Has Joost’s New CEO Got It Badly Wrong?
Around a month ago, I blogged about Joost - the new P2P IP television company. Joost aims to be the primary entertainment platform in the home. In that post, I made two points: 1) People don’t really want to watch TV on their computers, they want to watch TV on their big-screen televisions; and 2) the Net doesn’t have enough bandwidth to deliver high-quality streaming video into the homes of consumers on demand.
Fast-forward a month, and Joost has hired a new CEO - industry veteran Mike Volpi. In his early interviews, he’s made it clear he wants to see Joost not on computers, but working on big-screen TVs e.g. through set-top boxes. That addresses my original point #1 above, but it doesn’t address my point #2.
The Big Miscalculation?
Clearly, Joost’s new boss is betting that, in a few years, home broadband connections will have enough bandwidth to enable high-quality video streams to enter the home on demand, in real-time. The amount of bandwidth coming into the home is increasing exponentially, so at first glance, it seems a reasonable bet. However, when I look at the trends in detail, I simply don’t see how that’s going to happen.
Now, the whole Joost team is a high-quality outfit, and this is their core business - so I’m sure they’ve thought about this more than I have. I’d love to know where I’ve got this wrong. Here’s the problem, as I see it…
The Problem
The core value proposition of IPTV is to enable someone to watch anything they want to watch, at any time. That’s pretty disruptive, because you can’t do anything close to that with today’s cable/satellite television. Every piece of TV content will be sitting out in the cloud, waiting to be streamed on demand. Great. Now, let’s say, to be a reasonably compelling replacement for cable/satellite, you have to be able to service two televisions in the house at any one time. That means, you need enough bandwidth for two simultaneous high-definition video streams. At, say 8Mbps per HD stream, that means you need 16Mbps of truly usable bandwidth.
No problem, you say. After all technologies like ADSL2 and VDSL mean that 16Mbps or greater connections are here today for some people, and will be widespread in around two years. So Joost has perfect timing: build the platform now in 2007, and in 2009 it will be ready to dominate the world. Right?
Wrong (I think). Your home broadband connection is not a leased line - all that bandwidth simply isn’t yours. Your 16Mbps bandwidth is effectively shared with at least 20 other people (often 50 other people). That’s why home broadband is so much cheaper than the leased lines used by companies. If everyone sat down to watch IPTV at 8pm, their 16Mbps connection would drop to a speed of less than 1Mbps. That’s not close to enabling even one HD video stream.
IPTV With Contention
So, to me, it looks like IPTV won’t really be a serious contender as the primary home entertainment platform until home broadband connections can stream video at any time, even taking contention into account. My estimate, then, is that connections to the home will have to reach 320Mbps before IPTV can dominate; that would provide 16Mbps connection to the home when everyone is trying to max out their connection by sitting down and watching TV.
When will that be? Well, if you look at the history of home network connections stretching back to the 1980s and going to the present day, you get a nice exponential(ish) shape curve. If you extrapolate on this trend - one which has been unbroken for the last 25 years - it tells you this: IPTV will not be viable as the primary TV platform until 2017. And that’s way too late for Joost.
It seems to me, then, that Joost will have to change its aims of being the primary entertainment platform in the home, because it looks like IPTV won’t be able to disrupt the status quo for a decade…
lorenzo wrote:
Simon,
where did you get that Volpi quotes about getting Joost to the set-top box ?
Joost will be ready for primetime if NGN or FTTH or 100mbit full duplex, whatever you call it, starts being deployed next year. By 2010 that could have some critical mass, in the meanwhile Joost will be living off SD.
If a high-flying guy form Cisco joins, maybe NGN is ready for primetime too, beyond anyone’s expectations.
Don t get me wrong, I don t like Joost either, I have no ftth now, my DSL is the biggest bottleneck in the way from any webserver to my videoplayer, I don t give it for free to any p2p, only bittorrent gets that,
On the other hand Joost will be building on the PC interactive features that aren t yet available to the TV, those might become desirable. If Joost than manages to move those features succesfully to the living room, via TV or set-top box, then it could really shake that area from decennial torpor it is in. Then it might become a serious contender for traditional TV. There is juice there
Now i Know that you have some business in automatic analysis of HD videos. I really don t know much more, but if you were to deliver your innovative services, not to PC’s but to the living room, wouldn’t you need some sort of better set-top box which can offer a user interface for search, or you can deliver rich video syndication format to ?
then Joost moving in that direction does deserve some credit for the bold move and for the possibility of accelerating change.
Posted 15 Jun 2007 at 6:20 pm ¶
simon wrote:
Lorenzo, Volpi spoke to the New York Times about the set top box stuff. It’s not that I don’t like Joost, by the way - I just don’t see how it can work well enough to take over from cable/satellite for years.
100Mbps home broadband connections would allow workable SD streaming for mass usage - not sure it’s enough for HD because of the P2P overhead, combined with contention. I’ll be impressed if true 100Mbps broadband (i.e. up to multi Gbps contended over 30 users) becomes widespread in the next five years. As I say, I suspect that’s more like a decade away - it needs pilots to be deployed, and monitored, and then for the business cases for large-scale build-outs to be made, and then the massive infrastructure build-out to actually happen etc.. If I’m wrong, that’s great for Joost, and great for everyone that gets the connection!
BTW - as an aside, for information, our HD video analysis business is not focussed on consumers. We’re not big enough to go after that opportunity, I’m afraid. You probably have to be Google or Microsoft to attempt that, and even then, it might be too challenging…
Posted 15 Jun 2007 at 6:46 pm ¶
lorenzo wrote:
Simon,
I mostly agree with you and Joost does too since they stream at 800kbit, only SD.
That s why I think the set-top box announcement makes a lot of sense. Set-top boxes can become an area of great innovation, partly because of triple plays, but moslty for the same reason why Murdoch sinks money in every social network he likes. It s not putting television on the web, it is augmenting the television experience with the web.
Tv universe is still way bigger than the web, in the sense that TV has skills in programming, live events and newsgathering that will elude the web for sometime. Tv has the viewers and the advertsing money. Tv has still more bandwidth to the house than the web.
Anyone with a bold strategy can find in network set-top boxes a way to tap the rich Tv universe of blockbuster movies and live events and mash it up (forgive the buzzword) with the social features of the web. Of course it’s an hardware play in crowded ecosystem made of large TV networks and large consumer electronics firms, but they founded Skype and they re off a good start in VC terms.
Posted 18 Jun 2007 at 1:00 pm ¶