The party may just be over for CDNs, as more and more companies raise more and more money: Panther Express, Conviva, BitGravity, CD Networks, EdgeCast, Grid Networks, Highwinds, Velocix, Itiva, Pando, Rawflow, the list goes on...
The trouble is there are only a few factors that can differentiate you in the CDN world. These are:
Network Size – how big is your network (and there are any number of ways of measuring this, from nodes, to peering nodes, to number of servers)
Network Reach – how far does it reliably reach (serving 500,000 viewers in the Netherlands is a very different proposition to servicing 500,000 viewers globally due to the pressure on local infrastructure)
Technology used – fundamentally, it’s all pretty similar, using variations on two themes – replication or caching; however, services like multicasting can bring some cost benefits into the equation
Service – such an imponderable and something companies like Akamai have a lifetime's worth of experience with
Added Value Services – what above and beyond do you offer, especially when clients are looking for a one stop solution
Price – which is what it often comes down to at the end of the day, especially for the smaller players and the smaller clients
It’s estimated that the video CDN market is worth $400m and three companies have the vast majority of this – Akamai, Limelight and Internap. With AT&T and Level3 investing heavily into this market, what’s left for the new and smaller players ? Not an awful lot it would seem. The major trade buyers will find it much easier and cheaper to build than buy; associated companies like broadcasters and studios are unlikely to want to buy at a premium as they lack any economy of scale and the markets have been overly harsh on all companies involved in the Internet TV business.
On top of all this is the introduction of Edgeware's new Reflex2 box, which can serve tens of thousands of streams, significantly reducing the cost of self-build for smaller users of CDNs (although it clearly does not help sort out peering bottlenecks).
All of this said, however, I do believe that there remain some good niches:
CDNs with heavyweight tools and added value services – ironically being this didn’t do much good for Narrowstep, but they sold themselves much more as a technology provider not CDN.
Multicast based services – companies like GlobalMix have cut a profitable niche for themselves by beginning to establish multicast delivery, and this is likely to make them an attractive acquisition proposition in the future.
Market local CDNs – who can plug gaps in local delivery that the larger player might struggle with.
The reality is that the CDN market is beginning to mature into different sub-markets such as live v. on demand, streamed v. download and, most important of all, small customers v. big customers. The only strategy left for the companies without any real market advantage is to target the smaller players who can't get time of day from the big networks, but who are likely to achieve rapid growth.