On Friday, Dan Rayburn offered an update on CDN pricing trends for video delivery in the fourth quarter – as always, it’s worth a read. Amongst the data is that CDN pricing in Q4 was basically consistent with Q3. Could it be that the economic crisis has injected pricing rationality into the market? Year over year pricing per GB delivered fell 30-35%. That is still a pretty hefty pricing decline that roughly (and IMHO properly) matches what the IP transit market has done this year. But it is of course only half the story since traffic growth over CDNs has been strong. How much traffic growth does it take to balance a 30-35% pricing decline? The number is 50%. Was CDN traffic growth above 50% in 2008? Yes I think so, and therefore the pie is still growing.
I have in the past made the case that CDN pricing pressure is inevitable right now, there are just too many providers. Often I have used the IP transit price wars of 2001-2005 as a parallel. But today let’s mention a reason that the two cases are not all that similar. Back in the bubble, too many carriers pre-built far too much capacity. The money was spent already, so when the pricing wars got fierce there was no spending to cut back on. There was only unsold capacity and a need to get something for it, and anything was better than nothing. The business case for building the inventory in the first place didn’t matter for a long, long time.
The CDN world has too many competitors, but it hasn’t built that much actual capacity. There are no piles of inventory to work through. Sure there are too many participants, but the CDN business doesn’t have the same delay between capex and sale of inventory that the bandwidth business did back in the bubble. Many tens of billions went into the inventory of the last bandwidth bubble, the amount of venture capital spent building too many CDN businesses is microscopic in comparison. For small, new CDNs, the cost of a prolonged price war is measured in cash yet to be spent, which means that the balance is much, much different this time around. The economic crisis has therefore reduced the risk of a pricing demolition derby in the CDN business by taking away the cheap money that might have fueled it.
So what will happen? It’s all about scale. Some will get it, and the rest won’t. Those that have it can handle the price declines by scaling revenues against fixed costs. Those that don’t have it will be squeezed, and while they may be forced to try desperate measures to get it, they just don’t have the inventory to do it for long. So while I continue to see pricing pressure in the CDN segment as a major feature of 2009, the bigger CDNs as a group will only get stronger because of it. The smaller ones will just disappear, either via acquisition or not.