Dan Rayburn has a thought-provoking article out today in which he describes the video CDN business as flawed, with its downside obscured by subsidization by Google/YouTube. He has a point, but I think it's actually two separate points that deserve their own space.
The first point is that internet video has bitten off something that has so far been too big to chew. Everything on-demand all the time is really cool but also very hard on network infrastructure, whereas the broadcast/cable model of having a fixed number of channels with things on at one particular time only is much easier. Just because OTT video doesn't have to build the last mile doesn't mean the economics don't matter. It's no accident that OTT video quality doesn't come close to traditional formats, if the costs can't work their way up the chain then neither will the quality because the infrastructure (especially in the last mile) won't do OTT's bidding any other way.
The second point is one Dan alludes to when he says "Level 3 is the one CDN who actually owns their network and has a lower cost of delivering video, so for them, the Netflix business makes sense." To say it bluntly, the fact that Akamai and Limelight don't own (or aren't owned by) network has been a choice. They could easily have taken that route over the past several years, but they didn't. The fact that they are now finding video delivery to suck as a business model is a direct consequence of the way they avoided getting their hands dirty with the underlying fiber infrastructure.
Yet the everything on-demand model isn't going to go away whether Google and YouTube continue to lead it this way or not. That particular toothpaste has no chance of making it back in the tube, and the details will get solved eventually - with someone making a whole lot of money. In the meantime, the bits *will* get delivered, whether traditional CDNs take part in the process or not. That's one thing we learned from the dot-com bust - the internet backbone model's brokenness didn't change the trajectory of traffic growth even when everyone fled the wholesale model. Traffic growth waits for no business model.
So does all this mean that Level 3 is about to really cash in on its big CDN bet at long last? (They could certainly use a tailwind for once!) Because nobody else seems ready to step into the breech here if the pure CDNs taste for delivering video bits continues to decline even as traffic inevitably grows. Except perhaps Tata in the international arena- though we rarely hear much from BitGravity these days either. Perhaps some fiber operator will finally snap up Limelight this year and step up to the plate with some immediate scale. Teliasonera or NTT perhaps? Hmmmm...