As is the normal way on the internet, people love to read headlines. Of course they rarely read the substance of the article behind the headline. They take the headline and run with it as gospel. This is particularly true of internet video and the presumption by many that the internet will be so vast and powerful that its a foregone conclusion that video on the internet will replace traditional television delivery.
Ain’t gonna happen. I can go into depth about it, but my buddy Dan Rayburn does a far better job of it. He asks the simple question of “how is it the Youtube , with all of google’s resources, cant solve their buffering problem?” Then he answers much of it.
Now maybe Youtube will fix their buffering problem someday, along with the other issues that Dan addresses, but it wont be easy and it wont be quick.
You can find Dan’s work here. Make sure to also read the comments. He does a good job there as well.
Let me add a couple other thoughts. There are many that think that video over the internet will “set them free” from having to deal with a small number of big companies (think cable, telco). If that is what you think, you better think again. There are maybe 3 companies that can stream to 1mm or more simultaneous users. Google (GOOG), Limelight (LLNW) and Akamai (AKAM). And that 1mm simultaneous users isn’t just for your content. That is for EVERYONE’s content and they cant get much beyond 2mm without big problems. More importantly, if you want to stream your content to millions of users at once, its going to cost you an incredible amount of money.
Which leads me to a lesson for all you netizens who are jazzed up about over the top video. If you really believe the demand is there and your content will command 1mm simultaneous users , its probably cheaper to pay Directv (DTV), Dish Network (DISH) and Comcast (CMCSA) to create a channel for you and let your viewers watch it on tv.
Let that sink in. Its going to be cheaper to have the big traditional cable distributors offer your content to viewers than it will be to reach a large audience on the net. That’s for a one time offering. If you plan on doing it more than once or on a regularly scheduled basic, there is no question its cheaper to do it this way. And the picture quality will be dramatically better.
In fact, that is probably a great business opportunity for satellite, telco and cable companies. Open up times to bid to offer content over their networks. You want channel 1020 on Directv, its X dollars per hour minimum or the best price bid. Here is how you provide your content to us. You can buy marketing from us as well. Directv, Dish, Comcast could make a boatload of extra money offering this service.
And i can give you one more option. It may be cheaper to go to a movie theater chain and pay them to broadcast the content you want people to see via digital to their theaters. As long as its a slow night and they can sell popcorn, I can assure you it will cost you less than a content distribution network would charge to deliver to thousands of viewers.
Maybe someday over the top video will be a realistic alternative to traditional distribution of content, but its not now and its not this year or next or the next and probably not the year after that.
Mark,
I may be missing something, but Mr. Rayburn’s article addresses YouTube’s backend technology issues but does not argue that these issues are related to the large user base or have any impact on the scalability of the service.
Moreover, please explain how marginal bandwidth costs would be prohibitive, as you imply, in a model like Apple’s rumored TV subscription service, where a viewer would buy a basket of network programs for a $3-$4 monthly fee. In such a model, if the revenue is split between the network and Apple, the network would see more, much more, in revenue/month/sub than its affiliates would see under even the richest retransmission consent agreements, and Apple keeps whatever is left in its $2 share after paying for the bandwidth. Is the marginal bandwidth cost of the bandwidth necessary to deliver all the network shows a customer watches in a month (assuming the shows are delivered efficiently, i.e., on-demand, maybe 10 GB) really more than $2?
Incidentally, the majority of over-the-top services that are getting traction these days are on demand services rather than linear streaming services delivering a live programming feed of a particular channel. Is your analysis different in the former circumstance?
Thanks for your thoughts.
the math for an apple tv is not nearly as good as you think . Most industry people think that Netflix pays about 5c to stream a 90 min movie. Extrapolate that out and say that its 10c per 3 hours. Thats 80c per day. Thats 24 dollars per month.
Now you might think that Apple will just ask to be allowed to deliver each show ala carte ie, on demand. That would destroy all the other subscription deals they have with Comcast, Verizon, Directv, etc. Who would all ask for the same thing.
And in comparing live vs on demand. Its the same thing. An over the top video has to be posted at some point and “go live” Thats when most viewership takes place.
Finally, most networks and content for TV producers arent stupid. They had plenty of archival content to give Hulu and Youtube to make it viable. They wanted to experiment and learn how its consumed. But as they do new deals with MSOs, satellite providers, etc, those people are restricting what can be offered. Which is why Hulu will have to go subscription for new content. The pricing for which, when added to an ISP bill will make Cable look cheap