The market for content delivery is expanding (although as we will see, there’s no consensus on the profitability of that market), and the ability to place large content files close to the people who need them will be an essential service for carriers in the future. But should they build that asset themselves? Should they find a specialist partner who can help them to do the job? And will today’s CDNs last long enough to make good outsourcing partners?
To get some answers Capacity brought together four of the most important voices in the industry. At Edgecast, president James Segil has positioned his company as the “carrier’s choice”, keen for them to outsource as much of their business to Edgecast as possible. Akamai is by far the largest player in the CDN business, and Kieran Taylor is senior director, global marketing. In the future BT has committed to building its own CDN infrastructure. Simon Orme, general manager, content services at BT Wholesale, explains why it’s a core business, not one to be outsourced by a global carrier. And Frost & Sullivan principal analyst Dan Rayburn has strong opinions on the viability of the CDN business, often to be found at his blog businessofvideo.com.
“Content” has become a catch-all term, and so “content delivery” also ends up meaning different things to different people. In the next two years, what type of content will need specialist skills to deliver? Will the volume growth be matched by revenue growth?
James Segil: Players want to work with different partners for different needs. So maybe a BMW will want to look at the complex tracking codes so they know who is looking at their car ads, maybe they want to work with a best of breed CDN; but Microsoft wants to just download and cache as a service pack, they could work with a more commoditised CDN that a carrier could provide.
Simon Orme : By 2012 around 70% of internet traffic is expected to be video related, according to Cisco. We are seeing a meteoric growth of video services like Youtube, iTunes , iPlayer with a 30-minute TV programme requiring the same bandwidth as 78,000 emails. With BBC iPlayer already consuming 12Gb of data every second, a further surge in content demand can only be expected with the take-up of fibre-optic and 4G mobile services. It’s the sheer volume of content that BT believes will need specialist skills to deliver it.
Kieran Taylor: We immediately default to cacheable content, which is 60% of our revenue. But a growing proportion of our revenue is helping to get acceptable application performance.
Dan Rayburn: But I don’t think there’s enough profitable revenue. The problem all CDNs have now is becoming profitable. Network consumes bandwidth, there’s a lot of infrastructure costs. It’s great we have mobile, Tivo, Xbox. We can talk about growth all day long – but that gets us only so far. A lot of people say HDTV will change the landscape. It won’t, content in HD is a better viewer experience, but it is four or five times the number of bits, and there’s a cost to that.
Kieran Taylor: We continue to get good margins. But it’s our goal to diversify.
As carriers outsource low-margin business such as wholesale voice, surely delivering rich content is absolutely what they should be doing for themselves? If they outsource, is it just because they don’t have the capital right now?
Simon Orme: Content providers need direct access to a large end-user base so their content is delivered to an extensive end-user base efficiently at a suitable level of quality. To satisfy this requirement an open platform – accessible to all communications providers and LLU players – would be the most efficient way to achieve a model that supports these emerging business models. This is what BT is planning.
Dan Rayburn: Some carriers who really know what they are doing, like Level 3, are looking at the entire ecosystem. But if you’re just a telco with a network, CDN isn’t where you should be because you need the right mindset. That’s why Global Crossing and Deutsche Telekom and others have gone out and partnered. The CDN market is still so small compared to their other business. And if they can get an insight, test the product, they will simply acquire another player.
Kieran Taylor: But we have seen pretty rapid take-up from service providers who are avoiding the cost of network build-out. We also have something called our Accelerated Network Program, where they house our servers in their network, so they don’t have to backhaul for content.
Will there always be an independent tier of specialist CDNs, or will they inevitably be consolidated into global carriers – or will carriers simply partner while they build their own capability?
Kieran Taylor: I don’t see many of them building their own. If you look at our architecture, we have more than 900 network providers. You really have to have a broad distribution across a number of networks to meet a provider’s needs.
Simon Orme: What we can be sure of is that the number of CDNs in the marketplace is likely to decrease as the market needs economies of scale to work efficiently. Some smaller CDNs will not have enough revenue, customers, patents, applications or intellectual property to go it alone. Also content providers will not be looking for deals with multiple CDNs because this is inherently inefficient from their perspective.
Kieran Taylor: But we don’t see one telco having more than 8% of the access traffic, and there is a long tail. For a CDN to be viable, it has to have distribution across a large number of networks, which is only possible for independent providers.
James Segil: Today I think we are the carrier’s CDN of choice. How’s my last quarter? The best quarter ever. Demand is continuing to skyrocket: it is almost doubling every six months.
Dan Rayburn: There will always be independent CDNs. There will be plenty of business for a $2,000 customer. There will be regional service providers, offering CDN services starting at $99 a month.
On that subject, are CDNs doing a good job at marketing themselves to carriers as white label or outsourcing service providers, and are the type of people that they need to work with inside carriers open to partnership?
Dan Rayburn: I had a group of carriers on a panel at the CDN Summit 2009 and they said they were working with Limelight and Edgecast. There was comment from Deutsche Telekom that they preferred to partner with a young, agile company rather than Akamai, because they need to be quick to react.
Kieran Taylor: That’s not the case in our experience. We partner with a number of global carriers. BT resells us, Telefonica, NTT and Verizon.
James Segil: Having a strong CDN-known brand has been the most effective way to win white-label business. We have 40 strategic partners that resell our services, folks that are knowledgeable. The reason they chose us is because they know the customers know Edgecast, and they want our technology. On the other hand there are companies who have tried to do a just wholesale approach, and they have not succeeded: carriers want to be with a branded leader and not a white-label no-name because of the strategic importance of this business.
Simon Orme: The market is extremely competitive so it’s hard for companies to get their messages heard when there are so many players all pitching similar services. This is especially true in the CDN market, where so many vague terms are used to describe services.
Dan Rayburn: Edgecast has set up its company to support resellers from day one. Most of the CDNs created a reseller programme a few years later. As a result they’ve picked up some good traction.
James Segil: We’re nimble, entrepreneurial, when they look at our technology we have given them solutions on how to work with their billing systems, which really integrates with their IP services.
From the current roster of CDN providers, who has the best ability to become an outsourcing partner for carriers in the long term?
Simon Orme: The big drawback of many of today’s content distributors is that they are dominated by managing static content rather than live streaming for high-quality media and TV. In all cases CDNs reach the edge of an ISP network and the final stage is delivered over the top with no assurance guarantees to an end-user device. Those CDN providers who will thrive are those who have realised they need to try and sell more than just video delivery and diversify their revenue by making their platforms more agnostic and flexible so they can deliver video, games, software, small objects, applications and any other kind of content the market demands.
James Segil: Edgecast is taller and better looking than other CDNs. Akamai, Limelight and Edgecast are the top tier because they are full service, with video streaming, live and on demand, and have a robust cacheing program. And we have a comprehensive reporting stack and PoPs all over the world. A Bitgravity or Highwind or an Interoute either offer one service or in a limited geography.
Dan Rayburn: To some degree these guys are putting on a public face, pretending this is great when it isn’t. Some are fooling themselves. At one time there were 50 CDNs in the market who raised half a billion dollars. They raised more money than the size of the market. Investors make bad decisions every day: just because you have VC money it doesn’t give you credibility.
James Segil: And not everyone can handle every bit of business that’s out there. Akamai has 60% or 70% share, it’s by far the biggest. They are the only one that can handle iTunes or companies that are pushing 400Gbps. We are not going after iTunes-type business because it’s so commoditised. But you don’t take those deals to make money, you do it to lower your cost basis.
Kieran Taylor: We think there will be a consolidation. As a result of the downturn, there has been a flight to quality from carriers who want to partner, and so we are already seeing that consolidation happening today.
