One of the most talked-about startups came out of the shadows this week, as The Venice Project revealed itself to the world as Joost. The company invited The Reg to its West End offices for a demonstration and a chat with CEO Fredrik de Wahl.
Joost is an interactive, IP-based TV software system from the people who brought you Kazaa and Skype, Niklas Zennström and Janus Friis, considerably richer after the $2.6bn purchase of the telephony start-up by eBay.
The parallels with the founders’ earlier projects are hard to miss. Joost is P2P PC software; it’s free to download and use and requires no special hardware; it’s based on proprietary software; and the technology is cooler than the business case.
The most striking similarity, however, is that it challenges incumbents’ delivery systems. What Kazaa did to music distribution and Skype did to telephony, Joost wants to do for TV.
Only, this time round, Friis and Zennström are being a lot more co-operative with the incumbents. Around 25 of Joost’s 150 or so staff (no skunkworks project, this) are in London, and the steady stream of visitors from media production companies signals that Joost’s desire to partner from the get go. London is certainly the place to be if you want TV ideas, as it’s the birthplace of so many new formats in recent years, giving the world (for better or worse) reality TV. Sorry, World.
This is a shrewd move by Joost,and probably a necessity. TV is ultimately about watching something compelling, and compelling TV is made by people who know what they’re doing. Despite the buzzword-heavy sales pitch – we heard the words “long tail” and “user empowerment” repeatedly – Joost needs TV production companies and aggregators.
Joost in time
So what is it? Quite simply, Joost turns your PC into an instant, on-demand TV. No set top box is required.
There’s really three parts to the technology. Joost operates a server farm to cache the video, manage seeding, and maintain QoS, but the idea is that much of the content resides on people’s own PCs.
Like Bittorrent it takes advantage of your largely under-utilized uplink. The video is received over a proprietary DRM connection, using encryption of Joost’s own devising. Thirdly, there’s a Mozilla-based client, for advertisers to develop commercials or marketing tools, and software developers to create their on screen widgets. It comes with a chat client, but examples of Joost widgets include giving clips ratings, and noticeboards.
Because it runs on a PC (Windows now, Mac and Linux to come) applications could include anything that’s permitted to access the host’s services. deWahl cited one example, custom remote control app for a phone, similar to Salling’s splendid Clicker program.
From the outset of our demo, it was clear Joost has cracked some fundamentals. TV runs instantly, full screen and at high quality – there’s no lag from activating the TV software to getting something to watch.
It runs in a window too, of course, but there’s no messy furniture. And unlike Bill Gates’ whacky vision of TV, it isn’t a split screen with one half taken up by text ads. The on-screen widgets are discreet, alpha-blended overlays created using SVG and HTML.
So right away, this is YouTube done properly. The chat widget takes us into familiar Interactive TV territory -you can rate clips, chat with friends, and so on.
Joost has already been the subject of some concern, as most new budget broadband users in one of its launch markets, the UK, are capped. Joost uses a lot of bandwidth: a note on the company’s website explains that an hour’s TV downloads up to 320MB and uploads as much as 105MB.
Joost’s CEO Fredrik de Wahl told us the P2P effect, which diminishes bandwidth requirements, only really comes in at peak times, lowering the bandwidth requirement to 220MB an hour. Unless UK ISPs change their policies in response to demand, Joost will remain out of reach. (Joost tells us the bandwidth requirement is a 500kbit/s line – our demonstration was apparently running over WiMAX).
Joost has already been criticized for failing to strike deals with major distributors. But it has already signed up Warner Music and production company Endemol for the beta, which isn’t too shabby, and a good pointer to the future. Production houses such as Endemol stand most to benefit from Joost, as it gives them another route to market.
Will it pay, however? For the established TV players, this is the big question. For Joost, it’s a question of faith. For PC owners, it’s a fairly straightforward win-win.
Let’s break it all down.
Stripping away the endless buzzwords about the “User Empowerment” and “The Long Tail” (De Wahl referred to his server farm as “Long Tail Servers” throughout) Joost’s bet is that targeted advertising will succeed. Joost will take a cut of the advertising.
De Wahl said that TV advertising was currently immensely wasteful, as most people who see a diaper (nappy) advertisement don’t need to buy diapers.
But this is just the start. Adverts could be based on geography, or time. If a heatwave is due, you may prefer to see an ice cream or air conditioning advertisement.
More likely, said de Wahl, they will be based on viewing patterns, as Joost would collect this information and share it with advertisers. This could cause some discombobulation amongst viewers, but then already it’s pretty strange to see children’s toys advertised during Shameless (http://www.shamelessantisocialnetwork.com/). And even stranger to see health foods there.
Your reporter expressed delight at seeing diapers advertised on TV, as it meant that the company hadn’t built up a demographic database of its users, and didn’t know who I was.
De Wahl pledged that data would be anonymized and that Joost would protect users’ privacy.
In another example of how advertisers might gain, De Wahl said, Joost could give audiences access to a global audience. For example, there may not be an audience of 10,000 in the UK for a programme about pottery, but there might be an aggregate audience of 10,000 around the world
But who would translate it into Armenian, we wondered?
That’s where the widgets came in, said de Wahl – as fans could provide the subtitles. Then again, we could be watching an Armenian pottery program with added English subtitles. Be still, my beating heart!
“Because we’re sharing CPMs, we’re driving CPMs higher,” he said. MTV, he claimed, was seeing the same CPM returns on its ad spending through targeted campaigns.
Examples of adverts include Blipverts, Midrolls, Prerolls and Overlays. It’s new territory – search Google today for a combination of this jargon and the only story you’ll find, most probably, is the one you’re reading.
(In Joost-speak, Blipverts, we learn, aren’t the compressed high-speed, subliminal clips of Max Headroom fame – but “brought to you by … ” bookends.)
Then there were more familiar marketing tools – a car show that takes you to a web page, for instance.
Joost in time?
So it’s impressive technology, but is it a business?
Joost is great news for PC users, because it does video so well. It makes Google’s $1.65bn purchase of YouTube look particularly reckless, so much so that we can only presume that Google took a look at The Venice Project and decided it could do better by itself. After all, Joost wants a cut of Google’s advertising market, and anyone operating in that space eventually faces the same bleak choice as a software company competing against Microsoft: sell out, or prepare for destruction.
And it’s not just Google. Hollywood has embraced Bittorrent, and the major telephone incumbents have embraced VoIP, offering a seamless package (BT Fusion), their own cut price packages (everyone else) – or just cutting prices. With TV, and particularly IP-TV, the incumbents are already in the living room, are already interactive, so what don’t they have? When it comes to “user generated content”, Saddam isn’t hanged every night.
(De Wahl said he saw Sling as a “patch”, and left no doubt that he viewed it as temporary. This is risky talk).
There’s a bigger problem, however, and it isn’t just shared by Joost. As channel choices proliferate, and TV is in competition with alternatives such as the net and computer games, so there’s an ever shrinking audience chasing a dwindling pot of money. Joost is potentially good news for original programming producers because it gives them another route to market. But they’re unlikely to adopt a distribution channel that lowers their overall future earnings.
As we’ve seen with so many “empowering” or “revolutionary” technologies, they end up concentrating power with those who already have it. Sor the sad story of the Net Neutrality campaign helped AT&T, here.
Joost’s best hope is to convince advertisers that they can get from here to there, somehow, by converging the web and TV. But then you’ve heard that gag before. PCs don’t belong in the living room, and if they eventually get there, it’ll be without a mouse or keyboard – two things you really need to get the most out of the interactive, or “social TV” elements Joost brings to the party – the chat, ratings and message boards.
I wondered if De Wahl often paused for a moment to think about the negative consequences of the movement to an atomized media, with the emphasis on feeding the “Me, me, me”, that he’s helping bring about. (Time’s Person of the Year is you, you, you (http://www.theregister.co.uk/2006/12/18/time_magazine_poty/).) When a few websites destroy a town’s newspaper, this isn’t a net gain. Proliferation drives costs down, and programmes get worse, only there are more of them. Did it keep him awake at night? He couldn’t have looked more surprised if I told him I drank a gallon of seawater every day. Of it didn’t.
He did agree that the mid-market producers were key to Joost’s success, although he called them “mid-tail” (Long Tail, you see). We are prisoners of the metaphors we choose.
De Wahl said he and the founders were motivated by what “TV really should be”, but I don’t think many people share that vision. Substitute the word “PC” or the word “internet” for “TV” and they’re undoubtedly onto a winner.
As for what TV really should be, the answer’s never been in doubt. Good programmes, please.