Heavy Reading Research

What's Holding Back the SuperCDNs?

Content delivery networks (CDN) such as Akamai Technologies Inc. (Nasdaq: AKAM) and Limelight Networks Inc. (Nasdaq: LLNW) generate billions of dollars in revenue for accelerating content delivery and improving the end user's quality of experience (QoE).

But CDNs don't own the last mile – service providers do. Service providers can cache content closer to the end user, and thereby offer a better QoE. They can also use network management tools to better manage traffic and even guarantee quality of service. They also own the billing relationship and can create new tiered services, or target advertising for additional revenue.

These advantages should allow service providers to create a "SuperCDN" capable of adjusting faster and more effectively to network traffic, avoiding upstream congestion by serving video from the telecom edge and being more innovative with pricing models.

Some service providers have also discussed the idea of wholesaling CDNs. In this scenario, the operator-owned and -managed caches would be leased to traditional CDNs. This would allow for open access to the operator's network by other service providers, content owners and perhaps existing CDNs, but enable operators to share in the revenue generated by over-the-top content delivery.

Several Tier 1 operators are running their own CDNs today, but they don't offer the unique differentiators mentioned above. The factors holding up the SuperCDN include:

  • Operator prioritization: While they often complain about OTT providers making money over their infrastructure, operators aren't really tackling this space aggressively. This could be due to conflicting priorities and a lack of clarity around the business models for a service provider-owned SuperCDN.

  • Traffic management hand-offs: For the wholesaling model, developing interfaces for the hand-off of content assets from the traditional CDN to the service provider's cache is a challenge. The operator's traffic management systems need to be able to interact with the CDN's traffic management systems to ensure that content is cached and delivered in the most intelligent way possible, seamlessly from beginning to end.

  • Net neutrality: As with any network management approach, the uncertainty around net neutrality and acceptable traffic management is an issue. While transparent caching appears to be a generally acceptable practice even in the U.S., commercializing a content delivery operation, using policy tools to prioritize (or deprioritize) traffic, and exploiting any network ownership advantage could create controversy, bringing down the wrath of regulators. However, outside the U.S. this is less of a hurdle, and operating a wholesale model would likely calm many objectors even in the U.S.

  • Media delivery commoditization: The flood of CDNs that entered the industry a few years ago have succeeded in dramatically reducing costs for content delivery. However, the massive and continuing growth in OTT video still makes this an important opportunity for operators.

With video traffic forecast to grow steadily, caching and content delivery will be an important issue for operators in the future – not just to alleviate OTT traffic costs, but also to explore new revenue-generating services. Despite relatively few announcements, trials and tests have been conducted by several of the largest service providers in the U.S. and Europe. We anticipate this will be a key area for service providers as they try to cope with the flood of video on their networks and look for ways to generate incremental revenue from it.

— Aditya Kishore, Senior Analyst, Heavy Reading

ycurrent 12/5/2012 | 5:09:51 PM
re: What's Holding Back the SuperCDNs?

Exactly - there seem to be many ways to save money with the intelligent and efficient distribution/delivery of a CDN, benefiting an operator, a content provider, etc.  But actually getting more revenue, as a "big" business for operators to get into... still needs convincing, I think.

ycurrent 12/5/2012 | 5:09:51 PM
re: What's Holding Back the SuperCDNs?

Vendors driving operator-owned CDN marketing point to revenue opportunities leveraging last mile network ownership.But the question remains: which content owners are likely to pay for this benefit?

If it is global content providers, how much additional $$ are they likely to pay an operator on top of what they already pay global CDN providers to handle their content transport?  The business case seems somewhat stronger for local content providers, where they are likely more reliant on operator networks rather than global CDN providers (although not in the US).  But again, how much are local content providers willing to pay?

If we apply the 80:20 rule to the video driving network traffic, is 80% of the traffic being driven by 20% of the content being sourced from global or local content providers?  This matters.  Because if video traffic is more homogeneous (i.e., globally sourced from a few providers) then there is much less revenue for operators than some business modelling would seem to imply.

paolo.franzoi 12/5/2012 | 5:09:51 PM
re: What's Holding Back the SuperCDNs?


If you are a global content player (let's just say Disney), how many of these deals would you have to sign and then pay for.  30?  25?  That is the reason that it goes nowhere.  You do one of these deals as a content owner because it either:

- Pays you more money or;

- Saves you expense

So, which does this do for a global content owner?

I assume a local content owner has fewer people to pay and if the content was featured might make the owner more money.

I still don't get the idea that somebody like LR is likely to pay AT&T to make sure that its pages load fast.  Now if it is Netflix - perhaps there is a way to save Netflix money.



paolo.franzoi 12/5/2012 | 5:09:42 PM
re: What's Holding Back the SuperCDNs?


I think you are missing the point.  If you believe that IP distribution of video is a lower cost way of distributing content, then that is a fine thing.  CDNs HOST content for further distribution.  Hosting content and distributing content are two distinct things.



craigleddy 12/5/2012 | 5:09:42 PM
re: What's Holding Back the SuperCDNs?

Though the economics may be a bit hazy in the near term, there are long-term advantages for service providers to create CDNs. Near-term, a service provider can get a handle on the trafficking issues that Adi identifies. The long-term prospects involve a re-thinking and maybe even a re-definition of what a CDN is. In the cable arena, major providers like Comcast are exploring a migration of current MPEG2 TV video content to IP video delivery, a migration that will take many years. They can use their CDN in the near term for Internet traffic and eventually could use it to support IPTV delivery. With fiber backbones, a CDN and IPTV, they could reduce their reliance on satellite delivery for TV content and create a seamless IP  service delivery network. What do you think?                   

craigleddy 12/5/2012 | 5:09:41 PM
re: What's Holding Back the SuperCDNs?

Agreed. But maybe this is where a re-definition of a CDN is needed - or perhaps a new term to describe what's going on here.

And if you're a service provider that can both host content and distribute it, doesn't that make you a pretty powerful force? (Net neutrality bombshells aside)


paolo.franzoi 12/5/2012 | 5:09:40 PM
re: What's Holding Back the SuperCDNs?

I don't understand what you are trying to drive at.

You first talked about an IP Video distribution scheme (by the way FiOS is already doing that).

The whole point about hosting and having local access is that it is basically useless to the large content provider.  The problem is that then you would have to cut a hosting deal with every local access provider in the world.  Which is well beyond any possible cost to make it worth it.  So, basically there is no value to the content guys offering a service which requires a local access CDN.  If you want to be in the global CDN business and compete with Akamai then go for it.  If you think Comcast (example) creating a CDN to get content to its customers will sell to content owners, then I believe that you are completely wrongheaded.  If somehow you create a Global CDN and find some way to lower the cost to deliver to your own customers as a local access provider then that is a win for you.  Makes no difference to the content owner.  He has to build his service on NOT having such a deal.  Once he has done so then he will NOT pay you to give him something he does not need.

Remember from a content owners standpoint - he wants his service/content to work for any customer with an Internet connection.



ycurrent 12/5/2012 | 5:09:40 PM
re: What's Holding Back the SuperCDNs?

and so, the revenue model for operator CDNs remains elusive...

paolo.franzoi 12/5/2012 | 5:09:30 PM
re: What's Holding Back the SuperCDNs?


Who pays whom on those direct content deals that the Tier 1s have?

That's right - The carriers pay the content owners for the rights to distribute the content.  In fact, a Cable Company's number one external expense is buying content.

So, I agree if they can save the Content Owner costs on distributing content then they might get a deal.  But Content Owners will not make that exclusive, except in some really rare circumstances (see NFL Sunday Ticket).



yarn 12/5/2012 | 5:09:30 PM
re: What's Holding Back the SuperCDNs?

Hey Seven,

I think you are right in general but tier 1 network operators do have a sufficiently large subscriber base to make direct deals with content providers, avoiding the need for additional men in the middle. They already have direct content deals in place for regular TV service so they should be able to extend those for delivery over broadband access as well. Whether a content owner would have to pay a network operator for delivery is a different matter. In fact, some operators have offered to pay additional affliate fees for on-line distribution rights, so a content owner would actually save on CDN cost.

That doesn't mean the days of Akamai are numbered. Besides the fact that an Akamai does a lot more than video streaming, content owners would still want to have a default on-line distribution option for complete coverage.

Sign In