The Great Internet Video Lie

Internet Video. Its the salvation for content creators everywhere. Its the end to dependence on the big bad meanies, the cable and satellite companies. Right  ? Hell no. The concept that “over the top” video creates a valid business alternative for content creators is as misguided an internet business myth as there is.

For grins, lets say you want to start a business for which you want to stream, live or on demand, any video. Any quality. You want the ability to reach merely 10k simultaneous viewers.  Not a big number. In fact , its a tiny number.  Its certainly not competitive with any form of traditional TV, but its a starting point. So to stream 10k simultaneous streams, what are your choices ? 

Can you just do it from a PC in your dorm room ? From your house on your cable or DSL line ? No.  You will need an outside vendor in order to offer a mere 10k simultaneous streams. Which leads to the question of who can provide a service as simple as 10k streams ? Who can do it with any level of reliability ?

The natural response of course would be to say to use Youtube, right ? Except that their terms of service prohibit commercial applications.  You can pimp your content and do ad shares, but you cant have control of your content and its presentation.  Its branded Youtube. Its controlled by Youtube. 

So what are your options  ? You have a single option. You have to use a Content Delivery Network (CDN). CDNs specialize in delivering content that needs to scale to large numbers, and 10k simultaneous users, particularly if its sustained for any period of time, like a cable network would be, is considered a large number. 

When Youtube did their big  live event, they used a 3rd party CDN.  Any large scale event streamed to thousands or hundreds of thousands of simultaneous users is going to require hiring and paying a CDN like Akamai, Limelight or one of just a few others that can offer scaled streaming. 

There in lies the rub. There are only a  few CDNs that can offer any level of scale for delivering video to an audience that is large by internet standards, but very small by cable or satellite standards.  There is not a single CDN that can deliver 2 or more  video streams concurrently to more than 1mm simultaneous viewers.  Not one. Anywhere.  There are probably 3, maybe 4, that on a perfect day might be able to deliver a single video stream to 500k simultaneous viewers.

On the flip side,  there are at least 8 large cable and satellite video distributors that can deliver 100 or more video streams, concurrently,  to a million or more simultaneous viewers.

The Great Internet Video Lie is that the internet opens distribution to compete with the evil gate keepers, cable and satellite.  In reality, if you have a desire to deliver a large number of streams, and you want to compete with another internet video provider to offer a large number of streams, you are not in a very good position. You are at the mercy of 3 or 4 CDNs, the ultimate internet video gatekeepers  

Let me put this another way. Lets say you have your “Worlds Greatest Concert” that you are sure can draw 500k simultaneous streams (on demand or live) . Also planning to stream a large event, lets say the first broadcast of Dark Knight 2, which the producers will stream live at the same time, and oh my goodness, the remaining Beatles decide to have their final reunion with a single live concert at the exact same time. 3 Mega events, each with an expected draw of 500k simultaneous users. Who has the greatest opportunity and the most leverage ? 

Thats easy. Its the CDNs. You have so few choices of vendors that the CDNs can charge whatever they want to handle the event. And thats for one single event. You dont want to know about costs for 24×7 streaming for viewership levels of even small cable networks.

The internet is not an open video platform. Video distribution of any scale places you  at the mercy of just a very few CDNs.  You literally have to compete for timeslots for very large events.  If you want an interesting excercise, call up a CDN and ask them how much it would cost to support an audience that is never smaller than 10k simultaneous viewers for a 1mbs stream, 24 hours a day, for 365 consecutive days. Then call up one of the satellite providers and ask how much they would charge you to deliver to 100pct of their customers, and then call up a cable company and ask the same question.  Total up the cable and satellite numbers and compare them to the internet costs. You may be surprised to see which is cheaper.

If you have dreams of competing with traditional TV network viewing numbers using the internet, dream on.  You cant afford it. You have been sucker punched  by the Great Internet Lie.

141 thoughts on “The Great Internet Video Lie

  1. In his own words:

    Comment by Beiron -

  2. Pingback: Contentinople - R. Scott Raynovich - The YouTube Model Is Dead

  3. Pingback: .TV - Major League Baseball Gets It « Be Innovation

  4. The Great Internet Video Truth: This years March Madness

    I have a choice of four livestream games, right now…

    Comment by Eric Ryder -

  5. Mark – I’m curious, you stated:

    “…we started streaming audio in 1994, video in 1997. 10 years ago we were doing gigabits of audio, with hundreds of thousands of simul users, 400k video…”

    If you were streaming video at 400 kbps in 1997, was anyone able to actually watch it considering there was very little broadband access for individuals?

    I found this an interesting discussion topic. To me, it is all about using the right technology platform for the right environment. If you are broadcasting content that will be large scale, streaming is NOT the right technology vehicle for large scale productions. However, if you want to broadcast a “micro event” that has a strong loyal following but lacks broad appeal, then internet streaming is the way to go…you have lower technology costs vs. TV and can deliver very nice content quality with the right combination of new technologies.

    Comment by Marc -

  6. I might be thinking backwards here, but don’t we need the content that demands 10m simultaneous downloads before someone drops the technology to make it happen?

    I think grand point of the “internet video revolution” is that anyone can put something up and distribute it. If you think of the internet as a indie distribution system and not as the same class of communication as television or even radio then it fits right in. I’m not 100% sure that the future of online broadcasting even as reliant on CDNs as you think. I recently heard a figure that every episode of Heroes is downloaded frlom bittorrent 5 million times. For all those that poo-poo bittorrent, that’s a huge number. It’s possible that bittorrent could distribute as much content as some CDNs.

    I think looking at CDNs and CDNs only is short sighted. There’s bound to be a large assortment of ways to get files for years to come.

    – Nick Allain

    Comment by Nick Allain -

  7. To me, the answer seems simple.

    Someone needs to start hammering ISPs to configure multicast from the CDNs. Sure, the ISPs will leverage their position to get some money, but really, you are saving them money in the end which of course can be leveraged against them.

    And for a reservation fee (which isn’t necessarily cheap), I can reserve almost as many streams as I want for a live event. Configure your player to do a fail over to the next CDN and your list of URLs start with the cheapest CDN.

    Or, the Move player supports live streaming, which heads out to the end user as a progressively downloaded Flash hash. To me, live events should be handled in a way that if I arrive 10 minutes late, I can start from the beginning, scroll through up to live. This way anyone who arrives late is watching the “live” stream as a progressive download but asking for packets that are behind which . CDN costs are much cheaper that way. Content is cached at the edge and many of your users aren’t trying to get the latest bits to view.

    There are so many creative ways around the problem that I don’t think it is much of a problem at all. One thing is that true live streaming is only absolutely necessary for breaking news. If I am going to pay to see a Beatles reunion show and I start watching it 5 or 20 minutes late, I want to see the whole thing. I want to start from the beginning. I want to be able to catch up, pause, ect.

    Entertainment streaming just does not need to be true streaming anymore….

    Comment by Michael Dube -

  8. wow.
    I had never really considered the serious numbers involved with streaming video.

    Comment by Home Projector -

  9. Pingback: Running Naked: Mark Cuban Just “Pulled Down the Shorts” of CDNs Everywhere « SiliconAngle

  10. Pingback: Dzoic ClipHouse Review

  11. Interesting article… but not really an eye-opening revelation. What you state is a fact most internet addicts wouldn’t want to admit, but has always been true. We got so used to video feeds pausing every few minutes that the days when this was all news to us are long gone. What’s interesting is that the Internet has grown and advanced despite the obvious inferiority compared to other content distribution methods. The reason of course, rather than compare the content distribution methods in terms of technology, a factor that is in the favor of the internet – and also its downside – is the sheer availability of contents and diversity. Internet is a mean of transmitting information that are otherwise unavailable in other forms, be it satellite or cable, or even those exact same services provided overseas that are not available over this side of the lawn.

    What also is one of the biggest “draw” of the Internet is that significant amount of content – as much as you business folks might not like it – is “free”, after the initial setup of your PC
    and the internet service, legally or illegally (of course, we all know there’s no such thing as free, hence all the content providers scrambling to make any sort of profit, are failing miserably).

    I agree with you that content distribution over the internet (or any content distribution, period, as far as I can tell) is not a sound business for companies trying to make a quick buck. There are literally thousands of companies failing for one semi-successful one, and even when you are semi-successful in terms of brand, you are not exactly profitable (youtube comes to mind).

    What I think the businesses should concentrate on is using the Internet for its advantages, e.g. it’s reach and interaction.

    P.S. On top of it all, I’m amazed you actually read all of the response and reply to them. It makes me wonder if it takes most of your time, and how you exactly manage your time, as I’m sure you are quite a busy man. I’m pretty sure you wouldn’t read this as it’s a response to a fairly old article, but keep up the good work.

    Comment by James Hsu -

  12. Pingback: That’s Just Dumb Old Media Guy Redux |

  13. This is why I only watch streaming videos from one source, plus I can watch premium channels as well.

    Comment by Super TV -

  14. Totally agree with you here. Where is it said that the Internet enables you stream 1080p to 100k simultaneous viewers?

    I would note that cable and satellite companies are no more evil than IP lawyers and the FCC.

    The Internet isn’t a free lunch. Move along?

    Comment by Alex Leverington -

  15. I think you’re right about existing CDN’s and the delivery side. The current problem is of the chicken-egg variety, and nothing inherent to the medium differences.

    I think what has people here riled up is the implication that there’s something inherently wrong with Internet video distribution, versus existing broadcast standards. My biggest problem with these arguments is the assumption that current broadcast methods are somehow natural. They are not — they were created by a mix of historical technological/thinking limitations and short-sighted government regulations. If you forced all the incumbent broadcast monopolies to give up their radio spectrum and cable monopolies today, because (as the FCC knows but refuses to tell) replacement spectrum-sharing technologies already exist, you could replace their delivery systems with multiple times the bandwidth for bidirectional data streams, more than sufficient for every room in every home to receive a unique HDTV video stream. If broadcast video were a must-have application in data networks, you would see all the current compatibility issues with multicast IP standards get resolved very quickly, because ISPs that didn’t catch up would be left in the dust.

    I think the offense is in the implications made, and not so much your actual words about the state of the industry today. Part of the problem with the industry today is the force of existing monopolies, which don’t make sense when modern radio technologies are considered. IGMP support would be the rule rather than the exception.

    The one place I would say your comments are factually false are concerning the place of P2P technologies in current/future video distribution. Use of tools like BitTorrent implies users are gladly trading some of their unused upstream bandwidth in return for more downstream download avenues — no “stealing” is involved. This maximizes legal network use around the asymmetrical nature of most consumer broadband connections, and utilizes the fact that many download nodes are “closer” (both in terms of latency and router hop count) to each other than they are to any server. BitTorrent streams are dependent on a tracker, just like web sites are dependent on web servers, so it doesn’t share the same “inter-user piracy” implications as other P2P services might. Just as the existence of warez sites doesn’t mean the web is primarily a tool for piracy, the existence of sites like The Pirate Bay doesn’t show that piracy is the primary use of BitTorrent.

    Comment by TakeyMcTaker -

  16. Point #1 – You are completely correct that the existing internet infrastructure cannot handle a mass transition of video to online.

    And why would anyone want to? If the only interest is in sitting back and consuming an event, there is no better solution. Of course, this is only applicable to events that are broadcast. Plus, why would I want to watch the Super Bowl on a buffered stream when I can have an incredibly greater experience on TV?

    Point #2 – There is a huge potential market for Tier 2 events online.

    I happen to be a big MMA fan, but if I wanted to watch Fedor’s last fight, my only option was PPV. So I buy it and sit back to watch. Meanwhile, I am texting pals who are watching the fight, chatting via facebook, etc since I like a more interactive experience.

    Now there is opportunity created. As a fan, I would pay to have access to that fight (plus the online market is larger than the PPV market based on number of potential viewers). Not just for the stream since if that was all I was getting, there is zero motivation to move from the TV, but for the interactive experience. Whether it is utilized as a second screen or as the first, an interactive experience where fans can gather, chat, buy merchandise, create highlight reels etc is a way to monetize an event that previously was not done. What would a fan pay to have “behind the scene” access to the dressing room, the corner, or even the fighter themselves in a live interactive format?

    Now apply that to the traditional tier 2 sports or music events where there is a large fan base (not 1mm, but say 250 – 500k) that wants access to an event that may not be televised like the Ironman World Championship in Hawaii. If all you did was stream it online, it would be nice for the fan but hard to monetize. But what if you could give the fan the control of the producer? What if that fan could simply click on a racer’s number and see where they were on the track then jump to the camera nearest them? What if they could then create highlight reels of the person they are interested in vs the one that happens to pass by a static camera set up in one location? What if they could chat with other fans in real time and buy the shirt from the event? Would they pay for that experience? Some would not, many would, and the content owner would create an additional monetization strategy without incurring much additional cost.

    The answer lies not in recreating a TV experience on the web, that is dead before it ever starts. The answer in this space lies in creating experiences not available in a “sit back” environment and monetizing those experiences.

    And Mark, as a MMA fan, I am someone who not only would pay for that type of experience, but I can help create it.

    Comment by Bill Railey -

  17. Pingback: El Futuro de la TV at Rodrigo Peñalba

  18. You make some interesting points, Mark. I worked for Joost back when they were starting up (actually started the week they branded themselves Joost from the start up The Venice Project), and bandwidth issues were definitely a factor as to why they waited so long to start their online TV service. For exactly the reason you’ve talked about, they realized that the P2P model was their best bet – a cost-effective way to get high quality video online, and which could very well allow for massive audiences (bringing revenue to content producers). Maybe the demand just isn’t there yet? Commercial TV was available broadly in the 30s right? It took 50 years to bring cable television into homes. My guess is it will not take 50 years to find a way to bring TV into your computer.

    Comment by snapstreamenterprise -

  19. Pingback: Mike’s Points - Why Mike’s Points? I’m providing “points,” comments & links about PR, journalism, social media, branding, marketing & other items of interest. » It’s connecting that really matters

  20. –Live vs. On Demand–
    When flickr first launched it was first a live photo sharing experience. They scrapped the site and went to hosting photos because internet users didn’t respond to live. It required to much attention from users, something the internet has a hard time grasping when there are billions of sites online instead of your limited +/- 20 HD channels on live TV.

    –Live Fails–
    1.3M streams concurrently on CNN at inauguration.

    I personally was not able to view the inauguration due to maximum amount of viewers, so you had to figure they maxed out around this number.

    –Assuming Live Doesn’t Scale–
    What could I want to watch consistently live online? Only things I really enjoy live are Sporting Events. Maybe CNN news when I am at the office watching a plane land in the hudson or Inauguration.

    I don’t watch live on the internet that often.

    –Who Is a Content Creator–
    Fact is for most content creators that are finding salvation are making a name don’t do live as main outlet. They can get millions of views on unique content that has virability. These are the people I work with everyday at Viddler and they don’t have the bandwidth for live and not literally. Their bandwidth is energy/production restricted much more then physical CDN restricted.

    Content creators don’t like live.

    You say, “Internet Video. Its the salvation for content creators everywhere. Its the end to dependence on the big bad meanies, the cable and satellite companies. Right ? Hell no.”

    From what I see at Viddler, on-demand is the internet, and internet fully supports that for content creators that are highly profitable….. salvation.

    Rob Sandie (CEO – Viddler)

    Comment by Rob Sandie -

  21. Mr. Cuban – I think that you present a very well thought out argument
    on the situation with ITV. I imagine that you have identified the other
    reasons that make this an interesting nut to crack. I do think that
    there is a way to deliver content to anyone seeking it at any time with
    the combined results of traditional tv. There are a few folks making
    a good shot at it, but they’re missing a couple of pieces to make it

    Comment by Pam Melton -

  22. Few things bring me more entertainment than reading MC’s regular beatdowns of the internet video lemmings.

    “According to Deloitte’s State of the Media Democracy survey, three-quarters of Millennials (ages 14 to 25) view the computer as more of an entertainment device than their television.”

    That’s what they SAY, not what they DO. Young people spend multiples more time watching TV than they do online, and multiples more watching TV versus watching internet video.

    “Millenials are spending one-third less time watching their television than are other generations.”

    That’s nothing new. TV has skewed old for decades. People watch more as they get older, more domesticated, and more passive. This has not changed in the internet era.

    Demand for internet video is miniscule. The average American watches internet video for five minutes a day, up from three minutes last year, according to comScore figures out today. The average American watches TV for four and a half hours a day.

    As MC wrote, the dominant players of 15 years ago are still the dominant players and will be 15 years from now too. Internet video is the biggest fairy tale in media.

    Comment by Mikey -

  23. As the publisher of TeleMedia Strategy magazine ( in my spare time and a consulting VP for the guys that provide most of the software to the Tier 1 Telecom and Cable companies as my day job, I would love to get permission to reprint your tech blogs in my magazine. These are exactly the conversations we need to be having on behalf of our readers. Is there someone in your organization that handles such requests? Thanks.

    P.S. If it helps my cause, I have Mavs ticks in Section 313.

    from mc> you have my permission. go for it

    Comment by John Lee -


    O snap, I guess there is at least ONE CDN that can serve over half a million concurrent viewers!

    You’re off your rocker brother.

    From MC> Oh Snap, one more poster who has no clue. Highwinds has to take every resource they have in order to serve one event. Oh Snap, isnt that the exact point i was making. Oh snap, isnt that the reason Akamai and Limelight had to put caps on all the non inaugural event customers offering video.

    Comment by djRoME -

  25. I’m with Axion above– let us know when you launch that CDN startup and begin hiring, Mark. Just make it quick– unemployment sucks and I’m burning through severance!

    Comment by Randall Arnold -

  26. No question that you are correct today about live events, but
    if your argument is that we are still reliant on the cable and
    satellite companies to deliver video on a large scale I would
    disagree. Your assuming it has to be live, but why not on-demand.
    Even broadcast tv is moving toward time-shifted, on-demand delivery.
    And there are many options now for distributing on-demand video,
    both free and relatively low-cost commercial (like amazon).

    Comment by Bill H -

  27. Fair comment, which you have made a couple of times.

    Thing is as a user I don’t like the experience cable companies are providing(I assume they are like Virgin Media, I don’t have access to any alternatives).

    I find 80-90% of the TV programming distasteful and depressing.
    The interactive features are slow and buggy. 5-10 sec response time isn’t uncommon, and sometimes you have to reboot the box to reestablish the service.
    The speed of improvement I have seen in interactive cable TV over the past 4 years is pitiful.

    While I agree that the underlying distribution mechanism of internet video is inadequate, I think it will eventually be replaced with a p2p/multicast solution.
    On the other hand I see cable TV companies as defenders of their current position rather than user experience stars.

    Cable TV companies could easily change their strategy and become a key player, but I doubt they have the guts to make the right strategy.

    Comment by Henrik Vendelbo -

  28. Not only does the BBC have iPlayer for on-demand delivery; but now
    it also streams all BBC TV programming live on the Internet
    simultaneously with their terrestrial and satellite broadcast output.

    No idea what they economics are; but they’re clearly not prohibitive
    for major broadcasters.

    Comment by Simon Brocklehurst -

  29. att is basically (almost) doing what you (MC) say can’t be done effectively. they stream live concurrent streams to their subs (about 1M now) over their ADSL2 copper lines. sure it’s limited to 2-3 shows at a time but add in a network DVR and you’re good. bump it up to VDSL2 and you’re really good. the key is that on their network they control points of entry and multi-casting so concurrent views take a small hit. give your CDN peerings with end-user ISPs (they have them i’m sure) and ask them to accept multi-casting and you can do the same thing nationally or globally. that one webcam out of the dorm room can reflect off a server or two just to handle the set ups. the stream goes out multi-cast and needs little on the server or bandwidth side at least centrally. we will see TV as we know it go away. we won’t see cable and satellite go away (well maybe satellite). they will just turn into the big pipe and CDN that they really are but with more alt content on their more open data network.

    From MC> I never said IP video doesnt work. I said video over the internet isnt viable due to scaling and economic issues. And Multicasting is not as simple as that. You have router compatibilities and scalability issues. No chance any subscription TV provider just opens up their network to internet muxed with subscription video. that would be business suicide

    Comment by Stel -

  30. Mark let’s cut out all the bullshit and get right down to what is important to me. I want some of your money.

    Comment by The Audacity of Reality -

  31. Pingback: Links - 31st January 2009 « Curiously Persistent

  32. it can be done, but not in a dorm room

    maybe in a pro football stadium stacked to the brim with servers

    Comment by urtaggs -

  33. Mark, how would your argument hold up in light of the following. “THE internet could soon be made obsolete. The scientists who pioneered it have now built a lightning-fast replacement capable of downloading entire feature films within seconds. At speeds about 10,000 times faster than a typical broadband connection, “the grid” will be able to send the entire Rolling Stones back catalogue from Britain to Japan in less than two seconds.”

    From MC> until you can tell me that its going to happen in a switched wireless environment, no one is spending money to dig up your yard to put in the fiber it would take other than Verizon . And for them to support that speed for all websites they would have to host all websites. not gonna happen
    i think you have confusd the tests with internet 2

    Comment by Polo -

  34. Some additional things to consider:
    1. Move Networks
    2. AT&T’s new CDN (they own the pipes…)
    3. Level3 debt load (Mark- buy Vyvx…)
    4. ESI deployment complexity (Akamai)
    5. LLNW stock price/cash flow
    6. Silverlight
    7. FMS3.5
    8. IIS 8.0
    9. Problems with H.264 at high bitrates

    The next time I have the opportunity to redirect my career, I will not fail to gain a audience with you Mark.

    Comment by Fan de Cubano -

  35. Pingback: Will YouTube’s Closed System Kill Video Ad Tech Vendors? | Genuine Forex Trading

  36. “From MC> P2P supporters have been saying the same thing for years. And they will be saying the same thing for years to come. P2P will always be 2nd class because it depends on stealing bandwidth from consumers who have all you can eat accounts. That wont work and its even deader as more people drop landline broadband and go to wireless which really needs bandwidth consumption limits in all near term incarnations.”

    I realize that P2P let you down back in the day, but you really need to get your understanding of P2P up to date:

    1. ISPs are now working with P2P companies and are improving the quality of P2P delivery while lowering ISP costs (aka P4P). This is a game changer for P2P. Done right, it is no longer stealing consumer bandwidth.

    2. The capacity limitations of wireless (along with latency and costs) will prevent its widespread adoption except for business use. Modern P2P takes into account those users that cannot participate for whatever reason and those users rely solely on the CDN anyway without any action required on their part.

    Comment by SilentP -

  37. Internet streaming 720p movies(compressed severely) played via a computer that is connected to an an HDTV of 42″ and above, via even a lousy s-video cord, look spectacular. TV via the internet is the future.


    Comment by Chris Caffee -

  38. Mark Cuban’s right…
    Legacy cable infrastructure is built-out amongst many, many regional carriers across the nation. Each has their own networks of headend nodes akin to CDN edge servers. There is essentially local cached content fronted by software in the headend along with dedicated back-haul trunck networks. These infrastructure projects took decades to build-out (and with buckets of public’s money from then regulated government utilities) and it’s still ongoing, now with VRZN FiOS and others trying to upgrade last mile to fiber.

    For wide-cast ITV, you would need essentially need a similar and ever expanding multi-level tree, branches and nodes (and $$$, time, etc.) of repeaters to rebroadcast over & over & over. Some media server such as Helix, Adobe FMS, etc. have origin, edge and repeater configurations but imagine the size of such deployment…it would never feasibility reach same scale if cable/sat.

    Stick with cable/sat for mass audience reach and internet/ITV for niche content, Also keep an eye on Google. They just started doing ads on TV, so I would bet that other areas such as commercial and UGC (i.e. YouTube) videos might be next.

    Comment by MarkB -

  39. Mark, we’d definitely be all over putting our shows on broadcast or cable if someone wants. If you see anything you like that we’re producing, let me know!

    jim louderback (CEO Revision3)

    Comment by Jim Louderback -

  40. I agree with New Here above, I would never want to replace my tv. My
    TV time is the most important part of my day. The videos are great
    on the internet but watching a movie on my computer is nothing like
    watching one on a big flat screen tv on the couch.

    Comment by JAZD Tech -

  41. You are comparing apples to oranges here. Internet video is about snack sized video viewing. It’s not the same as watching programs on TV, Cable, whatever.

    People use their computers to watch video because it is convenient and because they have the option to if they desire. Not because they need a replacement for the “TV.”

    So in a way, to what your point might be, TV/Sat/Cab will never be replaced by the Internet because it’s not the same thing. The net is a communications tool – the TV is not.

    Comment by New Here -

  42. According to Deloitte’s State of the Media Democracy survey, three-quarters of Millennials (ages 14 to 25) view the computer as more of an entertainment device than their television.

    • Millenials are spending one-third less time watching their television than are other generations.

    Comment by New Here -

  43. Nice crowd sourcing of ideas. I wish I was famous and had a blog with a tone of commentors so i could pose a problem/business idea and get answers and ideas basically handed over to me for nothing but the time it takes to read them all.

    I’m not saying anything is wrong with it. It’s an excellent idea that I am jealous I’m not not capable of.

    Comment by New Here -

  44. Very valid argument for today, and here’s what may happen…

    I go from a $100/month 170 channel satellite package down to a $40/month basic
    satellite package which gives me access to all live content that
    matters (news feeds, major networks, sports events) while I move to
    internet delivery of all the ‘long-tail’ content such as special interest
    TV shows, biographies, movies on demand, etc…

    That way I get my low cost, reliable delivery of LIVE events and
    get my long-tail access for a fraction of the cost I pay my satellite
    providers…isn’t THAT the real threat to cableco’s and telco’s – skimming
    the profit off the top end?

    Comment by Jeff -

  45. Mark:

    In your previous post, if you kept on point or read my previous post, you wouldn’t have ended with a non sensical barbershop example. You know nothing about what we are doing and you missed my profitability statement in my previous post.

    In your response to my pervious post you also missed the part about the advertising being more effective at lower volumes AND you missed in my post before that that our “barbershop” maakes $5M in EBITDA PER STATION PER YEAR. You also missed my statement that we easily solved the bandwidth issue because NOW it is uber-easy to become a very large scale CDN as the first step of network building.

    Furthermore, in 1900 when the car industry was starting, rubber was very scare and very high priced when you needed to purchase it. First rule of economics … markets evolve to fulfill demand. If you transpose your “lie” post to date to 1900 you would be stating there is no way that car industry or automobiles would succeed which is preposterous. Internet distribution of PGC will evolve past Hollywood distribution companies or big sports events and the first beachhead will be in LPGC or local professionally generated content.

    With stupid bold titles such as this post all you are doing is making it harder for a new crop of innovators with the investment community. We have something that represents a new value proposition for both advertisers and consumers. I would like to be the guy that changes you stance on this but I doubt that you would come up to K&L Gates under an NDA for a demo and that even if you did you would probably just say we are full of crap instead of viewing our 8 years of tech and busdev work as the foundation for the next great American company.

    In our mind it is about replacing conventional broadcast and cable TV with Internet TV.

    In the next 3 months if we succeed with the $40M in advertising pitches I would be interested in you as a board member if you want to be part of the next generation of communications.


    From MC> sell 40mm in advertising and I will be happy to talk to you.

    Comment by Chris Caffee -

  46. Pingback: Interesting readings - January 29, 2009 | Sustainable Online Media

  47. Eventually the core infrastructure will be upgraded to support broadcast IP but in the meantime there’s a lot more that software can do to help minimize the issues. See Microsofts “Smooth Streaming” and Adobes “Dynamic Streaming” that use P2P and edge caching in the clients to allow for lot more than 1M simultaneous streams. For live events they simply take advantage of the fact that most of the needed content is already “nearby” rather than pulling it from the CDN let alone the originating servers.

    Obviously this isn’t as efficient as traditional broadcast infrastructure that is optimized for the purpose (and therefore cheaper) but it’s easier to set up, open to anyone and it allows for a vast array of interactive services that you can’t do with a fixed pipe, 1-way medium like cable or sat. Perhaps most importantly you get a lot of client data that is incredibly valuable for advertising that you just can’t get any other way since it needs a 2-way connection.

    Basically we’re in the dark ages of internet TV and there’s a long way to go. You should have more faith!

    Comment by Joe Nobody -

  48. Jared, at Revision3, we have streamed Diggnation live, and we did a live version of Wine Library TV with Gary Vaynerchuk back when we carried his program. And yes, there’s a core group of people who want to see that stuff live, but it’s really all about on-demand. The audience isn’t interested in live anymore – at least not for much of the video being consumed. Even when we do stuff live, less than 10% of the eventual audience clicks in – and it’s usually closer to 2%. Live has other benefits, including delivering more dynamic performances, but that’s another issue…

    With that said, Mark’s right. Internet video is more about serving a narrow audience deeply, mostly when it wants, rather than producing broadly appealing video. It’s the difference between watching the Superbowl, or the Mavs live, vs. catching curling on-demand.

    The problem with most internet video is it forgets the “internet” part. The path to success is not trying to recreate broadcast and cable – that’s just following the cowpath.

    Instead, Build a great product that appeals to a narrower subset of high-value viewers, and then sell that subset to brands that value deeper engagement with that group – as opposed to a 30 second drive-by shooting. Embrace the internet-ness of the media – you can reach anyone, anywhere in the world, right now, but not everyone simultaneously even in a mid-sized city.

    Live on the internet makes sense when you can help a core group of people be part of something that they couldn’t otherwise attend. Not streaming the Super Bowl, but streaming, say, the “Bil Conference” – like we’re doing the weekend of Feb 7 – or the top products and themes from the Consumer Electronics Show, which we did 8 times in early January.

    Next question: is that a business? We think so, and revenue grew 300% last year, so others seem to as well.

    Jim Louderback, CEO Revision3

    From MC> Rev3 does a good job. Its not about replacing TV. Its about producing low cost video and developing enough of an audience to make that show profitable. If its not profitable, you cut it and move on to the next show. Hopefully you can aggregate enough shows that work to more than cover the costs of those that dont. That said, if a show works well enough and the opportunity is there to place it, using the existing production cost structure, on a traditional tv network and expose the product to the audience and potentially generate volume network cable CPMs, Im sure you guys would be all over it.

    Comment by Jim Louderback -


    TVU uses a new broadcast technology called Real-Time Packet Replication. With this technology, all the viewers who are watching a channel at the same time are cooperating to give everyone the best possible signal.

    Unlike typical Internet video broadcasts that use the older streaming technology, TVU’s broadcasters do not need to dedicate more bandwidth for each additional user. This allows broadcasters to reach very large numbers of viewers at low cost. Because TVU can have an unlimited number of channels, TVU offers broadcasters opportunities denied to them by cable and satellite systems. Local broadcasters become global broadcasters; new channels can find a broadcast slot; and bigger broadcasters can create new channels to showcase content that they own but don’t have space to broadcast on their existing channels.

    Comment by Adam -

  50. You say these gatekeepers are few, I say that leaves room for competition. I await your offer for employment as a consultant in your new CDN startup.

    Comment by Axion -

  51. Mark,

    While Internet video may be a joke compared to traditional broadcasts. What about internet radio? Mainly talk radio?
    I enjoy a Liberal Radio network that is online. They offer content and views sadly missing from Terestrial and Sirius/XM.

    Can Internet radio be a major player in your view?

    From MC> For music, with the royalties, its pretty tough if not impossible. For talk, the challenge is the cost of building an audience. Audio delivery is cheap enough it shouldnt be a problem. BUt like sat and over the air radio, the cost is in building and audience and selling ads

    Comment by Spike Rogan -

  52. Fascinating article, Mark, and thanks for informing us of a bottleneck I did not know existed. Food for thought if nothing else.

    And I was going to kid you about your allergy to apostrophes, but I see that at least one commenter included extra ones so maybe it all evens out. 😉

    Oh, and on a completely unrelated note: I was at AAC for that Mavs-TWolves game where your guys had the biggest comeback for a win in their history. MAN what a great game! More of those please!

    Of course, now that I’m out of work I’ll be watching them at home… until the FIOS has to go.

    Comment by Randall Arnold -

  53. The BBC iPlayer had approximately five million page views PER DAY as of 25 June 2008.[69] As of 9 December 2008. On the day of the Obama inauguration 1.5 million visitors to the BBC website accessed video (or audio) on the BBC News site alone.

    Comment by GS -

  54. I do not deny your write up with regards to CDN duopoly / tripoli, in an article posted by Dan Rayburn @ StreamingMedia that Akamai decided to cap on and charges publishers at their discretion for Obama Inauguration live streaming.

    In fact, the big boys do not seem to have the enough capacity to serve more in any case of serving ad-hoc live events, it’s a turn off for any small video provider trying hard to keep the company afloat.

    Therefore I’ve always thought of a open CDN market, whereby webhosting companies and small regional ISPs may join forces and run a CDN collaboratively and collectively. There are simply too many webhosting companies out there running on excess capacity and making use of every extra bit of its resources would squeeze more juices out of it.

    Publishers (video provider) may choose from multiple CDN players instead of stuck with the few biy boys, opening up the market and options would be the way to go for CDN. Does it fit your appetite now? 😉

    Comment by Whei Wong -

  55. Mark –

    As has been pointed out in the comments above, the
    restrictions on using Internet as a delivery mechanism are purely
    due to scarcity of bandwidth. Not too long ago you posted how
    broadband scarcity was a farce and we had plenty available if you
    simply took it back from the cable operators.
    It feels as though this post contradicts that. Could you please
    comment on how you would reconcile the two posts?

    From M C> Its not a contradiction at all. There are gigabits of bandwidth allocated to traditional cable tv and satellite. Compared to what you get at home. Thats not going to change, although there is an opportunity for content creators to buy or lease that bandwidth. There are also still analog channels that eat up 38.6mbs that can be allocated to any number of things. That doesnt change that there are multiple hurdles for delivering video on the net. None of which will solve the problem that its an inefficient solution to deliver to all but the smallest audiences.

    Comment by Lee Milstein -

  56. I understand Mark’s point. And yes, you can’t beat distribution on cable or satellite rigth now. The problem is if you are a new programmer, you can’t get distribution on these channels. The cost of starting a new cable network exceeds +100 million and emerging networks are pushed to digital pay packages where they theoritically reach, lets say, 100 homes, but only 2 homes are paying for the service. Viewers are getting overwhelmed with subscription fatigue and not buying these services. So from their hay day 30 years ago, fewer networks are launching and reaching scale in the cable world as well. Then you look at rating, and with the increased fragmentation of the cable market, shows are drawing smaller audiences and therefore repeats, reality TV, and lower budgeted shows are dominating lineups. It’s like the newspaper business, you keep cutting quality, which leads to smaller audiences, which forces the quality to drop some more, and the circle continues.

    The Internet is opening doors for smaller players who can not afford to pay the gatekeepers in the cable market. If you would have told me 20 years ago that we would watch TV on the Internet, I would have given you Mark’s argument. But today, as someone who is nearing 50 years of age, I watch TV on my computer all the time. Why? Because there is nothing on TV.

    Comment by ChrisM -

  57. Wow.
    And to complicate matters, if CDN *could* carry what [TV] could carry,
    that much streaming video on independent servers, that much streaming
    video on my wifi…

    …no one would ever get to see anything.

    Comment by Ann Hartter -

  58. When I took Theater Appreciation in college, we were taught that the only people making money in the theater business were the theater owners. So, I thought, I should bail on acting or directing, get an MBA and buy a theater. Why could any one of us not decide to do the same now, and create a CDN company? That seems, from reading this aricle, to be where’s it at.

    Comment by Clark Bunch -

  59. Mark,
    A part of me hopes that this post is you venting over some sort of frustration over trying to find an acceptable medium for some HDnet content. Mojo was one of my favorite channels, and my DVR has seemed empty this year. I was sorry to see the channel go, and wish you luck in coming up with a viable solution to whatever the problems you encountered that led to its demise. How Mojo was unable to continue, while Al Gore’s CurrentTV continues to take up airspace, is a question that has been bothering me recently. In college I worked for a brand representation company, and when I was asked to represent Current, I could simply not imagine associating myself with such a poor product.
    Assuming YouTube has developed the capability to provide sufficient capability to solve this problem, it is a bit unsettling that they are unwilling to forgo their branding in favor of the possibility of working with customers to find a solution to this need. Many of the industry leading consumer products manufacturers have given up their corporate branding to create “store branded” products which can be sold at lower prices. They were able to increase production and distribution, while not having to focus on marketing. YouTube may soon find that it has reached its high water mark. While President Obama and some small content producers may be content with simply reaching the maximum viewers possible, more mature content will clearly need a better form of distribution that recognizes that the content is more valuable than the medium.
    Viacom’s recent push to gain Time Warner subscriber revenue relative to their overall market share looks like the first battle in what may become an all out war. While their tactics were somewhat ethically questionable, their point was clear. Content delivery has been sadly dominated by a few companies who enjoy the benefits of limited competition. Companies like Comcast and Time Warner rarely face pressure to change their practices, and are able to effectively impede competition and innovation by throwing their weight around Washington.
    I recognize your point about the limited capability of internet broadcasting in its current state. However, the internet as a delivery medium seems to be a much freer market than the traditional media industry. More importantly, the forces that impede progress in the traditional media market do not have the same clout when it comes to the internet. The solution must include collaboration and compromise. There is no reason for something as revolutionary and open as the internet to fall victim to over-regulation and practices designed to stifle competition and innovation. Perhaps this is the time for a media mogul such as you to recognize a market which is ripe with inefficiencies and which fails to fully meet consumer needs. These CDNs currently enjoy some of these same advantages, but the barriers to competition and innovation IMO are significantly less to the point where a focused, nimble, company may be able to effectively shake up the market.

    From MC> If the barriers to competition and innovation are less, where is that focused nimble company ? In this world, and until someone we dontk now about comes along, you play by the rules of the CDN or Youtube. And speaking of monoliths, is there a bigger one for video than Youtube/Google ? People make it sound like Youtube is some democratic , non profit that is there purely for the benefit of those that provide video.. I dont think so. There is far more power concentrated in Youtube than Comcast has ever had

    Comment by SteveK -

  60. Mark:

    I respectfully disagree with your response above. The issue is not simultaneous streams.

    It is low volume verses high volume distribution models for advertising effectiveness and the comparable investment return. The movie content is secondary.

    Clearly the internet doesnot work for pro sports, but there are many other applications where lower volume viewership builds higher customer engagement and potentially better sales comparables for advertisers. We are pitching these concepts right now to major brands that are receptive to this approach. High volume broadcast and advertising models are based on broadcast technology and with the over staturation of advertisers positioning fast paced advertising, it has become marginally effective to the point where you can lump this approaach into the entire failed economy. Newer technology will drive better brand engagement with lower viewership but higher regional sales for advertisers all via the internet. We are finding corporate C level people very interested in our approach as a differtiator.


    Chris Caffee

    From MC> No question, you are right, lower volume brings higher quality engagement. Just like at the barbershop. The problem is that unless your viewer can generate lots of revenue for you, you starve. At least barbers know enough to charge per haircut so that they can pay their bills.

    Comment by Chris Caffee -

  61. Wow, the commenters are sure riled.

    Mark is right, the system for internet distribution of video cannot handle much more capacity right now.

    I work for a large media company and help run both our traditional broadcasts via MSOs and I also help with our internet distribution of both live streams and on-demand viewing.
    No CDN stood up to the Inauguration. They all had problems. The infrastructure simply is not there yet.

    Even with on-demand episodic viewing, the main viewing tends to occur in the day and week after the airing on television, reaching significant concurrent sessions.

    Commenters tend to quote numbers of viewing (50k a day! Omg!11!) when that really means nothing without other key bit of information. You need to be able to discuss total video package size and minimum bandwidth levels (to avoid buffering). You need to understand how the cable and telcos are actually architecting their networks. If your central office or cable headend only has a DS3 back to the their main network and the net, then everyone they service is limited to the ~45Mb/s bandwidth. It doesn’t take much streaming to start eating that up. Toss in a bit of torrent users and you get some serious resource allocation problems.

    It’s a serious issue that will require a serious solution that costs real money and time to build. It will improve incrementally over time, but any that is declaring old media dead and internet video no ready for prime time is fooling themselves. Look, I like running Boxxee on my Apple TV as much as the next guy, but watching programming off a DVR fed by a satellite or cable company works and works well. The new DirecTV boxes are now network aware and can do on-demand viewing of content and will even find and play off a DLNA server in your home network. ‘Old media’ is not stupid and more importantly, they have the cash flow to make investments on improvements that ‘new media’ doesn’t tend to have.

    From MC> Ding, ding, ding ! Now this is the man that content companies should hire to help them. Finally someone who knows what is going on , good job michael !

    Comment by Michael -

  62. The technology is just around the corner, and it won’t require a CDN:

    Comment by Stefan Richter -

  63. The issue is one of displacement not replacement. Cable and satellite
    channels will remain the principal platforms for time-sensive
    mass consumption content while the web will intermediate more
    vertical, more narrowly focused content. The experiments with live
    concerts streamed on the web are simply that – experiments..though
    that something can be done doesn’t mean it should be done. Given
    that conventional television has 99.9% penetration of homes in North
    America, is ‘always on’ and with PVR access can provide the same
    benefit as downloading, it’s hard to imagine why anyone would want to
    replace it with internet streaming and its inherent imperfections.
    Similarly, that one could watch the Super Bowl on a 3G device doesn’t
    mean that most people would want to.. I think we are coming to the
    point where the appropriate device will be selected for media
    consumption depending on the time, place and circumstances and despite
    the wishes of the technorati, it will be the consumer who makes
    the choice.

    Comment by toddfg -

  64. hey mark—good post–but how could you forget about the top cdn of the early 2000’s—ibeam. your remarks brought back interesting memories of those days and that is how we initially all those infamous shows…where you always had your radar up on what we were doing. i appreciate the cable/satellite comparision as having worked for turner brdctsg for 10 yrs bf getting into the cdn biz provided a good perspective. i remember handling sales for the cable/broadcast sector for ibeam and how excited i got when i finally landed the mtvi (interactive) account with nicholas butterworth, damian manning, and genna borisov at the time. cdn’s are in fact bandwidth brokers or resellers, as our revenues topped out at $325,000/mth…handling there mainly audio streams. i forgot some of the details, but think they were paying us @ 1.2 cents/megabtye transferred (vs peak utilization) shortly thereafter, those prices came tumbling down as the model wasnt wkg for them and they knew it. it think we ended up getting about 1/10 of those revenues bc it was a commodity and others cdns that you referred to started coming in and offering lower prices.

    from what ive been keeping up with, there is still no model for scale within the cdn mkt and i think it is a tough road. ibeam got acquired in really an asset sell by wiltel, which merely meant that we survived for a yr or 2 longer bf being shut down….so you are still correct on making the economics work….come to memphis!!

    Comment by gwin scott -

  65. Cuban,

    Your logic is wayyy off, especially when you contradict yourself in your many responses on the threads here shooting down your flawed logic;

    You put in bold your original post that nobody can serve 1mm + streams and then you admit that LL and Akamai can both do it on a comment. What are you smoking?

    You’re the one that’s high brother, not us.

    You’re also forgetting one MAJOR flaw with all of this. Cable and Satellite systems cannot track with any degree of certainty who is watching, and don’t even start me on the nielson ratings..

    The internet is a wonderful place with two things going for it that cable and satellite don’t have. 1: ACCURATE analytics. You can see who watched, where they live, what computer and browser they have, the times they tune in / tune out, you can track when they pause or rewind internet video streams, and much more, and this isn’t even including the invaluable interaction with the user that takes place on the internet, like MySpace’s most recent broadcasts of the Presidential debates that featured live time viewer polls based on questions that were just asked and answered over the stream.

    Features like this make internet streaming not even CLOSE to what TV of any kind offers, and it’s old rich guys like you that don’t get it. New Media isn’t about just watching anymore brother, it’s about interaction.

    You’re STILL way off. Way.

    PS YouTube DOES support a B2B model with a rev share model, so again you were wrong when you said they don’t support any commercial purposes. You’re right when you say they don’t remove their branding, but that’s the ONLY part that you’re right. If you want your own branding you’re going to have to pay to develop the serving infastructure and the web infastructure to Manage Your content. In exchange for these free services, youtube gets to stick their watermark on your content while (and be srue to read this next part because you might miss it AGAIN) PAYING THE CDN BILL FOR YOU. That’s right, when your Youtube video goes viral and gets 1,000,000 hits, who pays that bill? Certainly not the end user, it’s YOUTUBE.

    You got it all wrong bro.

    From MC> You need to read something besides what you write. I said no cdn can do more than 1 mm plus event. As far as tracking, TNS and Rentrak both have second by second tracking for TV , DVR and VOD. Youtube does license content, but they control the display and presentation. You play by their rules. If you want to be a Youtube video rather than your own branding, thats your call. Nothing different than what I posted.

    And as far as your accurate analytics, the privacy people might have some issues with anyone knowing who is watching and all those people behind corporate firewalls have a say as well. You sure you have a clue about any of this ?? Stay focused , get that high school diploma and all will turn out ok..Oh and ask them to start classes on Tru2Way, might help you get into a good school if you graduate

    Comment by djRoME -

  66. Mark –
    While I would largely agree, I don’t think the point of the majority
    of web video use will be live streaming. In fact, there is a marked trend
    toward timeshifting viewing on the TV/Cable platform. The majority of
    web video is, effectively, on demand. Thus, one can achieve cable/broadcast
    audience on a web video piece quite easily. I would agree with a prior
    post that the live streaming will be a small or mid-size audience target,
    rather than a broadcast target for most, non Inauguration/Oscars/SuperBowl
    programming. When we get those rare mass market events, media supplement
    each other, rather than replace one another.

    Comment by Deborah Brozina -

  67. Pingback: Running Naked: Mark Cuban Just “Pulled Down the Shorts” of CDNs Everywhere « - Business & Technology Blog

  68. I hope HDNet shifts to the Internet delivery because more and more folks are turning there for their entertainment.

    Comment by The Internet on your TV -

  69. Mark:

    All valid points…QoS and the limitations of IP infrastructure affects all packet delivery even hosted applications delivered over the Internet….but it continues to improve. You can’t let the limits hold back the innovation. If you had applied that govenor, we might not have been able to do the following with your first start-up…

    I wax nostalgic for the good old days when we were too dumb to know any better!

    Great Blog, I read it daily.


    Comment by Jim Safran -

  70. Hi All,

    I’ve been streaming interactive video for over ten years now. Granted it’s not live streaming. Most viewers of online video (i.e. “The YouTubers”) really don’t care whether video content is live.

    My episodic TV shows at minimum reach 60-100k viewers per show online and mobile (Iphones). I’m not using a CDN to distribute the shows over the net. I have my own server and it’s been working find. I’ve also master the compression of the video/audio. Average file size is 120megs for a 30min 480×286 show. Don’t ask, Won’t Tell You! You can see a recent episode of my jazz show at

    Bottom line. Don’t stream live! “Individual On-demand”. …seeing TV when and where you want to see TV. This should be the direction of online video.

    Paul Revere has left the building.

    Comment by George -

  71. Mark – what are you trying accomplish?

    You have the ability to become the Ted Turner (CNN), Al Neuharth (USA TODAY), ESPN, and more! The way to do it, Mark, is to not be limited to what you are only seeing that’s out there.

    How did those guys do it? It certainly was a first.

    I envision a new global communication. You in?

    I don’t know about you, but I want my intelligence INSTANTLY and with me at all times. That’s why they say that we take our mobile phones to bed with us.

    OK, that’s a great thing.

    Now, if even if we have to create the device – that’s PERFECT – and maybe even launch our own satellite(s).

    I truly believe that everyone is ready for instant intelligence (eventually provided through something we wear – it does surprise me that we keep carrying these things), that provides everything. Yes, everything.

    I”m going to make it real.

    This is 2009, Mark, that’s going to become 2020 in seconds and becoming The Global Media Empire is simply a matter of a decision.

    It’s a first step. Let us all engage – instantly . . .

    Comment by Ian O'Maly -

  72. Pingback: » Is YouTube a Closed System That Will Kill Video Ad Tech Vendors?

  73. My HD is connected to an HD dvr (Cox) and a pc connected vi svideo.
    My only disagreement is the idea that “live” will remain a big deal.
    The closest to live we get is when we wait 20 minutes after the
    start of our Thunder games so we can skip the commercials. I think
    it will be a while before some form of localized bittorrent will
    have more bandwidth than the Cable Tv. 50 down/5 up Docsis is just
    a start. If speeds start doubling every year or so things could

    From MC> The issue isnt live, its simultaneous streams. If enough people wait for the game like you do, you have the exact same problem. And how many people are willing to wait a couple days for things they want now ?

    Comment by AllenC -

  74. Sounds like no matter what content distribution channel you use,
    it’s the content ITSELF that is key! Storytelling is still the
    most important element, because you could have the fastest, most
    dynamic system out there, and if the story stinks … you’ll never
    use it again.

    We’re always looking for ways to get our stuff out there — thanks
    for a great, thought-provoking post!

    Katy Widrick
    Executive Producer,

    Comment by Katy -

  75. Same world, different platform.

    Comment by patricia -

  76. Pingback: On Mark Cuban’s “Video Lie” « NewTeeVee

  77. Good post and very true. Comments and thoughts:
    – Using the Internet to boradcast concerts doesn’t make much sense. Cable and Satellite are broadcast media by nature and are much better suited for this type of thing than the Interent.
    – Broadcasting 24/7 on the Internet doesn’t make much sense either. That’s the old linear-TV model. The Internet is much better for on-demand, where you stream a piece of content only when someone wishes to watch it. This means heavy traffic during peak watching times and nearly no traffic during off-peak.
    – Having the Internet compete on large-scale std-def video distribution with cable/sat requires a MAJOR overhaul of Internet infrastructure, especially at telco/cable metro/access networks. Having those multicast-enabled and accepting multicast from the public Internet be a huge step in the right direction. Too bad there’s no business model to justify doing that.

    Comment by Itamar Gilad -

  78. Pingback: The Great Internet Video Lie « Technologic

  79. Great post. I guess this is the reason why so many big internet shows are
    recorded. Like diggnation, wine library tv, they never do live shows and
    I always wondered why.

    Comment by Jared O'Toole -

  80. Pingback: Finance Geek » Mark Cuban: Why Web Video Will Never Beat TV

  81. Pingback: The Great Internet Video Lie [Voices] | LinkTouch

  82. Mark:

    You got this totally wrong.

    You definately see the past clearer than the future. We got it pegged … 20 USA internet TV stations, each makes $5M per year end year 3. 50 employees @ $25/hour. 100,000 viewer peaks per station. 12 national advertisers with completed marketing materials. We’ve solved monetization of movies via technology and have overcome the COGS bandwidth issues easily.

    You are using past models as reference for the future which does not work. Sorry you can’t see it and are still looking to $10 tickets at theaters with Christie projectors as the future. That is the present.


    Comment by Chris Caffee -

  83. I hate to throw out the Long Tail argument, but it’s very possible that regardless of whether the capacity exists, people will not need to distribute to such large audiences since that large of an audience would rarely be interested in a single subject.

    I would also argue that given the infrastructure costs required for a traditional TV station/cable or sat provider to reach that many people, the internet is a much more accessible and ultimately cheaper medium. The cost to distribute to that many people via broadcast is getting cheaper at a much slower rate than the internet.

    But you are right that today, it’s quite expensive. If you need to stream to that many people, going to people who upsell other people’s bandwidth might not be such a good idea. 🙂

    Comment by Daniel Rhodes -

  84. While traditional broadcast is still far more efficeint in delivery
    quality, the internet has shifted habits. DVRs have shifted habits.
    Networks can no longer sandwich crap between other mediocre
    programming. They must compete! Bad for the fat cats of the old school
    programming, good for consumers.

    Local broadcasters should be pumping more into live/local coverage, it
    will be their only survival.

    From MC> Have you noticed that the fat cats of 15 years ago are still the fat cats of today ? That the poor artisans of 15 years ago are still the poor artisans and content creators of today ? What does that tell you ??

    Comment by Marc K -

  85. Even at the micro end of the scale, video streaming is demanding and expensive. For small companies with very limited hosting services, just a couple of hundred people watching a 10mb video can munch through their quota. I always recommend that these companies use a third party hosting system like YouTube: after all the branding won’t kill them, and at least it won’t bring their site down. Bigger companies can obviously go for a more commercial set up.

    But when you think about what video is, the frame rates and the increasing viewer demand for better quality and higher definition, it’s no wonder it can’t compete with conventional cable and satellite companies. Internet users in countries with download quotas, like here in Australia, are obviously more aware of this. But when you have unlimited downloads you don’t have to think about the size, you can just consume. You can leave it streaming while you sleep. And this is obviously very demanding and costly for the content provider.

    Comment by istara -

  86. There is a lot of progress in the field of live P2P video delivery. P2P will offer a scalable, low cost option for broadcasters, and it’s adoption in video streaming will likely cause a revaluation on how we are billed in the US.

    Copenhagen-based Octoshape, for example, successfully streamed the Olympics
    Buffering: non existing
    Startup-time: < 1 second

    They have been successful in Asia and Europe and recently streamed CNN’s coverage of Obama

    Further, there are efforts to bring similar P2P applications, like Tribler, directly to the TV through IP STBS that will be able to more readily compete with Sat/Cable services.

    From MC> P2P supporters have been saying the same thing for years. And they will be saying the same thing for years to come. P2P will always be 2nd class because it depends on stealing bandwidth from consumers who have all you can eat accounts. That wont work and its even deader as more people drop landline broadband and go to wireless which really needs bandwidth consumption limits in all near term incarnations.

    Comment by Eddie -

  87. Think I’m with
    @Steve Place

    It’s the legacy media’s job/obsession/reason for being to try to get as many viewers for a single piece of content at time of release – say one million hits, – rather than, say, 400 hundred pieces of content each getting 2500 hits.

    And that’s before you even consider an editorial ability to build a community around that stream, and the ability for the audience to comment, post video responses, make content to contribute to the stream (eg LG15: the show is yours,, or even

    (Surely the strength of internet content is that it isn’t a one way transmission?)

    I think I’m thinking, you don’t necessarily need 10k to view a single piece of content simultaneously in order for something to become huge…?

    Thanks for making me think – great post!

    From MC> What the “size matters” issue is about is making money. Content creators want to get paid. They arent going to get paid much , if anything with 10k viewers consuming over a multiple month period. The shorter the period from start to maximum views, the more money the creator makes. Thats the driver. Hundreds of thousands of fews over a year makes you enough money to buy a box of cereal every week

    Comment by StoryGas -

  88. To a layman like me it seems like you are saying that internet video
    can’t compete with live TV. But couldn’t content providers utilise a
    file sharing system? I know I would prefer that to paying for satalite
    TV here in Australia. I would love if the NBA would provide such a
    service for foreign fans. I would still keep my TV subscription. But
    I would think internet downloads rather than live streaming will eat
    into TV’s share substantially. Most content doesn’t need to be viewed
    simultaneosly does it? If TV is only left with simultaneous viewing
    as a market for content delivery they are in trouble are they not?

    Comment by peter s -

  89. Hi Mark,

    That was a good insight on Net videos. True, there are only a few CDN’s thus it would be extremely difficult to showcase good events over the Web. However, I think that Internet video will gain traction within a few years as tech companies start to develop better hardware to support the growing demand for Net videos.

    Comment by Paul Benavides -

  90. Mark,

    how did – and others as well – handle Obama’s inauguration then?

    Comment by Massimo Moruzzi -

  91. The Big Lie here is yours, Mark. You are assuming that Live Video – ie everyone forced to watch the same thing at once – is the goal.
    This is old media thinking writ large. People pay for products – Tivo, DVRs, iPods, TV series DVDs – that turn streams into files they can watch when they want to.
    The rare cases where millions of people want to watch the same thing at once – Presidential Inaugurations or faux Gladiatorial contests like American Idol, the World Cup Final or the Superbowl are great uses for broadcast TV or satellite, and lousy uses of the net.

    More here:

    From MC> Its not just about live, its about simultaneous streams. There are capacity limits. Akamai shows the number of simultaneous streams they are delivering on their website, across everything, live and on demand. 1mm is a big number. And they are the biggest at it. Whether you want to deliver live, or on demand, if you want to deliver enough video, at decent quality, and make money from it, you are beholden to the 3 or 4 CDNs that control the market. Sure you can go cheap and be on the long tail. Lots of people do. The thing about the long tail, is that people who live on it are ALWAYS STARVING. There are so few people on the net creating large enough audiences to make a living from it that we all name the same few when discussing the opportunity.

    Comment by Kevin Marks -

  92. Nice & interesting post.

    Like said in the comments it is a technology issue and not a structural issue. Amazon is stepping in the CDN market with their Amazon Webservices, but more interestingly P2P isn’t dead yet. It’s still an option, especially with the bigger streams (millions) and the CDN’s know that.
    The best example so far is the inauguration of Obama where a plugin (Octoshape) on the Flash Player was used to stream to 8 million concurrent users. More information can be found at

    The future is bright, but be patient!


    Comment by Joey -

  93. Pingback: The Great Internet Video Lie | Mark Cuban | Voices | AllThingsD

  94. Pingback: Why Mark Cuban’s “Great Internet Video Lie” is Merely a Delusion | Michelle's Blog

  95. OK, except that it almost never happens that a million internet users
    want to watch something live for hours at the same time. It’s a
    very rare, superbowl-scale event that could get that viewership.

    Actual popular internet videos are short and not live.
    The most popular YouTube video has 113M lifetime views and is
    a little under 4 minutes, so a CDN capable of 1M simultaneous
    streams could serve that entire audience in 7 hours.
    That’s plenty for any current popularity model.

    Comment by Trevor Blackwell -

  96. This was very necessary post Mark. I’m a media buyer with just a few years experience and I’ve always gone off a cost per point, cost per thousand and frequency combination. One of my current employers is looking at going strictly web with their newest ad campaign. At least the cost per thousand can be calculated on the net but you’re absolutely right… When I think about it, my client’s in a top-12 market and with the cost to reach everyone they’d want to over the internet, I could go after a sizeable buy on CBS or NBC and blow up their commercial (while hitting people with a high frequency). Someday I think the information super-highway will carry similar clout, but that’s a loooong time coming. Thanks for putting things in perspective.

    Comment by COlax1980 -

  97. Mark –

    I think you missed the point of internet programming…Although some people would like to takon the big evil empire that is broadcast television, most internet programming, particularly live streaming programming is for a more intimate crowd. Which in return is what makes it so inticing. The ability to watch live and now to interact with streams is the draw. The LIVE part is meant for a medium size crowd so the producers of the content can interact. The archived version is meant for mass consumption. By the way why does the comment box cut off so I can’t see what I am typing completely? Look forward to hearing more.



    Comment by Dennis Lankes -

  98. Mark,

    Absolutely. Forget the traditional monopoly Internet connected CDN. Game over.

    I outlined my thoughts at IP Possibilities last year.

    I’m at the (2:45 mark) and the preface for me was that rural operators
    have all the assets to provide programming -now- that far exceeds
    anything that the big boy networks (Comcast, TWC, Verizon, etc.) could
    ever pull off — it’s called local content.

    I’m looking for companies that can support or have thought through providing
    an Akamai like model of “ubernode” close to the subscriber at the POP as
    a managed appliance/service.

    Seriously. If a 15 year old kid can rip and house everything he’s ever rented
    from Blockbuster, there is a legit way to make it happen at the local POP.

    When people conflate IPTV with YouTube I get really really antsy. It is NOT
    the same thing and we don’t want it to be the same thing. The TV format and
    the handheld format are two entirely different things. Just like David Lynch
    said about the iPhone — it’s true.

    I also hope IPTV does not become a consumer used word — and hopefully
    this is contained to the insular groups you outlined.

    Eventually, North American spending patterns will emerge and players
    will have to describe their value to markets in terms of threats and/or

    10G, 40G, and 100G are all within reach before someone has to rethink the
    local access layer. The key is keeping it nimble and forklift ready just
    as datacenters today don’t have time to keep the big iron anymore.

    Yes, I had a bit of coffee before they hit record.

    As I’ve said elsewhere, I still think that the Telcos can launch -now- vs.
    later. Middleware trips them up and integration trips them up but there
    is a much clearer path now than there was years ago when this was
    rocket science when you did

    IPTV as a term has been applied to anything “video” delivered over
    the Internet. Since this could make South Park’s “Spirit of Christmas”
    be argued as IPTV unicast delivery from nearline storage
    (it occupied a lot of file servers in corporate America)…
    it’s just the wrong term finding its way into the common dialog.

    The Joost play is still something Telcos could embrace. They have
    the best “CDN” path to subscribers and always will until something
    comes along to disconnect the copper, fiber, or cable on the side
    of the building.

    The mobile broadband IPTV that takes place on handhelds in
    London makes dozens of channels available — but it is coming
    from a service provider, telco oriented, but when I tried to
    explain this to a group of STB minded folks, they said it
    can’t be called IPTV.


    Comment by Jay Cuthrell -

  99. > Which leads to the question of who can provide a service as simple as 10k streams

    There’s some very cheap solution in China that you can transfer a mms source to a p2p live broadcast network, I tested and it can handle 10k streams. Keywords PPStream, TVAnts, etc.

    Comment by est -

  100. Well now we know computer can never beat Television
    and satellite in this regard.

    I was hoping for another way to get independent media but you
    can’t beat big business.

    I wouldn’t be surprised if the CDNs are controlled by the
    Networks and Satellite. It seems that they have not developed very far
    in the last decade. Seriously it is like they do not want to compete
    with them.

    This reminds me when they got rid of the electronic car because Ford
    and the gas company’s said so.

    The people at the top control the media. If they can shut down
    Aljazeera(newspaper) in Iraq. They can control 3 CDN’s.

    Triple Kings.

    Comment by Triple Kings -

  101. Mark,

    I read every thing you post and have for years. I buy into every post whole-hearted. But not this one. I think you are missing the point. Internet video is not a lie and will come very close to destroying tv distribution as we know it. It won’t be because the internet is a better way to ditribute huge events to millions (I agree with you there) it will be because the internet can distribute billions of events to smaller, happier, audiences who are watching exactly what they want, not what someone at a network says they should be watching. Law & order is on NBC right now, I love Law and Order but this episode sucks so I am on hulu watching a old Law and Order with a cast I remember and like better. (Just a hypothetical example)

    The inaguration was another example. I didn’t want some dumbass network anchoring telling what I should think of what I was seeing, so I bopped around 20 or so different live feeds I found on the internet and found the one I liked. Perfect example of internet video giving all of us more options than Cable/Satellite could ever hope to.

    I still think your a genius, you converted me to be a Mavs fan (u and Dirk) and your still at the top of my feed, I am still throwing a voodoo curse on the SEC for you, but I rrspectfully disagree here.

    From MC> I watch law and order too. I set my DVR to record all the eps, and then i watch the ones i like when i want, on my tv. MSOs can add the same hard drives to their networks that Internet VIdeo Providers can to theirs. And they will.

    Comment by cody -

  102. Mark,

    You and I discussed similar topics a few years back – SBC/ATT IPTV play was part of it. One of the problems I have always had to counter in my field is the reality of BW vx. Storage and the ability to deliver content. I can appreciate your thoughts on this topic and greatly enjoy reading your posts.



    Comment by Wayne -

  103. Everytime I read a mark cuban post where he’s writing with supreme confidence I’m always reminded of the saying ‘Do it once your lucky, do it twice your good.” He ripped off AOL once and thinks he’s a web genius. My grandmother could have hustled AOL for a billion dollars. Make a great web company or shut up.

    Comment by durbin -

  104. use torrents

    Comment by lou -

  105. I have been streaming IP video since around 1988 internal to Stanford to
    monitor experiments around campus. I was an image processing expert
    back then.
    Then built a CDN inside Sun Microsystems global intranet in 1990.
    It did 13000 simultaneous viewers on Scott McNealy’s Christmas
    announcement. (AKA empcompvideo)

    I had my own CDN running in 1994 Sokol & associates /
    doing video streaming and static web content.
    It was later acquired changed to Internet Broadcast Systems
    and eventually sold it off in 1997.
    It was too tainted with adult that going public wasn’t an option.
    The new owners end up sinking it I think, not my fault.

    We owned the adult video business back then.
    Mostly my “Livecam” jpeg push technology. Maybe you’ve hear of it?
    I also had streaming MP3, MPEG2, H.263 back then too.
    We had over 1M video views per day of live content from 20+ channels, mostly
    girls. in 1997.
    We supplyed over 2500 top adult sites with content.
    We doing over 500,000,000 image hits per day too.

    After selling all that off, I went to Korea and created the first
    CCTV DVR. It did 16 Cameras using a custom capture board.
    Then created DVBS to push Livecam in a more serious way, we were
    getting $80K per server! each could support up to 4000 streams per PC
    assuming you could get enough bandwidth to it.
    But this all fell apart in 2000 shortly after we went public on the VSE.
    I coded most of this stuff in C myself.
    Including over 10 video codecs from scratch.

    So don’t punk me like I haven’t been around.

    That said.

    > OTA HD and sufficient broadband speed are the core reasons I was able to cancel my DTV service.

    Here is Santa Barbara there is not 1 single DTV broadcast
    that I can receive. Let alone HD.

    Dial-up. MC are you nuts?
    Oh my god dialup was painful, dopped lines, call waiting,
    creeping downloads.
    > I bet you barely notice the difference in the speed of your
    > basic browsing.

    > Are you high?
    Are you?

    1/10 second vs. 30 second page loads. If I was high maybe I wouldn’t
    >How much bandwidth do you think most regular web browsing and email uses ?

    It’s not about bandwidth but latency on some things.

    I also don’t have a POTs phone as many others are going away
    from POTs too.
    The telco charge $50 a month just for the basic line before any usage.

    > Apply the savings to your cable bill, and you get far more content, a DVR to watch what and when > you want.

    $50 for my cable modem vs. $50 for basic land line phone + $10 for
    the Dial-up ISP.

    savings? where do you see savings?

    So for me it’s cell and skype, and Gmail voice and video conferencing.
    I am talking with China and India almost 4 hrs per day. cost $0.
    I don’t even want to think what the telco’s would charge.
    Or calling cards. Yuck, and no video conferencing either.

    In addition

    > and you get far more content, a DVR to watch what and when you want.

    What kind of DVR are you talking about? Certainly not my Cox cable.
    I have wait till it’s broadcast to record it, assuming I can even
    navigate through the sluggish terrible UI of the Cox channel guide to
    find the program I want.

    It so much nicer to see a whole season of Dr Who, and almost
    1 year before it airs here.

    Tru2Way , ah hum. This look like it’s just RTP/RTSP IP based video streaming
    over Docsis 3 to back end servers in the Cable company.
    There ain’t no free lunches here.

    So it’s not all that different then regular internet streaming.

    And again Commercial P2P has been heavily suppressed. But there
    isn’t any reason it can’t bury CDN’s.

    Also Youtube was over Peer1 in March 2006, see “Exploring YouTube”
    Not sure now.

    It should be possible to build a CDN that is all peering agreements
    and greatly reduce cost, assuming you have sufficient demand,
    which youtube should have more then enough at this point.

    I think peer1 was doing this back then, it was hard to tell from just
    poking around from the outside.

    If so Youtube was getting reamed.


    PS: What’s with this blog entry form.
    I can type off the right end of the window
    and can’t see what I am typing.

    Comment by John L. Sokol -

  106. Come on, Mark. You should know better. Yes, this is the case now, but it will eventually scale up to meet demand! That’s the way things work! Five years ago, *any* kind of video online was expensive and out of the reach for most people, but now look at where we are. YouTube, Vimeo, Skype video, iChat video, Ustream, Mogulus, Tokbox — none of that existed five years ago. The technology scales up to the demand. Five years from now, I highly doubt this will be an issue. There’s too much money to be made and bandwidth will only get cheaper. Someone will figure out how to make this work. It’s only a matter of time.

    From MC> How many years have people been saying “5 years” ?

    Comment by Radical Bender -

  107. I truly do not get the point of this post.

    And your response to Adam Bruce can be summed up by color…

    The green you carry with you.

    If you had not succeeded at what you had, I doubt your attitude would be a know-it-all with no solutions to the problems you throw out there.

    I work in the music industry.

    I see a problem… I am not posting about it.

    Guess what?

    I am actually attempting to fix the problems and make things better.

    Perhaps you should do the same?

    Comment by Question Mark -

  108. Cable, Satellite and TelcoTV (e.g. AT&T, Verizon) are still well
    positioned to continue to deliver the 52″ 1080p living room experience
    as a cost-effective, low-hassle service.

    Cable in particular is evolving to look more like the Internet than
    one might imagine. Content is already being moved around via IP for
    transport purposes, switch digital video is real, we’re very close to
    seeing widespread deployment of interactive applications on currently
    deployed set-tops and the on-demand content libraries are growing

    There are vendors today developing and deploying cable solutions that
    transcode content in the headend to MPEG-4 streams and deliver them
    over the DOCSIS network to UWB modems and low-cost MPEG-4 IP set-top
    boxes or even directly to a computer.

    The trick of course for the CableCo’s and Programmers has little to do
    with the technology, but will be more about figuring out how to merge
    their video and Internet business models and make it as painless as
    possible for all involved, but do it before it is driven by desperation.

    Comment by mrv -

  109. Sooo, explain to me the presidential inauguration. msnbc, hulu, joost… all of these streamed live for hours with well over 500k people watching on each site.

    Comment by Anonymous -

  110. Mark is venting about something I do not believe is true. There’s a couple of huge errors on his part:

    Error #1: Assuming Youtube does NOT use CDN services all the time. They DO. Last time I checked they belonged to Limelight Networks.
    Error #2: Not counting Youtube as a viable way to stream internet video for FREE. Yes, that’s right kids, it’s part of Youtube’s model to PAY CDNs to stream your videos yfor you! In return, they get the ad revenue and of course the content. For end users like Joe the Plumber, that means COMPLETELY FREE CDN SERVICES through YOUTUBE! That’s exactly what the site is Cuban!

    Error #3: Comparing this shit to cable or satellite in any way. Buying airtime on cable and satellite networks can cost from $200 to $5000 per 1/2 hour and this is conservatively speaking on relatively obscure channels. Not to mention all of the hoops you’ve got to jump through with the censorship that is rampant on those networks. Censorship does NOT exist across all CDNs, they are happy to serve any data you have.

    You assume that when you purchase time on Time Warner’s cable network you’re getting your message out to the world when you’re not. Youre getting your message out to maybe 1/5th of the US market on that cable network, and yes, you’ll have to buy time on another cable network if you want to reach a different market in a different area. 15 cable networks and satellite networks later you may have the worldwide coverage that you were considering with a CDN. at a conservative estimated price tag of $65,000 per day.

    I dont know anyone that spends that much on bandwidth in a day, and I REP CDN service.

    CDNs do hold the cards when it comes to bandwidth, and the best we can do is help them improve their businesses by continuing to renegotiate at more aggressive rates. I’ve had clients pay up to $1 per GIGABYTE streamed, and now they are paying around $.11 per GIGABYTE. Why? Because they didn’t allow themselves to get taken for long, and asked the right questions. That coupled with creating competition by pitting two CDNs against one another, the end user can then start the ‘battle to the bottom’ where both CDNs undercut each other to death while the end user wins.

    Cuban, if you want some advice on how to make CDNs affordable and effective for you, I suggest you email me at the email address I supplied with this comment. I’d leave my number but I’d rather not talk to your blog scabs all claiming to be you. You will find on the other end of the line a fellow intense personality who knows the internet and new media like you think you know your space 🙂

    From MC> You might want to read my post before you comment on it. Youtube wont subsidize commercial businesses. Read their TOS. Nor will they remove their branding. If you think 11c per Gbs is cheap, you need to do some math and maybe re evaluate the value of your services. Let me give you a point of reference. Lets say there are 30mm satellite subscribers. If you went to them and told them you would pay them 2 or 3 cents per sub per month, Im guessing you could get them to put up your 500kbs stream of video for a lot less money and reach a much larger audience with a much better picture than you can get from any CDN. You could have a real tv network. Or you could extend that to all 100mm cable and satellite households. Thats about 2mm or 3mm per month and you could put up a nice little stream to everyone in the country that they could watch on their TV. Not 10k streams,not 250k streams up to as many people as you can convince to watch it. Just like it works on the net.

    You see MSOs and Sat providers like getting paid for their bandwidth. If you do it right, there is no reason to go through a CDN. CDNs are just intermediaries that resell bandwidth. Smart businesspeople can cut deals right with the network providers and get far more bang for your buck and far more upside. But it would take someone who wants to provide the right solution in order to figure that out.

    Comment by djRoME -

  111. Mark,

    This is an awesome post. Where do you think the future of “Internet TV”
    (if you think there is one) lies? Is it live content? On demand content?
    Something nobody has thought of yet?

    We keep hearing everywhere that we are in a new media revolution and that is coo
    cool, but do you think old and new media can be blended together?

    I try to do this with my TV show and it is really an interesting dilemma.
    I wish there was a glass ball to see what is going to happen in the next few yea
    few years…heck even the next six months.

    All the best,

    Comment by David Siteman Garland -

  112. Hey Mark, good to read your latest post.

    You raise some good points. I remember back in the early days with the Clinton-Lewinsky testimony when we at InterVU were crowing about delivering 28,000 simultaneous viewers at 28 Kbps. The cable people must have been laughing their heads off.

    You make a number of points:
    1)If it is scalability and reach, correct, cable has the Internet beat, hands down; totally agree here
    2)If you are talking about business model, on a cost per viewer basis, your model holds up that cable/satellite for large scale events is much better
    3)You say that video distribution of ANY scale places you at the mercy of just a very few CDNs. I disagree. My company (not a typical CDN) has a well known cable TV network as a client paying us less than $3K/month for all of their encoding, user gen, and on demand streaming needs. I would bet their transponder bill and labor for video editing bill is much higher than that.

    We aren’t one of the 3-4 major CDNs you suggest and we have to be very competitive or else any of our 15 other competitors will take the business away very quickly.

    4)How many people that don’t have your clout or that of a larger media company, can call up one of the satellite providers and ask how much they would charge you to deliver to 100pct of their customers, and then call up a cable company and ask the same question, as you suggest?

    Shelf space on most MSOs, even with digital lineups and fiber buildouts like FIOS, is still precious. Having worked in cable before the CDN space, I remember how long negotiations for The Weather Channel took with the MSOs, and Weather is a well known entity. Don’t PPV events require some type of commit?

    If you decide to start up a company and deliver your content, how do you get to the stage where people know who you are, know what you are promoting, and then can be directed to a large enough event or popular enough video? The cost for advertising on cable and satellite to do just that is not cheap. My point is in order to get to the point where your business model logic makes sense, scale really has to be there, or else satellite and cable, if you are lucky enough to get on, will eat you alive, which is truly a gatekeeping function.

    Maybe your point works well for large broadcasters and well known individuals like yourself who greatly benefit from multicasting, but for the average start up guy, in order to get to the scale where traditional broadcasting makes sense, it is a long and expensive process in itself.

    That is where the gatekeeper CDNs fit in. They don’t discriminate and will take anyone with a buck. Most companies these days have pricing with a minimum that scales with a usage based model (GBs delivered) vs the old way of reserving capacity for huge numbers of viewers, and people are using 4-5 CDNs with load directing software to balance capacity for mega events.

    Comment by Todd Loewenstein -

  113. Mark,
    I disagree with some of what you have to say. First my qualifications such as they are, I worked in Both AT&T an MCI’s Tier 1 Enterprise Hosting and managed solutions for most of my career. I’ve designed, sold and implemented bandwidth applications that far exceed the requirements of YouTube and other UGC video providers (Akamai was a customer of mine), but on that front I launched the first UGC video sharing site on the Internet that kicked off the current Video revolution.

    First and most broadly I disagree that CDN is the only way to stream 10K concurrent streams. My own Start-up far exceeded this during our broadcast off the Hurrican Katrina WebCam with a modest 10Gbps redundant connection out of an tier 2 Equinix IDC pushing into MAE East. I could discuss the reasons (Backbone peering out of DC, etc), but the main point is a CDN is not required, merely advisable for the best user experience. We have found that a CDN does scale better (for reasons I can discuss below), but it is a “should have” rather than a “have to have”. If you don’t want to use a CDN you can go with Global Load Balancing and effectively build your own regional Distribution network and the costs are not unaffordable even for a start up, I just wouldn’t advise it for video.

    Where I agree with you. Yes, the Internet is inferior to BROADcast infrastructure for streaming ONE video to MANY viewers. This is due to the fact that the Internet is a 1 to 1 design rather than a 1 to many broadcast design. But they key here is that Internet consumption is not concurrent. On-Demand viewing is the most common video consumption use case online.

    (from above on CDN’s)
    My Start-up was doing 2 Million in-page video streams a day in 2005 against thousands of pieces of content, our costs were low 5-digits, our move to a CDN was precipitated because it’s harder to scale a thousand videos streamed to a thousand people than one video streamed to a thousand people. Lots of reasons for this (Serving out of cache Hard Drive I/O, NIC I/O etc) but in the end CDN’s came to the table with a more cost effective solution that also increased viewer experience.

    Where the Internet exceeds television for democratizing content is asynchronous viewing of content. Overtime a show on YouTube can exceed the viewing of a single time-block on NBC. There are hundreds of video hosts now, and all of them are anxious to sign content and distribution deals, so there is a model there.

    From MC>
    First, if you want to talk firsts, happy to show you this, where we said “ and SimpleNet plan to work together to introduce self-service audio and video streaming to SimpleNet’s customers. Users will be able to create and broadcast their own personalized audio and video programming, including Internet-only radio and television shows, business presentations, or home movies, which will be accessible within minutes of setting up their service, according to both companies.” in 1998

    As far as the rest of your comment, if everyone had a meager 10gbs connection, they too could stream far bigger events. But Im guessing that not everyone has 10gbs they can swamp at their leisure.

    Comment by Adam Bruce -

  114. Mark, I know from your blog posts that you are aware of the
    potential for IP multicast to allow live event streaming to
    scale to “millions of viewers.” Many closed multicast systems
    exist today and deliver SD and HD video to network subscribers.
    AT&T U-verse is probably the largest installed example today,
    but it was certainly not the first.

    The problem with multicast remains that access providers think
    that they don’t know how to monetize it, and without support from
    large access providers, there is little point in offering multicast
    streams. The way to monetize this technology is to offer it before
    competitors and perhaps to bill separately for it. If I want to
    watch a streaming indie band’s concert or my high school football
    team, I am probably willing to pay for it, as long as the billing
    and usage is easy to understand. Pay-per-view has worked for many
    years with a charge scheme that subscribers can understand. If
    access providers would enable IP multicast, they could certainly
    generate revenue from it this way. Content producers, on the other
    hand, can support themselves with advertising.

    I agree with you, Mark, that unicast “Internet Video” is not suitable
    for live events; but I know you agree with me that multicast is a
    viable technology. I wish some Internet tycoon like yourself would
    negotiate multicast distribution deals with ISPs and assist them in
    creating the necessary back-end infrastructure to meter and bill for
    its use. It would revolutionize live video the same way that youtube
    has for recorded content.

    Comment by Jeff -

  115. “PVR’s are a HACK!
    Youtube/Hulu are basically the ultimate PVR, with ever show
    ever aired recorded for future playback.”

    John Sokol, tru2way is about to kick your ass.

    Comment by Jeff -

  116. Internet Video is not a viable business at this point, with the exception of porn. Porn is a huge business tailor made for Internet delivery.

    Cable TV has enormous bandwidth now, and it will be capable of 4X capacity (Wide Band) once we go to total digital signals.

    Wide Band takes the bandwidth dedicated to four cable TV channels and bonds them together into one stream that can deliver 100 megabits per second. Wide Band can download a full length motion picture in High Definition in about 4 minutes. High Speed DSL service would take 6 hours to download the same movie. This is truly amazing speed that will be available to millions of Comcast customers this year.

    Video On Demand – Comcast is building a library of 1,000 high-definition movies and hundreds of channels of on-demand programming by end of 2008, and will eventually have more than 6,000 movies a month with more than 3,000 of them in HD.

    Tru2Way DVR/DVD – Comcast and Panasonic jointly introduced AnyPlay, a portable combo tru2way DVR-DVD. It looks like a regular portable DVD player that you could take in the car or on an airplane, but it is also a DVR that works like a Tivo. It sits in your living room by the cable box and then undocks and becomes a portable DVD player with recorded TV shows or other content.

    With Wide Band, VoD, and Tru2Way available to distribute video I don’t see where Internet Video fits. The cost structure is prohibitive, and even if you are willing to pay for it…it isn’t available.

    YouTube has never made a profit and probably never will. If YouTube can’t make money with their brand, scale, and Google’s mighty ad sales machine…I don’t see how any other service could possibly do it.

    Hmmm…why not buy the cheapest, wimpiest, existing cable channel and turn it into a Video On Demand service for special events, corporate announcements, and a distribution channel for “Internet Video Content”? Sort of like, but using a cable channel for distribution.

    Comment by Don Dodge -

  117. Mark,

    No question video content constitutes the bulk of my broadband traffic. With all of that content on demand though, DVR is a non-issue. For those few ‘weekly’ shows that my wife and I do enjoy, I grab those OTA in HD and DVR to my HTPC. (I’ve discovered I really dig watching and browsing simultaneously.) OTA HD and sufficient broadband speed are the core reasons I was able to cancel my DTV service.

    Your suggestion implies that ‘basic digital cable’ would provide the content I want to view.

    For me this simply wasn’t the case. (I had DTV, all I really miss is HDNET, honestly.)

    The content I view via the internet (some free, some paid) is significantly more ‘valuable’ to me.

    I am a firm believer that broadcast TV and satellite as we know it today will cease to exist in the not too distant future.

    As you so accurately point out, the CDN’s combined ability to handle mega-events will play a direct role in that transition. (i.e.: the inauguration web video debacle.)

    Comment by BillFitz -

  118. Sounds to me like the best opportunities for making a biz model from distributing content are on the CDN side of the market. A smart CDN could make deals with content providers on a long term basis trading equity stakes for free or reduced distribution…or just buy or start their own content creation. It doesn’t compete directly with cable and satelite and I believe your point that they won’t in the foreseeable future. But, you’re talking about huge marketshares, so if I can successfully create AND distribute keeping the cost down, it’s a viable business. While it may never take over cable and satelite, it still caters to a growing market.

    Comment by Michael -

  119. from MC> Depending on what you use the net for, if content is your king, you would be better off using dial up for your web browsing and buying basic digital cable for your content viewing. Take out the streaming video from your broadband bill, and I bet you barely notice the difference in the speed of your basic browsing. Apply the savings to your cable bill, and you get far more content, a DVR to watch what and when you want.

    Are you high?

    From MC > Are you stupid ? How much bandwidth do you think most regular webbrowsing and email uses ? How much effective bandwidth do you think you use for non multimedia ?

    Comment by Brad -


    > Which has advanced more ? Satellite, Cable or the Internet ?
    Mark each has it’s place.

    Some content is great for Broadcast.
    But most broadcast is just prerecorded crap anyhow.
    It would be easy to do this almost identical model using
    live streaming P2P , something I was playing with into set top boxes.

    But Narrow casting, like youtube will eventually kill TV as we know it.
    PVR’s are a HACK!
    Youtube/Hulu are basically the ultimate PVR, with ever show
    ever aired recorded for future playback.
    Why the hell should we wait for the show to be broadcast?
    Once youtube comes out with a subscription STB, TV will never be
    the same. it will be interesting to see how this plays out.

    Think about how much actual live content there is.
    Even most new’s is prerecorded.
    So sports, and major news events like the inauguration and the
    Los Angeles favorite, car chases.
    But little else really.
    They don’t even broadcast concert live. Which I think is crazy.

    I think the home shopping network is one of the few channels that
    really does live in a major way.


    Comment by John L. Sokol -

  121. what’s your point? just because the infrastructure doesn’t exist today doesn’t mean it wont in the future. the demand is clearly there. this industry (online video) is only a couple years old. slow down fireball calling the end of online video or calling it a big lie, imo

    From MC> Internet video has been around for a decade.

    Comment by mark -

  122. Moreover, local cable systems are quickly evolving to provide all the functionality of Internet video.

    UPLOAD, SEARCH, SOCIALIZE, AND WATCH IN HD is all coming to your local cable headend.

    Comment by Morgan Warstler -

  123. Correct me if I’m wrong but my impression is CPM is generally higher for internet videos than for cable shows. So that explains why people are willing to pay higher distribution costs.

    Plus the key for internet videos is it’s perfectly on-demand and asynchronous. You don’t need to sunk in a lot of ad dollar up front to get them to all watch the show when it airs. You can leave your videos there and let word-of-mouth works its magic. So your marketing cost is a lot lower and your operation cost in most cases is not about paying for maximum throughput but a more spread-out average usage.

    However, I do agree with you that internet is not the most cost-effective channel to stream huge events like the inauguration, which everyone wants to watch at the same time. But it does bring a lot of benefits to a different type of video. Videos that have small marketing budget, long shelf-life and viral!

    Comment by Jamie Lin -

  124. How much amplification for a CDN can you get from a swarming service such
    as Bittorrent DNA?

    Comment by Don Marti -

  125. You seem to be operating under an assumption: that “Internet video” is just like TV/satellite/cable, only “on the Internet”. You then extrapolate from that and point out that video on the Internet, with its inherently point to point topology, is ill-suited to one-way mass broadcasts.

    The problem with this view lies in two areas: what constitutes “Internet video”, and what TV/etc. actually *is*, from the consumers’ perspectives.

    I’m working from myself as template, as it’s the topic I’m most expert in. I don’t own a TV. I don’t watch live broadcasts of any kind, not concerts, not movie premiers, nothing like that.

    When I want to watch video content, it’s most likely either been DVD in the mail ( or downloaded one way or another. Lack of instant gratification in downloading isn’t a big deal, as the extra flexibility in choice makes up for it.

    The trouble is, I have far more interesting things to be doing with my time than watching TV, consuming a one-way medium. I’d much rather be reading interesting material and – this is crucial – partaking in conversations, just like this comment.

    The one-way consumption has become *highly* selective, where only the very best material is consumed, with a complete absence of advertising – and paying a premium for that. Moreover, the only reason to do that one-way consumption is so that I can spend time with my significant others and have a shared experience.

    The “default” is not to sit in front of a TV; the default is to interact. And it’s a lot better than TV ever was.

    So there you have it: Internet video is a myth, not because satellite/cable/broadcast is better, but because TV is dying.

    from MC> More people watch TV now, in terms of absolute numbers and hours per day than in any time. More money was spent on TVs this past year than on any other CE device, and probably more than on any other consumer item period. Thats how dead TV is. If you think content creators are going to invest in content based on your economic model, you are delusional. In fact, those who have invested in making content for the net have found failure rates equal to and worse than on traditional tv.

    Comment by Barry Kelly -

  126. Why so upset Mark?

    It took cable awhile to scale up to reach millions of homes.

    It will take internet CDNs awhile to do the same.

    But rest assured it will happen.

    From MC> Lets see, we started streaming audio in 1994, video in 1997. 10 years ago we were doing gigabits of audio, with hundreds of thousands of simul users, 400k video, and had a multicast network. Now 10 years later, 1mm simuls at 500k will get you accolades. In the meanwhile, DirecTV was launched in 1994 and look at it now. Cable went from analog with what, 60 channels to HD and multiple gigabits of bandwidth. Which has advanced more ? Satellite, Cable or the Internet ?

    Comment by Ken L -

  127. Mark, I have been following your blog for years. I am always impressed with your insight. You are a great man for your works and efforts to impart your knowledge and experience. I loved the book “the number” by the way. LOL!

    Comment by Mike -

  128. This is good information to have Mark.

    I am going to research this as you suggested on the costs of presenting.

    Who are the major CDN’s? Can someone provide a list?


    Comment by dave -

  129. With all due respect Mr. Cuban, I happen to disagree.
    You seem to define internet video as a video that can only be streamed in real time via CDN’s. Not so. Companies like NextNewNetworks and Revision3 have proven that you can aggregate consistent reliable audiences in the tens of millions (NNN hit 300mil views alone in their second year over a dozen or so shows).
    As a past employee of NNN – and a current entrepreneur with my own video venture, we find significant value in the “super distribution” model; connecting with your audience where they consume media– not where we dictate them to watch it. Distributing through YouTube, iTunes, VEOH, Break, Yahoo, etc., you’re able to hit critical mass exponentially quicker- thus allowing us to target big brands and get bigger spends.
    Yes, you can never generate a scalable business off of 10,000 viewers- but, with reliable, consistent, professional episodic content, you’re able to generate a huge following. If you play your cards right and actually focus on a specific vertical, you can draw endemic advertisers that are willing to pay a premium (see Gary Vaynerchuk, Wine Library TV).
    Media buyers are comfortable buying television/video as they understand the unit. Media buyers are comfortable buying display IAB ads. Web video is becoming a format that introduces contextual relevance coupled with sight, sound, and color—with quantifiable results. You can measure the efficiency and effectiveness of ads TODAY.
    The biggest challenge to web video, IMO, doesn’t involve hitting critical scale or distribution – it revolves around standardizing what exactly a video view is. How you measure it. What’ll be the generic ad unit. Preroll/postroll/overlay? Companion ad?
    Traditional banner ads; 300x250s, 728x60s, 160x600s just aren’t effective enough anymore for big brands. While I’m not a member of “POOL” ( ), I do believe that they have a legitimate shot at creating an IAB-like standard ad format.
    Web video’s success isn’t contingent on streaming distribution, its contingent on the definition of what is a view – as well as the scalable ad unit.

    Comment by Kenny Herman -

  130. Mark Cuban knows what he’s talking about.

    Comment by Johnny Smith -

  131. Mark,

    How does Amazon’s web services come into play? Would you consider them a CDN?

    I came across the same cost issue, but not for a commercial application of internet video. I keep a family website and wanted to make videos available via a YouTube clone script. This script works great within my home network, allowing me to upload videos, which get converted to flash, and can be viewed very well over a 100Mbps home ethernet network. As soon as my relatives hit the site via the internet cloud, the videos came to a crawl.

    I investigated doing an Amazon image of my linux box so that I can serve my family videos at a quicker pace, but even the nominal cost of Amazon’s “per cpu hour” rate adds up fast when it is 24×7. So I can’t justify the cost from a family video perspective.

    Thanks for the insight,

    Comment by John Nguyen -

  132. Hey Mark, at what bitrate are you using as the example? Akamai and Limelight can both deliver to 1M+ simultaneous viewers for something like a 500K stream. Are you talking traditional broadcast quality of a couple Mbps per stream or HD? What’s the metric for the video quality?

    From MC> I acknowledge they can reach 1mm simuls. But the problem is they can only do 1 event at a time at that size. As a result, its impossible for a real market for over the top video to develop. If each CDN can only do 1mm simuls at any given time, what happens to everyone else ? The point of the post is that if the 2 biggest can only do 1 major event at a time, the concept that a reliable competitor to cable and satellite can develop is a joke. And if in the next 2 years they each can do 2mm million total simuls, you still have the same problem . If they can do 5mm simuls at once, you still have the same problem. Its a market where all the power is consolidated in just a couple companies. That isnt freedom, thats a controlled market with minimal scalability.

    The perception that internet video can and will compete with cable and satellite is and will continue to be a joke. It can offer video services for specific events and applications, but that is a long way from competing with TV

    Comment by Dan Rayburn -

  133. I bet the receptionists at Akamai and Limelight are a bit pissed off at you right now with all the calls for quotes they’re getting.

    Comment by Shawn Shepherd -

  134. Interesting thoughts; I totally agree about not being able to simultaneously cast to, errr I don’t know, a billion people a live event, therefore, in a way, it is the ‘Great Internet Lie’. It is impossible to beat the cost of streaming a large live event like the concert you mentioned via the internet.

    However, interesting to note; there are a lot of small CDNs starting up, the 4 you mention are the big ones. Although they are not as big as Akamai or CDNetworks, they provide a service for simultaneous download and stream. So in a way, it will get cheaper as more companies enter into this market, but it will take some time for CDNs services to get cheaper. It was the same case with satellites and cable companies before, I believe.

    So it will eventually be cheap enough for me to put out a live guitar concert in my pyjamas in my room somewhere in Kuala Lumpur, Malaysia to reach 10k simultaneous viewers?

    Just my thoughts.

    Comment by razman -

  135. But with the internet you don’t need to stream live. The content is
    on-demand and for that, handling the bandwidth isn’t that hard

    From MC> The issue is simultaneous streams. Live is harder, but simultaneous streams defines absolute resources. it doesnt matter if those streams are live or on demand.

    Comment by Cameron -

  136. Love your blog, i’m a diehard Mavs fan. Although you’re right that 3 CDNs control about 80% of the market, funding for CDN startups had been accelerating (before the meltdown, atleast). Therefore I think the cost of hosting on CDNs will also fall – similar to how network and then cable television were prohibitively expensive before their ubiquity. I predict the same will happen with CDns, only even faster.

    Comment by Orijit -

  137. Mark,

    Excellent article, but I am not sure if I’ve been sucker punched yet.

    What would those numbers have been if we tried to write this article 5 years ago? What might those numbers be 5 years from now?

    The CDN’s know what their weaknesses are.

    Web video of the inauguration was a clear demonstration of how weak mass content delivery is today.

    I’ve successfully ‘cut the cord’ from my cable and satelite TV providers for programming.
    Ironically I still pay my cable company for the broadband I use for internet TV and movie viewing.

    Do I miss some content? Yeah, but I like the money I am saving more.

    from MC> Depending on what you use the net for, if content is your king, you would be better off using dial up for your web browsing and buying basic digital cable for your content viewing. Take out the streaming video from your broadband bill, and I bet you barely notice the difference in the speed of your basic browsing. Apply the savings to your cable bill, and you get far more content, a DVR to watch what and when you want.
    Just a thought 🙂

    Comment by BillFitz -

  138. I think you’re not making the proper distinction between content and the medium it’s delivered on.

    You’re main point is the lack of technology, which is a medium, not a content problem. The lack of a proper “streaming” CDN is, according to you, prohibitivly expensive. To me, it seems like a business opportunity (see

    But you’re also assuming that internet content will try and compete on just a large scale as TV. As you said, it’s very expensive to do that. But it’s also *very* expensive for satellite and cable as well. And if advertising trends continue the way they are, TV companies will have their margins squeezed and will have to become lighter.

    I’d say that the trend is going to be a mean-reverting one. Internet media is and will continue to catch up in terms of attention and eyeballs, and larger media companies will have to reduce overhead just to have a business model.

    Comment by Steve Place -

  139. The current internet video providers seek a future of widely-available wireless internet, and the capabile of streaming vast catalogs of on-demand video. The goal of wirelessly delivering video to homes sounds, well, like a 1960’s idea. Didn’t the TV wirelessly deliver video? So why is it thought novel, or a ‘new paradigm’ to get video transmitted w/out wires to your home – and that ad content will support its free delivery. It’s less innovation, and simply disruptive technology. One wonders whether the broadcast spectrum might become more valuable as a reliable, accessible, clearly-programmed source of video, than a slew of internet video webcasters, with some much varied content as to be all but un-searchable or with programs put to the top of your search results by those willing to advertise. Some broadcast, some cable, some internet – all have some place in my entertainment room.

    Comment by Thomason -

  140. This is the same conversation you and I had about multicast back in 1997.

    Bandwidth is getting cheaper very slowly. Storage is getting cheaper much faster.

    Comment by Erik Schwartz -

  141. Mr. Cuban,

    Very interesting blog you have posted, but I would like to know what the purpose behind it is. Just curious to know why you would spend your time writing something like this. Is there some underlying reason that is not obvious to the “average joe”?

    Blog fan,

    Comment by Jeramy -

Comments are closed.