It's as easy as IPTV

Packet Vision Managing Director Patrick Christian tells MMM Editor David Murphy why he's so excited about IPTV

UK company Packet Vision is developing the first all-in-one network box for enabling addressable advertising on IPTV. The company says it will allow advertisers to move from 'spray & pray' campaigns to highly focused ads that are relevant to the viewers, as IP has the ability to tell the broadcaster what's being watched on each TV. While its aimed primarily at the TV-over-broadband market, Packet Visions offering will also be relevant to mobile operators offering 3G Mobile TV services.
The company is privately funded, and to date has raised 1.75 million from two round of funding through 16 Angel Investors. It is about to start a new funding round with a view to raising a further 5-6 million in the autumn.
Packet Vision will launch its TV advertising delivery platform at the IBC2006 exhibition in Amsterdam, which runs from 8 12 September. So what's the big idea?

MM: Where has Packet Vision come from?

PC: Im an engineer. Ive always been motivated by technology. In 1998 I co-founded a Voice over IP company called Vegastream. At that time. VoIP was pretty new stuff. We sold the company to Pace in 2000, and thats where I really started taking an interest in TV, and specifically, IPTV. I left at the end of 2001, and since then Ive been looking for a niche to build a new company around. I started Packet Vision in May 2004, because I believed I had found the niche. That niche is addressable advertising. TV ads that are driven by and relevant to the viewer, as opposed to the traditional form of spray and pray ads, where you chuck it into the ether and for some people who receive it, it will be relevant, but for an awful lot more, it wont be.

MM: So why are you so excited about it?

PC: The IP element gives you two things that traditional broadcast TV does not have. It gives you addressability, which means you can stream personalised advertising content to viewers, down to individual TV sets. And it gives you bi-directionality, so you are able to get feedback from the viewer. Those two aspects are critical to delivering addressable advertising in the TV environment. So for the last two years, weve been working to create the platform to make this happen. The nearest equivalent is the way cable companies in the US have done local ad insertions. The national networks deliver TV coast to coast, but as it goes through the local networks, they have the opportunity to insert local ads into the commercial breaks. So this sort of addressable advertising, but its just based on geography. We want to go far deeper than that, and target viewers based on their interests and all sorts of other things. To do so, we had to create out own platform because the technology that delivers local ad insertion to cable companies is just too big and expensive. 7ft. high racks of expensive boxes. Its fine for the cable world, where the networks have hundreds of thousands, or millions of subscribers, because when you divide the cost of the kit by the number of subscribers it is delivered to, it is still cost effective. But its no good to us. Our dream here is that addressable advertising will go the way the Internet has gone. More personalised, more focused on viewer, and with more feedback from the viewer. So we have taken the Internet approach and created a  platform that sits in the IP network, but the box has been designed like Cisco would design a router or a switcher. Low cost, easy to manufacture, highly scalable, and easy to build up the network. So you start small, focus on large demographics groups and as the advertising becomes more targeted and more relevant to more people, then you can scale up.

MM: It sounds like a great idea in theory, but is it really practical? You get all this slicing and dicing of data and targeting people based on their interests in the direct marketing world and yet most DM campaigns are still considered successful if only 95 people out of a 100 throw the mailshot in the bin.

PC: Its a question of timing. A question of when it will happen, not if it will happen. You only have to look at how Internet ads have taken off and look at the factors driving that. If you could apply the same factors to TV, it would be equally attractive, because TV is a more powerful medium to deliver the messages. But it wont be a big bang. It will be evolutionary, and certain aspects will happen before others. So things like analysing viewers activities and interests to target them more accurately will be a little further down the line.

MM: So what sort of things will be possible in the near-term?

PC: Measurability is one. Just knowing how many viewers saw and received your messages is a powerful piece of information. This is an implicit thing that comes out of IP. Its a 2-way process that allows us to count what viewer activity there is on a set-top box by set-top box basis.

MM: And are the TV providers happy with such a degree of accuracy? I get the impression that in the broadcast TV world, there is some resistance to measurement methods that go beyond saying that if a person was in a room when an ad played out, they must have been watching it. As far as they are concerned, the bigger the number, the better.

PC: The IP providers are OK with it. But to get the most accurate information, we need real viewer activity, so if there is a set-top box in the room,  it may be answering polls from the network to say it is switched on, but that does not imply that someone is watching it, so there are still some grey areas in terms of measurement. But rather than try to answer all these types of questions from the  beginning, we have created the platform to be very usable, so that it can be used to answer some of the questions no one has answers to. We have designed the platform to be open to third party software that will allow everyone to get a handle on this and answer all those questions.

MM: And where does mobile come into all this?

PC: We talk about IPTV in a TV context. So the sort of people we think will be the first users will be the telcos who provide the broadband services, and who are delivering TV and video services as well. But part of our vision of the future is that everything will go to IP and people will be watching television on all sorts of different devices, including mobile phones. And all these delivery mechanisms will be IP-based. As a company, we are not being prescriptive. We dont want to be about how it is delivered to the viewer, so long as the advertising the viewer sees is personalised, relevant and driven and controlled by the viewer, in a way that is satisfactory to the advertiser.

MM: Wont this be very expensive though? There is software that can personalise email marketing messages so that if you clicked through from a previous email to a retailers website and looked at training shoes, youll get a different email to the person who clicked through and looked at T-shirts, but if you start making 10 different versions of a TV ad, isnt it going to get out of hand?

PC: There is some work being done by companies like Visible World. Normally when you create a 30-second ad you would storyboard it, divide it into scenes and then film the individual scenes. Visible World are creating software to allow you to do a two-dimensional  storyboard. So you have the different scenes going in one direction. Then in the other direction, each of these scenes can be one of several possibilities, and this allows you to piece together different ads using different permutations on these scenes.

MM: Which would still cost more, presumably?

PC: It would yes, but the point is that if you have a 30-second ad divided into 15 two-second scenes. With just four permutations for each scene, the permutations run into millions. So although there is more work to be done to create the individual components, once they have been created, a small number of permutations can produce a large number of personalised and individual ads.

MM: Do you see this being popular among any particular sectors of advertisers?

PC: Ask me again in six months time. We are just about to announce our first trial with a network in the US and one in Europe, and another two a few months down the line.

MM: And have you had much interest from mobile operators?

PC: We have had some discussions, but we are a small company so we have had to choose where to focus and for the moment, we are concentration on the telcos providing TV over broadband. Having said that, a large proportion of the development work we have done to make the platform work in  the Telco TV environment translates into mobile environment as well. IP networks have a lot of commonality, whether you are delivering via 3G to mobiles or to PCs over broadband. For us as IP guys, this is great.