Telcos Not Yet Willing To Run Their Own CDN: Bharti Airtel Enters Via Limelight

Airtel_Logo Over the past two years, more than a dozen telcos and carriers have entered the CDN space, all with the strategy of re-selling or partnering with a pure-play CDN provider like Akamai or Limelight. In that time, Level 3 has been the only company with the strategy of wanting to own and operate their own CDN, laying the groundwork today for what will be a big business in the future. This morning, Asia’s largest telecom provider Bharti Airtel announced they too would enter the CDN space to serve the Indian market via a partnership with Limelight Networks.

With this announcement, more than 20 telcos and carriers have entered the CDN market and I know of other carriers who plan to make similar partner announcements this year. While Wall Street and others want to continue to speculate that telcos and carriers will one day rule the CDN business, thereby forcing pure-play CDNs out of the market, that won’t be happening anytime soon. Yes, carriers who own their own fiber have a distinct advantage over pure-play CDNs who don’t operate their own network, but the delivery of bits is only one of the many pieces required for a real CDN offering. Vendors have to be able to do more than just own their own network and deploy a bunch of boxes. If it was that easy, all of these carriers would have done it already and would not be partnering with pure-play CDNs.

Right now, the CDN market for video is too small for most carriers to spend the money to build and deploy their own network. By my estimates, the entire market size for video delivery services globally was only $500M last year, which while a big number, is small in comparison to the other markets that telcos and carriers are in. Until the market size and opportunity is larger, I don’t expect we’ll see many of these companies who have partnered with pure-play CDNs doing anything on their own and bringing CDN services in-house to their product portfolio.

  • It depends on the telco, and it depends on how well informed the telcos are of the opportunities and options out there. If a telco wants to move in with a short term strategy to offer a commercial CDN service, reselling any CDN may do.
    But in mid-term and long-term, it doesn’t make sense to outsource to a third party network if you own a great network yourself. Besides, selling CDN services to the market isn’t the only strategy for telcos. Optimizing the network for content services (lowering backbone and peer loads) is another strategy. Offering on-net services (IPTV for instance) is another strategy. Most telcos are looking to combine these three strategies into a single CDN and global CDNs can’t really deal with the latter two strategies.
    Another issue is QoS for OTT services. All global CDNs, regardless their infrastructure, throw traffic over the fence into internet exchanges, carriers and telcos networks. No QoS. Telcos however can guarantee QoS from the delivery point all the way down to the end users local termination point. Serious OTT content providers need telco-powered CDNs.
    With todays technologies, a CDN can be setup and running in a few weeks. This is a huge difference compared to the 1/2/3 years time-to-market that telcos had/have to deal with when they have to develop an in-house CDN solution, or when they have to replace their entire network infrastructure if the solution is integrated into the network level.
    The reason why it takes telcos time to get on-net CDNs on the market is because they have to learn how to market, sell and support these services. It really depends. Some telcos have many years experience in this field, others are just starting.
    Besides telcos, hosting providers and broadcasters are also deploying CDNs. Even enterprises are.

  • Paul

    “The reason why it takes telcos time to get on-net CDNs on the market is because they have to learn how to market,”
    I completely disagree. The barrier to entry set by Limelight and Akamai is high. Its not just bit delivery, its reporting, analytics, storage, and more. Its just like if you wanted to start a package delivery company to compete with FedEx or UPS. You can’t buy two planes, park them in Cleveland, and say you are in the game. You need the global scale and the value added services…and then and only then can you start marketing against two well-established brands.
    “With todays technologies, a CDN can be setup and running in a few weeks”
    I’m not sure how effective that CDN would be. It took Akamai 10 years+ to build their platform, and Limelight 8+. Both companies keep on extending and innovating capabilities based on feedback of thousands of customers and carrying 15-20% of the Internet’s traffic. No CDN setup in two weeks will have that kind of operational intelligence – it will take years for it to catchup.

  • Hi Paul,
    (risking a delete by Dan for potential spamming 😉 I invite you to read my blog that exactly covers this. Click my name: the articles ‘Why are telcos deploying CDNs’ (strategical), ‘CDN layering’ (architectural). We have dramatically lowered the entry barrier.

  • shaun noll

    Hi Dan,
    do you think all these partnerships between Limelight and the telcos make Limelight a less attractive acquisition candidate for a telco looking to get into the game?
    It just seems like a relatively cheap way for a telco to become a big player, since LLNW could likely be acquired for <$250M (note LLNW has $150M in cash) but if LLNW has tons of telco partnerships then it becomes more complicated for a telco to acquire them since they would likely immediately lose some business? would be interesting to hear your thoughts. thanks! great blog!

  • Hi Shaun, I don’t think these partnerships stops anyone from acquiring Limelight or any other CDN. Right now, none of these resellers or partner deals are responsible for any large portion of Limelight’s revenue. While that could change down the road, today’s it’s not having a big financial impact on the company. Also, I don’t Limelight would go for $250M. They will have about $140M in revenue for 2009, so if it went for 3x revenue, that alone is $420M. I see them being more a $350-$400M price tag, in today’s market conditions.

  • Streamingguy

    Dan as usual there are no shades of grey in your posts. AT&T most definitely runs a CDN and continues to invest in this service offering. In 2009 AT&T expanded the CDN’s international footprint and continued to add features. AT&T is continuing to investment in 2010.

  • Where did I say anything about AT&T not running a CDN or not investing in the space? AT&T is listed on my CDN list at – always has been. I don’t understand what your argument is as this post has nothing to do with AT&T.
    If you are taking offense that I only mentioned Level 3 and not AT&T, you didn’t read the post carefully enough. I said, “Over the past two years, more than a dozen telcos and carriers have entered the CDN space.” AT&T would not be one of them. AT&T has been in the CDN space for a lot longer than two years.
    I don’t see how the fact AT&T is investing in the space is relevant to this post? This post is not about who is or who is not investing in the CDN space.

  • Streamingguy

    Apologies if I did not read your post carefully enough and you are correct that AT&T has been in the CDN space for more than two years, but I think this is easily done considering the caption “Telcos Not Yet Willing To Run Their Own CDN”. Also it is my belief that Level3 has also been in the CDN business for more than two years.
    My only point about AT&T’s continued investment is that they are willing to run their own CDN and this is evividence of this fact. FYI, I did not take offense, I just wanted to clarify as it appears others have not read your post carefully enough.