The Video CDN Business Is Flawed, YouTube Subsidizes Video Bandwidth On The Net

While many want to imply and insist that video delivered over the internet is going to one day rival or surpass cable TV as the main broadcast medium, which is an incorrect notion on its own, most are missing the bigger picture and ignoring the business side of the discussion. They want to argue about network capacity, improved compression algorithms, server deployments and other technology pieces without realizing or even acknowledging that the business model of delivering video does not work. Delivering video over the web, even at scale, isn’t profitable, unless you’re YouTube and you can afford to subsidize the bandwidth of nearly the entire Internet.

Even at scale, you can’t make money from delivering video on the web. It’s a flawed business because unlike traditional broadcast TV and radio models, with video on the web, each new viewer you get costs you money. You don’t have a fixed cost when it comes to delivering video and the cost to a CDN always rises. Just ask any content delivery network why they are diversifying their revenue away from video and focusing on high-margin value add services. They can’t make money from video only content, and that’s not going to change any time soon.

Last year, Limelight Networks cut their capex from about $40M in 2011 to about $25M in 2012 as the company decided they could no longer spend money to build out scale for video customers they don’t make money from. And last month, Akamai, the largest service based CDN, said on their earnings call that they will be, “winding down some contracts with a few media accounts in the first quarter that are not of long-term economic value.” They also said that moving forward they would, “refuse to pursue deals that we deem to be unprofitable or of little strategic value”, pertaining to their media and entertainment CDN business. This is Akamai’s way of saying these contracts didn’t have margins that were good enough to make it worth their time.

Delivering video on the web isn’t profitable. No CDN today is profitable, based solely on delivering high volume, low priced bits on the web. The only reason there is so much web based video to begin with is the fact that Google subsidizes it. And while some will argue that the price of bandwidth will always decline, it’s not declining at the rate it used to. Between 2009-2012 the industry witnessed the slowest rate of decline in CDN pricing ever (www.cdnpricing.com). On average, pricing was only down 20%, versus previous years dating back to 2000, when CDN pricing would drop at least 45% every year. Today, CDNs can’t afford to give it away, their costs are always rising and no one can debate or argue about their lack of profitability.

Many are quick to point out that online video advertising is growing and that’s going to allow content owners to monetize their content and as a result, make more money and be able to pay for more video delivery services. That would be the case if video delivery pricing fell at 30-40% each year, but it’s only falling at about half that. It’s one of the reasons that all of the projections made years ago of just how big the online video ad market would be are completely off. None of those numbers came true and part of that is due to CDN prices not falling as fast as some thought, which has had a direct impact on the growth of the online video advertising market.

Broadcast TV networks were built to distribute video and the Internet was built to distribute everything but video. Large objects weren’t even around when guys like Akamai started building their CDN and even fifteen years later, the Internet is still not capable of delivering TV quality video to the number of subscribers cable/satellite have, at the same quality. When a webcast does take place, for everyone who says they got the stream, there are always others who had problems. Just look at the recent webcasts of the Super Bowl, launch of the iPad Mini, Red Bull Stratos or the recent PlayStation 4 press event. They all had problems, for a large portion of viewers.

Every time I write about the topic of cable versus Internet video distribution, some always want to argue about technology. But no one can argue about the business model of delivering video on the net, which to date, has never been a profitable business for any CDN, at scale. In 2000, I remember many in the industry saying that in five years, everyone would watch all video online, cable TV would be dead and the Internet would give you the ability to watch anything, anytime, on any device. Remember Quest’s “ride the light” commercial? That was thirteen years ago yet still, many continue to preach that Internet video will displace cable TV.

Some might argue that Netflix has already proven this model as they currently have more subscribers than a lot of MSO’s, but comparing a service that costs $9 a month to a service that costs $50+ a month isn’t exactly fair. Not to mention, Netflix’s quality isn’t even close to what you get via cable from an HD channel and they have less choice. Also, no one seems to mention that Netflix is being forced to change their CDN strategy and move away from using service based content delivery networks due to the fact that the CDNs don’t want Netflix’s video delivery business anymore. Sure, they love using the Netflix name as a customer, but they all make slim margins on Netflix’s traffic and the capex required to support Netflix is huge. In Q4 of 2010, Level 3 disclosed they spent at least $14M in capex to add capacity to their network, just for Netflix. Level 3 is the one CDN who actually owns their network and has a lower cost of delivering video, so for them, the Netflix business makes sense. Three years ago all the major CDNs were fighting over Netflix’s video business but today, they no longer see it as a smart use of their network capacity or resources. What does that tell you about the business of delivering video on the web, even at scale?

And even with the business CDNs already have today, many still struggle to deliver it with good quality. Conviva measured the video quality of 22.6B video streams in 2012 and found that roughly 60% of all streams experienced quality degradation. Viewer interruption from re-buffering affected 20.6% of streams, 19.5% were impacted by slow video startup and 40% were plagued by grainy or low-resolution picture quality caused by low bitrates. And these streams are the ones that matter as they come from content owners like ESPN, HBO Go, Turner, Disney, Vevo and others who actually have content consumers want to watch and a way to monetize them. 4% of all streams Conviva measured never started. When was the last time you turned on the TV and your video signal didn’t work 4% of the time? Video delivered over the web simply isn’t reliable, for large audiences, at good quality and things like multicasting, better codecs, more fiber etc. isn’t going to change that.

The reality is that without Google subsidizing a large portion of the Internet’s bandwidth, thanks to YouTube, we’d have far less video on the web today. But even by YouTube’s own data, they monetize less than 25% of all their video streams every day. So they are paying to deliver billions of streams each month that no one can make money from. Does this sound like a business model that makes sense or better yet, one that will allow streaming media technology to displace cable TV as the broadcast medium that the majority of people will use to get their video? No chance. Those who suggest that cable TV will be replaced by video being delivered over the web can debate and argue all day about the technical details, but you can’t argue with nearly sixteen years of CDN data which proves that it’s a not a business companies can make money from. If there is not a profitable business model behind any technology, it can only go so far.

I’m not down on the CDN market, it’s still a very important part of the video ecosystem and many vendors will still be in the industry for years to come. But video delivery is not the portion of their business they will make money from and it’s not the product that will grow their top line. Profitability is now the measure of a CDN’s success and not the number of video streams they deliver.

  • Dan – Thanks for definitively validating what I expected (since 2003) was the case. I do think there are or will be a few novel twists on OL video that will make business sense, but having you publicly validate that the current mainstream (no pun intended) models are broken is helpful.

  • BroadcastAssassin

    None of this matters. The current infrastructures are inadequate, sure. The current models aren’t profitable? Doesn’t matter. The sea change is not about technology or corporate interests. Consumers want video, and they’ll get it. While you and I can remember a time before the internet, an entire generation has now matured to adulthood with the internet as an integral part of their life experience.
    People who recognize that will make money. People who fight the inevitable will not.

    Here’s what’s actually going to happen: companies, either future or existing, will either upgrade their infrastructure or find new technologies for delivering content. Consumer demand for bandwidth today is minuscule compared to what it will be in the future, and our society is increasingly connected for communications, media, and other aspects of their daily lives. Profitable or not, the CDNs (or their successors) will have to continue to invest in fatter pipes to meet (or preferably exceed) the demand.

    It’s not like if Akamai decides video isn’t profitable people will stop watching. “Whelp, video was fun while it lasted but the CDNs can’t make money so I guess we’ll all stop watching it,” said no one ever, nor will they.

    You can’t base a thoughtful analysis on the assumption that all there is now is all there will ever be. 13 years (or whatever measure you want to use) is an insignificant amount of time. It’s not like the TV infrastructure you so fervently applaud existed a decade and a half into broadcast TV. It got built because the consumers demanded it. The arrogance to declare the death of online video in the face of overwhelming global demand because the business models are still developing is laughable.

    If you really want to do some bankable predictive analysis, you need only look at the history of the television industry that you get paid to reassure. After all, that industry had more than its share of early doomsayers (as did cable). They, like you, used inconsequential factors to support their analyses while ignoring the only factor that ever really matters – market demand.

    • bobsponge42

      BroadcastAassassin, Right, but the question is how long will it take to make online video a more profitable proposition? Simply because people want content cheap, fast and now doesn’t mean it’s going to happen soon. Nor does it mean if you jump out ahead and meet these new demands that you’ll have a profitable or sustainable business. Your business may simply be another rail laid down the long track of the path to the future.

    • danrayburn

      If you think “profitability” does not matter, then I can’t really debate the topic with you as companies don’t survive on ideas, they survive on revenue.

    • Most infrastructures in all walks of life are inadequate, The closest thing I can liken this to is Concorde, it was the super fast luxurious fat pipe we all want to delivery our video over but it wasn’t sustainable, therefore it got canned and we are all now subject to a slightly slower more cramped alternative but at its price point it works and completes the job at a level that mostly everyone is happy with.

  • Steve Lerner

    I’ve been saying this for over 13 years… video cannot be profitably delivered, from a bandwidth perspective, over the public internet. This is due to video’s business model requiring fixed costs, and network bandwidth being variable cost. Yet, for the same reasons that large amounts of the internet’s own infrastructure aren’t profitable (telcos, colos, startups, other technology firms) yet survive, so too will video. Its sort of like a public good at this point- it will never make money, but you can’t turn it off.

    -Steve Lerner
    Senior Member of Technical Staff, Network Engineering
    eBay, Inc.

  • I appreciate what you are saying, and point out a couple of things:
    1. Comcast, the cable company, is converting to IP delivery. That’s a sea change. I’m wondering exactly how they are doing it. Are they moving content closer to the customer to avoid congestion, with caching and additional servers? Don’t know, but they obviously have found a way to sell and deliver content at affordable and profitable rates.
    2. I have costed out 2 hour delivery using Amazon AWS and Wowza, and factored in the cost of development and advertising. Content (movie) delivery is profitable using this model. If you go to the CDNs that provide all the services for you, it isn’t profitable unless you get into really big numbers. Advertising is a major factor, but if you are smart about it, you can do it profitably.
    I enjoy your articles and your seasoned perspective.

    • danrayburn

      Hi Dorian, you can’t compare video being delivered over a closed, private network like an MSO, versus video being delivered over the public Internet. MSO’s own the network, can bring technologies in-house to manage the delivery of video. Most are doing it purely to save money, but over time, will try to monetize. Service based CDNs can’t do that, they don’t own the last mile.

      Video of how Comcast is doing it: http://bcove.me/qxeg83sh

      Delivering movies via a service based CDN is almost always profitable because most consumers pay per stream to watch it or per download to own it. But most content owners delivering video over the web don’t have the advantages of a major studio.

  • Cary

    I think the problem is a lack of smart people trying to be competitive inside the video streaming industry and what we are seeing isn’t so much a failure of the concept as it is a stagnation of innovation in tech and content production. It feels like a lot of streams I see are companies saying “throw an internet stream up, that popular these days … right?” and then treating it like a TV production (at best) instead of thinking through the separate logistical/content production concerns of internet streaming.

    I’m curious if you have looked at twitch.tv. It seems like they gets overlooked because people naturally assume youtube/major TV broadcasters will be running with the latest tech but when you compare them to twitch it feels like you are comparing a coal powered steam engine to a 200mph bullet train. Twitch is an interesting example as they came into a market (Streaming professional video gaming events) where internet streaming is the only way to consume content. The viewers are always connected to their online community/twitter while watching content so any problems with the service and the pitchforks go up visibly very quickly. That has forced twitch to address all the problems with better tech and forced the content producers to focus on creating a more seamless broadcast. I watch steams on twitch several times a week and all of the issues most people have with streams are non-existent on twitch.I can’t remember the last time a stream didn’t load or the quality dropped suddenly or I experienced bad lag. For me trying to watch the stream for the PS4 event (or the Olympics over the summer) was really jarring because I was so used to streams on twitch being flawless.

  • Lon David

    1. Sure Youtube subsidizes video, but so what? You seem to think the only “value” is the old school model of delivery. Dollar per video, or per minutes, etc. If you think that Youtube isn’t valuable to the overall Google business model, you’re missing the point. Just like Amazon and many others, they sell some things at a loss in order to make money in another area. Even if the broadcast is a loser, the ability to bring people to my site and have them purchase other things is of great value. I might pay (to the CDN) more than I take in for the content, but there’s other sales to be made.
    2. You keep saying “no fair” when people bring up the idea of direct IP delivery of content. But it’s quite fair. Sure it’s more expensive from a broadcast perspective, but the granular detail with which it is delivered is VERY valuable to both the consumer and the producer. This one-to-one exchange is the future. I pay several CDN’s to broadcast (deliver) my content. Amazon S3, Vimeo, Livestream, & Watershed. I pay them to stream my content (live and recorded). My customers pay me for my content (and other things). This business model of delivering video DOES work. And yes, it is the future. You seem to be approaching this from a theoretical view, but from someone engaged in it, it works. For all of us.

    This is still a relatively new industry and will grow as we learn to use it better. Go the way of the Old School print media & music industry if you like and claim it will never die. How did that work out?

    • danrayburn

      “it works. For all of us.” – No it does not. The CDNs delivering your videos are all losing money and aren’t profitable. Ask them why none of them are profitable and are moving away from the video delivery business.

      • Lon David

        Nice try. None of them are moving away, they are just refocusing their target on pro (paying) users versus giving it away for free to anyone and everyone. Ustream, switching to pro customers, Livestream same. Vimeo same. I for one stopped using a free service (Youtube) and instead PAY Vimeo. Amazon’s doing pretty well thank you very much and growing their cloud delivery services as fast as they can. It will work because people like me will (and do) pay for it. I assure you I am a profitable customer to them. And there’s a lot more like me out here, they just need to find us. Music industry – changed. Publishing industry – changed. Newspaper industry – changed. Television…on it’s way…ignore it at your peril.

        • danrayburn

          Ustream is not a CDN. Neither is Livestream. Or Vimeo. They are customers of CDNs, they are not CDNs themselves. And they don’t sell delivery, they sell a cloud based platform that enables you to broadcast live content. That’s not what an Akamai, Limelight, Level 3, EdgeCast offers.

    • danrayburn

      “Just like Amazon and many others, they sell some things at a loss in order to make money in another area”

      Exactly. And Google can afford to do that with YouTube because like you say, it’s valuale to their business in other ways.

      But CDNs can’t compete with free. They can’t afford to give away video delivery to get other business. That’s the whole point.

  • This is very in line with what we have been saying and predicting:

    Internet CDNs were never really built for video delivery. They started as website caching platforms and tried to bolt on video support afterwards which resulted in inefficient architectures. They don’t own the pipes. Because of this approach, technically they can scale but cost-wise they can’t.

    Internet CDNs use the best effort Internet and use best effort technologies such as caching and DNS to deliver video. Best effort is not good enough for content publishers and consumers. Best effort cannot compete against broadcast grade IPTV and digital cable.

    In fact the entire OTT value chain is broken: CDNs don’t perform and lose money on video delivery. Telcos face rising costs and no additional revenues. Because streams don’t perform, consumers don’t mass adopt new services. So OTT content publishers can’t build a large scale business. So they can’t pay CDNs a reasonable price to invest in better performing delivery. This is a deadlock situation.

    Basically all Internet CDNs got inspiration/mimicked/copied/stole (however you prefer to call it) their approach from Akamai without any real innovation. And although Akamai as a first mover got a massive chunk of the video delivery market, their profits are not from video but from app acceleration, cloud and enterprise services. And is refocussing on those drivers. Why invest in video if it is a cost driver but not a profit driver? So they walk away from video delivery. Limelight has never been profitable. Highwinds already reshifted focus. Level3 only went for large volume prospects who typically are not loyal and are low margin customers, like Netflix etc. Some Internet CDNs tried to move up in the value chain by offering OVP services. But that meant they lost focus.

    After being in the streaming business for 18 years now, all I can conclude that the Internet still does not scale and does not offer the required performance for mass audience high quality video. And it probably never will.

    So to think out of the box is to bypass the Internet. Last mile, deep-edge access provider enabled delivery. Cut out the man in the middle (Internet CDNs, carriers, exchanges) to guarantee better performance.

    And the next step out of the box is to rethink the way CDNs are architectured and which technologies are being used. DNS is unreliable and uncontrolled. Best effort. Instead you need active request routing engines to reach broadcast grade uptime levels. Caching is passive as well. Using DNS and caching you hope and assume that an end user is sent to the right edge, and that this edge will be able to cache and deliver remote content in realtime. No guarantee whatsoever. That is not broadcast grade.

    What good is an Internet CDN for a premium content publisher if they can’t guarantee performance, capacity or uptime? What good is an Internet CDN for any content provider if the CDN has no sustainable business case? They may get a cheap deal but that’s short term thinking.

    I did manage to find a sustainable business model for video delivery with StreamZilla. As far as I know it’s the only CDN that 100% focusses on video, bypassing the regular Internet, using more controlled CDN technologies. That resulted in being the only CDN that has approx. 50% profit margins and year on year growth since 2004. So it is possible, proven.

    And the above is also why I strongly believe in (software /cloud based) wholesale CDNs within telco networks, whether they are managed or licensed or deployed via other constructions. The telcos that fail here are those who again simply try to copy Akamai or use regular Internet CDN technologies, since there is no real competitive USP. The telcos that win are those who have a combined strategy:
    – Defensive: On-net premium delivery of retail content to their subscriber base;
    – Offensive: On-net premium delivery of wholesale OTT content;
    – Cost reduction: efficient traffic flows and operation

    So it is possible to break through this OTT value chain deadlock situation. In countries where we deploy CDNs for telcos, traditional Internet CDNs lose business, telcos find a new revenue stream, consumers like the performance, so publishers make money. And the entire value chain costs are lower. Proven as well.

    About OTT ever replacing digital cable: no it is not going to happen. Cable and IPTV are superior for linear broadcasting. It scales. However you need critical mass which is quite hard. And VOD is limited. There is a huge market for online video that cable and IPTV don’t and never will address. So for this market, online video is superior. There will be some overlap but the two will mainly coexist.

    p.s. It will be interesting to see whether LTE broadcast will get some traction. DVB-H never took off.

  • Acct_exec1

    Dan, LLNW has Playstation as a customer and was an active participant in the sale of Gaikai to Sony. With the upcoming PS4 allowing full library downloads as well as streaming demos, I expect that LLNW will be a primary partner in content delivery. Google fiber is putting pressure on the big boys and there needs to be a real reason for customers to want high speeds. As of right now video is a novelty that’s nice to have, but with the next generation of gaming there will be an increased push for higher speeds to facilitate a need. I agree that video streaming may be unprofitable but what are your thoughts around the next generation of gaming being a catalyst for growth in this space?

  • RBR

    Excellent post Dan, Stef,

    Eyeballs vs Content. This is where the worm shall turn. Under legacy business practices, the owners of the eyballs (aka service providers) must purchase content (aka tv programming, owning parts of the PSTN, etc…), or access to content (aka Internet Access) wholesale and mark it for retail to the eyeballs. This surpresses incentive to increase local access capacity and presents local access providers insecure sources of revenue. CDN’s represent a unique position to aggregate and manage the quality delivery of access to the eyeballs, on behalf of the service provider, and sell that access to the owners of the content. Allow the market to select and determine the value between Best Effort and CDN managed access and remove the chimerical cost of purchasing Internet Access. $25 / M, $17 M, now $1 / M with 1G or burtsing up to 10 G no more than the 95th percentile. Oh, and when you’re big enough lets just peer directly and pass traffic bill & keep. Can you say Netflix Open Content Delivery Network. CDN’s usurp Internet Access for those who know how to monetize their value. Sure, the big players will still vertically integrate their local access with a CDN, but there are lots of smaller providers who could benefit from the scale and talent base offered by CDNs.
    The days of differnetiating Internet Access by speed or capacity are fading. As are the content based revenues for access providers. Time to refocus and monetize the comparitive advantage each industry posses. Access to Eyeballs for the local network access providers, quality delivery of content for the CDNs. We’re beyond “If” entering “Who” and “How”.
    Cheers.
    Robert

  • Imagine if Intel walked away from the chip space in the 70s instead of following the guiding principles of Morse law. The CDN industry has a similar guiding principle called Wheaton’s law. It states investments in R&D usually increase the efficiencies of delivery to the point that prices can decline at a rate that offsets the general growth in a publishers traffic resulting in a generally flat to moderately increased fees year over year. Argue that confusing sign up screens for TV Everywhere are hurting some publishers by keeping consumers out which lowers ad avails and the corresponding delivery deals. Argue that YouTube is taking eyeballs away from publishers reducing their overall traffic contracts at CDNS. Argue content costs went up over 20% last year putting the squeeze on publishers. Saying CDNS don’t want to sell video delivery is like predicting the next pope will not wear a funny hat.

    • danrayburn

      Except that the CDN’s “are” saying they don’t want to sell video delivery. It’s not that I’m saying, I’m reporting on what they are saying. They don’t want to sell video delivery where the volume is high and the price is low.

  • Kathy McEwen

    Great article Dan! It does appear that YouTube is subsidizing their own CDN costs. Are they subsidizing any of the broadband access providers costs? What % of the total web delivery cost is the caching/streaming server vs the access/network cost to carry it?

  • Sonnati

    Interesting article Dan, but I agree only partially: I agree that internet streaming will not substitute anytime soon cable tv because of inerent infrastructure limits, but I disagree that online video is not a sustainable business. Among the cable tv subscribers there are a certain percentage of users that make a limited use of their subscriptions. A web based service is an interesting alternative for them, even because 1hr of streaming at 3mibt/s avg may cost not more than 5c$ and so an avg consumption of 2-3 hrs daily would cost not more than 4.5$ per user per month which is less than cost of the cable. It’s a matter of business model. Adv based models are much more critic, subscription based models are better. the fact then CDNs are.running away from big, low redditivity, video delivery contract is a contingent problem of their specific business: if they have low margin is simply because they have strugled with each other to gain big contracts lowering too much redditivity. It’s a momentary problem: they have too low matgins ? They can simply raise their prices. That said, internet video will lower the huge broadcasting market of negligeble amont for many year, probably

  • bigpaws1969

    Great article Dan. It should be well known by now that Akamai’s high margin enterprise business subsidizes the low margin (yet sexy) Media & Entertainment business. But they can’t do so forever. In contrast, Limelight never had an enterprise business and that’s why they can’t make a profit. They tried briefly, but they didn’t have the staff, tech stack, or patience to make it work.

  • MK

    Dan – for a moment (hypothetical) assume video delivery costs are zero. If publisher video acquisition costs are nonzero, then for pure ad based sites it’s all about ensuring CPX’s are high enough to defray video and subscriber acquisition costs. I fear that, video CPX’s over the Internet for an average site is far lower than ad CPX’s for an average cable channel. Also, cable channels typically get $$ from cable MSO’s. So the problem with Internet Video Monetization in my humble view has nothing to do with CDN pricing. As a consultant, I am living and breathing this with publishers. Internet publishers also incur a higher cost wrt customer acquisition, compared to cable channels. While your article may be a restatement of facts you hear in the industry (from CDN’s), the issue of CDN pricing is actually less of an issue for publishers. It’s really a revenue problem. Having said that, I feel the train has left the station – Internet video here to stay, and the market will figure out the economics soon.

  • James Van Leeuwen

    What happens when we get publicly controlled open access fibre networks that minimize cost of carriage for everyone, like our publicly controlled open access road networks?

    Also, what happens when the $$ that cable customers are forced to pay for content they DON’T want is suddenly available to pay for content they DO want?

    The real question is whether cable operators can maintain the status quo of owning their customers, to which I would answer, “Not a chance.”

  • Akamai is being smart. Who wants to race to $0? In the last couple of years I’ve seen pricing go from $1/gb (to start) to under $.30 or less (to start). Some of the requests from customers I’ve seen have been in the sub 1 cent range for not a lot of traffic.

    Really, it was the discount CDNs who stepped up their performance who started forcing the premium brands to drop their prices to compete. It used to be you’d just tell your customer that if the discount guys were inferior. Now the discounters are on par, or better in some cases. I say it’s the CDNs own fault for letting the pricing get so low.

    Now I have to disagree a little Dan. CDN costs are not constantly going up. As they add more and more traffic, their bandwidth pricing goes down. Albeit not by the same percentage as the discounting in pricing goes. Still I think costs remain about the same if not go down a little bit each year. As long as the CDNs can fill the valleys of traffic with some decent, they are making a profit.

    But, you are right, the margins aren’t there like they were. Any CDN which doesn’t diversify and offer a full stack of products like Professional Services, OVP, Transcoding, Mobile Delivery, Advertising, CMS, DSA, FEO, etc will be in trouble in the next couple of years.

  • Guillaume

    Good article, but maybe a bit too US focused. If you’re looking at France for example, triple play is massive. Every ISP provides a TV service via its box, at the same HD quality as cable. Cable/satellite there is not working at all, just a few subscribers, where near 100% of the personal internet subscribers is using the internet to watch TV.

    Also, when you’re saying that CDNs are going away from video contracts, it’s not true; Akamai, Limelight and Level 3 are fighting at the moment to win France Television, for CDN but also video delivery.

    To finish, you’re right when saying that CDNs don’t make their money from video delivery, they’re doing it on professional services sold around video delivery, like iOko / Kit Digital is doing.

    Anyways, that was a good article, thanks for that!