If Akamai returns fire…

May 26th, 2008 by · 5 Comments

Yesterday I discussed the benefits of owning the network in the CDN business.  I do realize that the issue is a disputed one, but for the sake of argument just suppose that owning the network does change the cost structure of the CDN business substantially.  What happens next?  Surely the pure CDN providers won’t just sit there, waiting for annihilation, so let’s look at some possible responses:

  • Refocus on more managed offerings, premium services, etc.  Things a carrier would have no advantage in.
  • Ally with a network provider that has no CDN
  • Build a network
  • Buy an existing network

Smaller pure CDNs obviously would find the latter two options to be beyond their means, which means they would choose some combination of the first two.

Focusing on the high end, differentiating yourself by service capability and avoiding the commoditized segment of your market – that is something we are already seeing, of course, and not necessarily because of the entrance of network operators.  It is a natural response in a crowded marketplace.  It is also the fast track to niche-ville.  Being a niche player is fine, but it would be a bit of a reduction in scope for much of the capital these guys raised to take on the world with.  Wouldn’t be the first time though.  One CDN player, Internap, is already aiming at this – which isn’t a surprise because Internap has always been a niche player, serving the high end managed IP transit market.  They like it there and know the terrain, and thus probably succeed.

Allying with a network provider that has no apparent CDN aspirations, e.g. Verizon, Qwest, or Sprint on the large side, or a Savvis, Global Crossing, or XO on the small side.  I suppose it might work sometimes, but alliances generally have difficulty reproducing what a unified company can because the two parts don’t really have the same interests.  Also, there is always the risk that the carrier you ally with may decide to do it itself one day.  Some might try this route, generally in an attempt to avoid getting stuck in a niche and perhaps to eventually get bought by the network they ally with.  Personally, I think Limelight is most likely to follow this route, but others might.

Building a network?  What exactly would that mean?  Obviously all these companies have some sort of network hooking up their nodes, but in this context it means building a network with an excess of capacity and perhaps selling that capacity in addition to CDN services.  It means lighting longhaul fiber between markets with DWDM gear and using metro fiber to bring multiple datacenters on-net.  The thing is, it is very hard to achieve the necessary scale doing this, you don’t see the cost savings you were originally hunting for until you are huge, either a Tier-1 backbone or close.  It is just to much trouble, a giant cash sinkhole with the rewards beyond the range of today’s investment horizons.  Nobody short of Google or Microsoft is going to even think about it.

That leaves buying a network.  Of all pure CDNs, Akamai is the only one with the means to pull this off.  But they really could – their stock is good currency, they could buy an XO or a Savvis or even a Cogent without breaking a sweat, then sell off the pieces that make the least sense (e.g. in the case of XO, the legacy voice/clec biz).  Would Akamai do this?  IF IF IF owning the network matters enough, then it is a no-brainer.  If it makes financial sense for a network to own a CDN, then logically it makes equal financial sense for a CDN to own a network.  And Akamai is not in this to become a niche player, nor would they fit on the shoulder of another carrier like a talking parrot.  Akamai dominates the CDN sector, they own everything and they aren’t going to walk away from that, nor should they.  If Level 3 and AT&T are right to attack from this angle, then Akamai would be equally right to defend in the same.

If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!

Categories: Content Distribution · Internet Backbones · Mergers and Acquisitions

Join the Discussion!

5 Comments So Far


  • Dan Caruso says:

    what kind of network does savvis have? I lost track of what they retained and what they sold over the years.

  • Rob Powell says:

    Savvis still owns at least one if not two fiber pairs on the Level 3 network, along with all that datacenter real estate. They are only using it to a limited extent, but I believe they are still tier-1 in the definitional sense of not having to buy transit.

  • Dan Caruso says:

    having a couple pairs of fiber doesn’t strike me as a meaningful “solution” to Akamai’s potential disadvantage….i will comment more on this in a post later this week that links to your posts

  • tech101 says:

    CDN without networks may lose out like those facility-less CLECs, and the lack of network support at least limits AKAM’s future growth as Goldman Sachs recently commented about AKAM:

    “Beyond valuation, we are also concerned about intensifying competition as private entrants proliferate and selected large network operators begin to eye the space … While portions of Akamai’s customer base are likely relatively immune from competitive pressures, we believe that the majority of Akamai’s 2,700 customers will be able to put increasing pressure on the company as contracts come up for renewal this year.”

    The only carrier that fits AKAM’s scale and that is left for pick up is probably XO. However, could any one imagine Carl Icahn would let XO go without a minimal $2 b, considering XO’s current $109 m ebitda plus a minimum $100-300 million recurring saving in the synergy from both companies?

    This value even doesn’t include XO’s wireless business NextLink and the $3.3 billion NOLs, which alone may save the profitable acquirer at least $1.1 b.

    If AKAM announces a merge/acquisition with XO tomorrow, AKAM may jump 10-20% like when LVLT announced the Broadwing buy-out.

  • Paul says:

    “Building a network? What exactly would that mean? Obviously all these companies have some sort of network hooking up their nodes, but in this context it means building a network with an excess of capacity and perhaps selling that capacity in addition to CDN services.”

    LLNW does this. They own their own network, and sell both transit and CDN services. Visit http://www.limelightnetworks.com/network.htm .

Leave a Comment

You may Log In to post a comment, or fill in the form to post anonymously.





  • Ramblings’ Jobs

    Post a Job - Just $99/30days
  • Event Calendar