't normally put the words reasonable and Comcast in the same sentence).

And Netflix sets a low price.

It's a negotiating ploy.

But I don't believe it will work.

Lowell McAdam is unlikely to fall for it. He is not going to settle for single-digit millions. The tables have turned and the carriers are getting market power over Silicon Valley - and this is probably the quickest and easiest place to exert some power.

I think Reed Hasting wasted his few million.

Seeking different opinions as I have been wrong often.


No position in Comcast, Netflix. Long Verizon.



Anonymous said...

This is the simplest explanation I have read so far:

Kirit Vora said...

If Reed Hastings wants to raise the price for Netflix subscribers, this will give him a good excuse. If the price were to go up by $2 per month, Netflix can pay a part of it to the ISPs, and still come out ahead.

Anonymous said...

Another opinion that came through TWTR feed

Anonymous said...

I did early work as a network engineer with ISPs and have kept up with the chatter since then.

Larger internet providers make private arrangements to exchange internet traffic between their networks. Historically the most expensive links to build and maintain have been the long haul links across the country (or the ocean) and an ISP could save significant cash by dropping traffic onto a peer as quickly as possible. As a result networks with significant long haul bandwidth were always more sought after for private interconnects. And since networks with significant long haul capacity were in the business of selling bandwidth as well they were typically fairly picky as to who they would peer(interconnect) with.

The initial peering arrangements used to be done by network engineers over drinks at networking conferences like NANOG and most of the negotiations were around where the traffic would be exchanged. Most of the ISPs with significant long haul networks were pretty selective as to who they would peer with but they were also regulated telecoms. So most of the larger networks had policies that on their surface were neutral but were designed to screen out smaller players. For example a large ISP may have a policy that they would only peer with providers who could meet them at widely dispersed geographical areas and exchange a minimum amount of traffic.

Since then long haul bandwidth has gotten significantly cheaper and most large networks now have significant capability to sling traffic around the globe. For example Google (which isn't a traditional ISP) has a very significant global long haul network. This has made last-mile providers more sought after as peering partners because it is now relatively easier to move a few hundred gigabytes per second of traffic across the country than it is to get it the last mile.

Comcast was the first last mile provider to recognize this and move peering from the realm of network engineers to the MBAs and started systematically refusing to upgrade existing private interconnects and in some cases systematically de-peering in other cases. Comcast neatly side-stepped the entire net-neutrality debate by degrading service to everybody who wasn't willing to pay for a private interconnect. Comcast has had a relatively free hand because their customers are blissfully unaware of the politics of global peering and instead will just go somewhere else when a website is 'slow'.

This has put a lot of pressure on companies like Amazon who know that a 100ms delay in the order process can result in a 1% decrease in sales. Since private interconnect arrangements aren't public my guess is a lot of companies have caved and are paying Comcast to peer.

Google seems to be calling their bluff and launched a Google Fiber (a last mile ISP). In my opinion that was predictable because Google is very protective of their core search product and are willing to enter any market that they feel may pose a long term threat. The Chrome browser, Chrome OS, Android, Google Plus and Google Fiber make no sense other than to protect their core search market from Microsoft, Apple, FaceBook and Comcast (respectively).

Netflix had more at stake given the volumes of traffic they represent. They attempted to put pressure on Comcast by limiting HD to ISP's who agreed to host caching appliance in their network or upgrade their interconnects but most of Comcast's customer base is captive. Comcast customers can 'want' anything but with no choice, what are they going to do?

Dasan 888 said...

I think you're right. There is an analog with the initial Starz deal. It was cheap, way too cheap. Once the other content providers realized how valuable their sht was, that was the end of the cheap deals. No way NFLX gets free delivery of their content, using 35-40% of nightly bandwidth and gets it for free or cheap for very long.

Anonymous said...

A good explanation here:

Anonymous said...

This linked article might get all the technical details right, but that does not mean that the media are all wrong. Malone repeatedly stated that Netflix has profited from net neutrality like no other company and that now it's on the cable providers to charge heavy bandwidth users like Netflix a premium. I don't think that he doesn't know what he is talking about.

Anonymous said...

No way NFLX gets free delivery of their content, using 35-40% of nightly bandwidth and gets it for free or cheap for very long.

NFLX is only delivering packets that Comcast's customer's requested. NFLX has paid Cogent to haul their packets to Comcast.

Comcast has been paid by their customer to deliver the packets they have requested and is essentially demanding of NFLX that they be paid twice.

Comcast's argument is that they currently aren't making enough from their customer base to upgrade their networks to handle video traffic and the burden should be shared.

Personally I think that argument was blown out of the water by Google's broadband roll-out in select markets. Google's service offering is not only cheaper but faster. When Google entered the Provo, UT market Comcast quickly responded with bundles with higher capacity and lower prices. That wouldn't be possible if their networks were saturated.

Kid said...

Fun post, but this seems like a stretch "Comcast had more reason to dislike Netflix than Verizon though. After all Netflix competes with Comcast's core service (video delivery) and Reed Hastings of Netflix used to complain endlessly about Comcast's breach of net neutrality rules." Verizon wireline business relies on fios videos at least in part. Verizon wireless business will soon reach for ARPU with mobile video streaming.

I think what happened is that Netflix wanted better access to Comcast infrastructure to provide better service (install stuff in central offices etc) so paid a fee and maybe you're right Comcast had unusual circumstances and did a deal.

But I don't see how this is a potential new revenue line item for VZ, T and others who have the network and eyeballs; netflix paying has subs without it and doesn't have capital or interest in needle moving deal (for VZ/T), VZ/T doesn't want anyone touching their networks without paying up, particularly given their plans for the future.

Anonymous said...

Your hypothesis is not at all disproven by the post linked on your update John. That post doesn't address the question of why NFLX would choose to do a Comcast deal now v.s. continuing to pay e.g. Cogent.

Kid said...

Please tell me there's more to this VZ thesis than increased IP interconnect pricing to companies such as Netflix (which had formerly used intermediaries)...

Martin Barry said...

There is also a good take on it here

The reality is that Netflix pays somebody to take their traffic. Handing it directly to Comcast gives them a contractual relationship to enforce performance/quality standards. Doing it indirectly via Cogent or another 3rd party gives them no such leverage.

The only point of concern is the pricing power Comcast exerts on both the customer side, by being a monopoly provider of decent broadband in a lot of areas, and on the transit/peering side, by being the only way to reach it's quite substantial customer base.

As long as the price premium Comcast is charging Netflix is balanced by the benefits they receive there is no problem here. Unfortunately it's impossible for any of us to know for sure because the details of the deal are secret. We will, however, be able to see how much Comcast's performance improves in Netflix's reporting.

Anonymous said...

The post at StreamingMedia that you now link to has a informed take on the opinions, but is somewhat deceptive about what net neutrality is and why the current arrangement might be considered problematic.

The heart of the problem is the conflict between two sets of internet providers (ISPs): the ones serving content providers and the ones serving end users. There is some overlap, but basically, companies like Verizon and Comcast want to charge companies like Level3 and Cogent for the rights to reach end users. Comcast's argument is that they need to charge money to build out infrastructure--the counterargument is that Comcast is already making plenty of cash, and is paid by users precisely so that those user can access content. However, the ethics are less important than the realities of the market: end users are less likely to switch ISPs in response to issues. If Netflix's users are getting slow connections, it's going to be worse for Netflix than for Comcast, even if Comcast hasn't upgraded it's infrastructure in years.

However, it's rare for ISPs to actually disconnect each other. Instead, the concern of advocates of "net neutrality" is that providers like Comcast will switch up their algorithms to give preferential treatment to companies they like (and that pay them). Maybe Netflix comes in crystal clear but Hulu gets a laggy connection. This is what Comcast was ostensibly doing in this case, and may have been the reason why Netflix decided to pay them. For Netflix, securing a better experience for users is worth a few million.

The problem is that once this norm is established, you end up with a situation in which content providers are bidding for better quality from last-mile providers. You end up with perverse incentives--the worse a network is, the more important it is to secure reliable throughput. In other words, you end up with a different structure for the telecom industry, one with worse service and higher profits for last-mile providers. Which is why I'm sympathetic to the net neutrality folks, even if sometimes they don't know what they're talking about.

General disclaimer

The content contained in this blog represents the opinions of Mr. Hempton. You should assume Mr. Hempton and his affiliates have positions in the securities discussed in this blog, and such beneficial ownership can create a conflict of interest regarding the objectivity of this blog. Statements in the blog are not guarantees of future performance and are subject to certain risks, uncertainties and other factors. Certain information in this blog concerning economic trends and performance is based on or derived from information provided by third-party sources. Mr. Hempton does not guarantee the accuracy of such information and has not independently verified the accuracy or completeness of such information or the assumptions on which such information is based. Such information may change after it is posted and Mr. Hempton is not obligated to, and may not, update it. The commentary in this blog in no way constitutes a solicitation of business, an offer of a security or a solicitation to purchase a security, or investment advice. In fact, it should not be relied upon in making investment decisions, ever. It is intended solely for the entertainment of the reader, and the author. In particular this blog is not directed for investment purposes at US Persons.