Last week seemed to be the week of many CDN rumors, one of which was Microsoft being interested in buying Limelight Networks. While I have no idea if that is the case, the blog posts and articles that followed the rumor were a great example showcasing how many people really don’t understand the product offering that the different CDN providers have in the market today. Too many are still comparing one CDN provider to another unfairly. The term "content delivery" is used way too generically these days and folks tend to speak to it as if it covers every type of content under the sun, even when it doesn’t.
Update: I am getting a lot of comments from those who are saying that Limelight does lots of other kinds of delivery and my choice of the word "only" is incorrect. While I know they do some, the majority of what they deliver is video and rich media content. But I did contradic myself when I said "only" video but then also said "mostly" video in another post. To be exact, I should have said "mostly" video in this post as well.
Numerous blogs all talked about how it would make sense for Microsoft to buy Limelight Networks so that Microsoft could accelerate it’s own CDN build-out. SiliconAlleyInsider.com speculated that, "perhaps it (Microsoft) believes Limelight’s infrastructure and expertise will help accelerate its transition to cloud computing. Specifically, instead of buying CDN services from Akamai, et al, Microsoft could now float MSN, Office Live, Silverlight, and other Software-As-A-Service products on top of the Limelight infrastructure." How is that? Limelight Networks does not offer application delivery, software as
a service, static caching and most other forms of content delivery over their network today. Limelight specializes in running a network that is optimized specifically for the delivery of video content
only. So Microsoft acquiring Limelight does nothing in the way of advancing their "cloud computing".
On the same day of this rumor, DataCenterKnowledge.com reported on some of the details surrounding the build-out by Microsoft of it’s own "edge network". This gave some good insight into what Microsoft is working on however, the Microsoft person quoted in the post never used the word video. They said "edge network". Most video delivered on the Internet today is not delivered from the edge and most video is not cached like many seem to think. So what kind of content exactly is this Microsoft network going to deliver? I’d love to see a follow up story by DataCenterKnowledge.com talking to that.
Around the same time as this rumor, GigaOM.com ran a short post about Limelight’s share price and revenue guidance and said, "Limelight
spends about 60 cents on every $1 it earns just to provide service,
whereas Akamai spends about 30 cents, I’m not sure how low Limelight
can go. Or for how long." Yes, those numbers are correct, but not
in the context they reported them. No one knows how much it costs
Akamai to operate their network specific for video since they delivery
many kinds of content. Again, Limelight delivers only video, so
comparing Limelight’s costs to deliver one kind of content, to Akamai’s
costs to deliver many kinds of content, is not a fair comparison. Using
those numbers in that context is just plain wrong and set expectations wrong in the market.
I’ve said it many times before on my blog, you can only compare the CDN service from one company to another and not the companies themselves. The CDNs are out in the market making very clear statements about what they do and do not offer, yet it seems like no one wants to listen. At Akamai’s recent analyst day, what did they focus most of their time on? Not CDN, but application delivery. They are trying to tell the market they do more than just content delivery of static images and video content. Limelight on the other hand is trying to educate the market about how they
only delivery video and are the specialists when it comes to that kind of content. Level 3 is focusing on offering a bundled service of transit, co-lo, static CDN delivery, streaming etc…. same with Internap who is focusing on multiple products. In most cases, the CDNs are delivering a clear message to the market on what they do and do not offer in the way of a product portfolio yet it seems very few are paying attention to that and have a clear grasp of the products that make up the numbers.