Breaking the net-neutrality deadlock is about quids, not QoS

By Rob Gallagher

As debate continues over whether the latest micro-amendment to the European Union’s telecoms regulation will defend or destroy the supposed neutrality of the internet, one major question remains unanswered: whether it was ever in danger in the first place.

In keeping with the tone of some of this net-neutrality debate, here’s an overly simplified recap. Operators want to charge internet firms such as Google for using their broadband networks. The internet firms don’t want to pay. The operators say they won’t be able to afford to support the traffic generated by internet services unless their providers pay. The internet firms say this would transform the internet from a “neutral” space, where even the smallest of startups can succeed, to one dominated by operators and their giant-corporation clients.

The operators cannot simply demand that the internet content providers pay for the “free ride” they have enjoyed to date. Rather, they’re trying to bring something new to the table, something only they could provide. “Someone will ask for something that the Web doesn’t do at the moment, and of course there are ways of fixing it,” an executive with BT once told me. “But they don’t come for free.”

That executive was Paul Reynolds, who was then CEO of BT Wholesale, and the year was 2006. I’m not aware of any instances in which a major content provider has seriously approached an operator with such a request since then. But BT and a number of its counterparts have at least come up with a number of products they think content providers might pay for. Typically, these involve operators helping deliver content in ways that are somehow superior to what can be achieved over the open internet.

There are good reasons why some of these products might actually appeal. Most internet firms have long paid to use the content-delivery networks (CDNs) of companies such as Aramaic and Limelight, which employ servers in major cities and towns worldwide. Operator CDNs promise to do the job even better, by using even more servers even closer to broadband users’ homes.

Online video, traffic-shaping and QoS conspiracies

But the reasons some operators give for why their services will succeed seem to miss the point somewhat. They argue that content providers will, at some indeterminate point in time, have to strike deals with them. As broadband traffic continues to rise while subscription fees remain flat or fall, ISPs will slow down, or “shape,” heavy traffic to keep down costs, the argument runs. Internet firms will in turn have to pay operators to guarantee that their services reach the consumer in an acceptable state.

That is unlikely to happen. Although most ISPs already shape traffic, no one would dare do it so heavy-handedly that a popular internet service such as the BBC’s iPlayer would be unusable. It’s a PR battle the (largely unpopular) operators can’t win. Disgruntled subscribers would simply switch to a rival, probably one whose CEO had promptly stepped up with some bluff about how his company would never control what its customers can see and browse. A more likely outcome is that those ISPs that can’t keep costs down in other ways will simply sell up to those who can.

Some operators also argue that content providers will be willing to pay a premium for new special features. These could include enhanced quality of service (QoS), which would enable content providers to make their services download faster or run in a higher resolution than those of their rivals. A variant of this argument suggests that online video providers in particular will have to pay for such features if they want to make the leap from the PC to the TV.

But there are several reasons why this model is unlikely to take hold either. Ever-increasing broadband speeds are giving content providers more and more room to enrich their services. Many are also employing technologies that can sense what sort of connection a consumer is using and dial the quality of their services up or down to fit.

CDNs are all about the cache

But the main reason involves cost. Content providers like the open internet because it costs them very little to use compared with traditional media, such as TV, radio and print. In fact, the main reason they use CDNs is to save even more money.

Akamai and other traditional CDNs cut down content providers’ bandwidth costs by storing, or “cacheing,” copies of their content on their worldwide network of servers. When an internet user tries to access the content, Akamai delivers it using the server closest to him, rather than from the content providers’ servers.

Operators have the potential to cut bandwidth costs even more by placing their CDN servers even closer to the home. But such “local CDNs” face one major problem: Their servers might not cover enough internet users.

Content providers like working with providers such as Akamai because they have to sign and manage only one deal to get access to a global service. Operators can get closer to some users, but not all. The world’s broadband subscribers are spread among thousands of ISPs across hundreds of different countries. To take advantage of the local CDN concept on a widespread basis, a content provider would have to strike deals with multiple operators.

Operators with high national broadband-market shares have a temporary opportunity in that many popular online video services are limited to specific countries for licensing reasons. The BBC iPlayer, for example, is available only in the UK, while Hulu is restricted to the US.

Such content providers might be the first to weigh up whether the benefits promised by operator CDNs would be worth the hassle of connecting to them. Let’s not forget that any deal would require costly and time-consuming commercial negotiations and systems integration. But if an operator does succeed, it will have achieved something truly remarkable: convincing a content provider that it can do something the open internet can’t do – save them even more money.