Yesterday, telecommunications giant AT&T (NYSE:T, news, filings) announced it is re-entering the CDN marketplace, an action which they have been telegraphing for several quarters. They last did this back in 2008 just before every telecom on the planet started salivating over over entering the CDN space, planning to spend up to $70M to roll their own system that would compete against Akamai, Limelight, and the rest of the field. Well, we all know that never materialized in any real way, much as the overall telecom threat to the CDN space fizzled out except for Level 3. This time around AT&T is no longer doing this in-house, but instead is using licensed technology from EdgeCast and Cotendo to assemble its new CDN platform. But do they really mean it this time?
Dan Rayburn says things look quite different now, in that the company has a much healthier outlook on the industry and on matching its own strengths in infrastructure and customer relationships with proven industry technology platforms. I tend to agree. After several years of making little if any dent in the sector, the company appears to have learned a few things. But I still doubt we are talking about any material changes in the competitive environment in the near term.
For one thing, it's the numbers. Suppose AT&T does seek to compete directly with Akamai, Limelight, Level 3, and the rest of the field, and does really well - quickly reaching a revenue of $100M/year in revenue. That would be important to the CDN segment, but it wouldn't even be a rounding error next to AT&T's overall business. If that was their goal, then frankly why bother?
The other thing is culture. Yes AT&T has those relationships with content providers, but this is a very dynamic sector that requires a very different mindset. Look at how hard Level 3 has had to work to adapt its culture to get as far as it has, and then remember that even an eager, rejuvenated AT&T CDN business is still part of an ILEC with baggage that will always slow it down some. It's hard to imagine AT&T moving nimbly enough to dictate the game.
A more active AT&T CDN effort will IMHO be more focused on controlling the costs of their own network by optimizing caching and the like. Rather than looking at this as a way to take on Akamai, I still see it as AT&T's way to quietly build itself the fast lane that the network neutrality guys hate so much. And when it's done, I think they hope not only to sell access to it directly, but via the CDN federations which will eventually develop. By being a power in the industry when that happens will allow them to affect the economics of those federations to their advantage.