Which Network is 'Best' and CDN Implications

July 21st, 2008 by · 10 Comments

As a followup to my thoughts on that WebHostingTalk poll, consider the top 4 in that poll:  Level 3, Internap, AT&T, and Savvis.   These companies have something in common, the first three are recent entrants into the CDN space, and the fourth had a CDN until last year.  In fact, the first three are the only networks with a CDN offering. 

This is not a coincidence, it is a customer focus thing.  Many of the customers that buy CDN services are essentially the same ones that buy IP Transit in datacenters.  In other posts I have tried to make the point that other network operators seem less likely to enter the CDN business than those who already have.  For instance, on the surface Verizon seems very likely to enter the CDN space since AT&T did – what’s good for one RBOC is good for another, right?  In my opinion, this does not necessarily follow – it could, but AT&T has a substantial presence in the datacenter market that Verizon doesn’t have.  It would take Verizon more work, more money, more focus to enter the CDN business than for AT&T, or for Level 3 or Internap, because Verizon starts from a lower rung on the ladder amongst these customers.  That doesn’t mean they won’t do it, only that the hurdles are higher.

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 · Datacenter · Internet Backbones · Old · Polls

Join the Discussion!

10 Comments So Far


  • Dan Rayburn says:

    Verizon will enter the CDN market this year with a delivery offering, but it won’t be by themselves and will be in conjunction with another company. I will post all the details about it when I am allowed.

    I would have to disagree with the statement that most customers who buy CDN also buy IP transit services. Akamai and Limelight who own 70% of the U.S. market for CDN services for video, based on U.S. revenue, don’t sell IP transit services at all, yet they are still the ones dominating the market. And many content owners don’t have any need for transit at all. Yes, many of the big content owners and major studios and broadcasters do, but that’s not the majority of CDNs customers. The average content owner who does $3-4k a month in video delivery services over a CDN will never have the need for IP transit products and services.

  • Rob Powell says:

    I would look at it from the network providers point of view. A CDN would normally be run by the same part of an organization that runs the datacenters, which happens to serve IP transit. Therefore, when a network provider looks at entering the space, they naturally look for synergies and bootstraps from that asset base, from that skillset. If they don’t have it, or have less of it than others (Level 3, Internap, AT&T), then it is harder for them to enter the new business.

    Interesting about Verizon. If they are to enter the space, doing it via partner rather than directly like AT&T would make some sense, because they can tap the partner for the expertise they don’t start with. But partnerships often work better on paper than in practice, so it will be interesting to see who they choose.

  • Aaron says:

    Limelight doesn’t sell transit? Tee hee…

  • Brian Scully says:

    Dan,
    Are you referring to Velocix going to market with Verizon?

  • U Wanna Know says:

    The world is flat & the US is not the only place where revenue is generated. Glad to get vision that Dan is exclusively focused on the US – but the growth currently across the globe is still in Asia (where CDNetworks has 70 % of the marketshare – with ChinaCache also having a decent amount of revenue – exclusively in China). There are metrics to show that more groups are utilizing IPTV on the other side of the earth than the US and also broadband is more mature in other markets than the US. Time to look at the global perspective…. not just in your backyard.

  • U Wanna Know says:

    Before you also try to figure out which CDN is the best fit – you need to look at what the particular vendors provide.

    a) Streaming – WMS or Flash (which is what Dan is focused on – not all providers provide this today – e.g. PantherExpress)
    b) Object Caching (not all CDN’s can provide this)
    c) SSL caching – (only a couple have this globally – heck only a few offer this in the US)
    d) video/ file download (most can provide this)
    e) ESD (Electronic Software Download) – 7 do this well
    f) Application acceleration -unique to AKAM for a network based solution

    BTW – there are really only 2 top providers with all of the above (except for the Application acceleration – currently unique to only AKAM) One is the largest CDN in the US – one is the largest in Asia – who also has US based sales executives.

    You gotta know the whole market and respect what the whole global market brings to have full vision of the choices in hand…

  • Dan Rayburn says:

    “Dan is exclusively focused on the US”. No, but the U.S. market has the most providers, the most customers for these services and generates more revenue than Europe or Asia. It is also the region that has the most data available about the market.

    And if you look at my list of providers at http://www.cdnlist.com, it contains CDNs from Asia and Europe as well like Accelia, Aflexi, Broadmedia, ChinaCache, CDNetworks and PCCW.

    If CDNetworks has 70% of the CDN market share in Asia, then based on CDNetworks 2007 revenue, that means you are saying the entire market for CDN services in Asia was only about $100 million last year. That would be about 10% of the market share for all CDN services in the U.S. – so how do you justify that as being where the “growth” is coming from? You want to share your opinion great, but the moment you mention numbers and market share, you open yourself up to the hard facts and the data.

    You make two posts, both championing for CDNetworks, not disclosing your name. Clearly it is for a reason.

  • nottlv says:

    Dan,

    As Aaron mentioned, Limelight absolutely sells a transit product. And correct me if I’m wrong, but don’t most customers have origin servers and therefore need some transit?

  • Dan Rayburn says:

    Yes, Limelight sells transit, but have you ever asked them what percentage of their revenue it equals each year? It’s so small, they don’t even break it out and it continues to decline each year. Even if guessing that 5% of their revenue is from transit, that’s only about $7 million a year, which is not even a drop in the bucket compared to Level 3, Internap, AT&T or other telcos. Years ago, Limelight had a large portion of the revenue from transit, but I think too many people are thinking of the Limelight from 3 years ago. They are a very different company today.

    For content delivery specific to video, no, most customers don’t have origin servers. For content delivery of static content, images etc… yes, most do. I may have assumed incorrectly, but I thought this post was specifically talking to CDN as it pertained to video delivery.

  • Rob Powell says:

    I agree with Dan in that if/when pure CDN players sell transit it is purely on a resale basis and not material to their operations. Network operators looking to enter the sector would factor in savings due to their existing IP Transit resources as part of their calculations. It doesn’t work the other way unless a pure CDN goes out and buys a network, which only Akamai has the resources to do if they wanted to or felt they needed to. At the moment, they are not likely to because any pressure from network operators is purely theoretical.

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