SCOTT FERBER: TV is a portfolio-managed enterprise. There’s approximately 80% of business transacted in a futures or upfront market like we are in now. Premium content matters. It’s scarce and if you query the history of marketers, video…is the most powerful for generating sales not at the direct-response moment. The power of TV [is] to brand something so when I’m not in front of the communication medium it puts that brand at the top of my mind.
Do you consider yourself a data play?
We are agnostic to the data. We view ourselves as systems integrators. The difference is the math, so how do you pull together big data and use that big data to manage the television portfolio problem. TV is 80% upfront and it’s negotiated as we discussed in these big bulks on behalf of many people but then you have to allocate it at the individual impression level when you deal with digital. So how do you buy 15 to 18 months out but then allocate on a one-to-one basis in real time? That’s a tough math problem to do both and that’s the math problem we’ve solved.
The tech may be there, but what about the inventory? What’s still the biggest hindrance to selling linear TV inventory programmatically?
First, the agencies are the negotiators in the upfront, which is why the six agency holding companies are our clients – [we provide] the platform to the agencies to activate those buys. So, you’ve got to activate their trading prowess. Programmatic success in television is automating the execution of these upfront trades. Now, it turns out that we can do that no problem on the digital side. The linear feed side, the traditional side, has very many legacy systems. The cool part about the Internet is it was one standard. It didn’t care what language you spoke. If you spoke this thing called OSI 7 layers, boom, you’re good. Legacy systems (in TV) it’s the biblical Tower of Babel and we’re not really understanding each other. There are many languages in the legacy systems and this is why programmatic traditional is taking so long.
Is this an agency or programmer problem?
The media company is not used to allowing other people’s systems in to theirs. In order for automatic processing to happen, you need a bidirectional communication. That’s the issue. If you think about your house, you’ve got to be willing to let guests in. You can put parameters on the guests, where you can’t go to the upstairs and nobody’s going to the bathroom in the master suite, but you've got to let me in the house. So you have to meet the agency needs for how they trade and do business, and you’ve got to get the media companies, the broadcasters to become comfortable letting them in their house and if neither of those happens, you don’t get programmatic. And it’s hard. What we’ve done is we’ve approached both sides and given both sides tools to enable that and put them in control, and that creates the opportunity for automatic transactions.
This was one of the first upfronts where we saw NBCUniversal, for instance, talk about programmatic reserve.
I think this is the inflection moment in the industry where everybody says, “We need to do it,” so I’m not surprised. If you said when does the real traction occur, I think next year’s upfront because the complexity of meeting the agency needs and broadcaster needs is not a snap of the fingers. It’s easy to talk about. Everyone knows it’s important. It will take quarters to put the right systems and processes in place to really make it effective. This is that inflection point right here, and I think next year will literally be programmatic in the upfronts in a major way. It’s talked about this year, executed in a major way next year.
What about video DSPs that claim to do programmatic? This is still not linear television inventory, correct?
There are ways to create audience capabilities for linear TV and Simulmedia is a great example of that. To do linear programmatic, and it depends on how you define programmatic, but human interaction in the transaction is highly necessary still. It’s going to be awhile before it’s fully programmatic but you can automate components in the process, if that makes sense. We have a fully addressable TV offering you can plan against and it’s fully ingested set-top box data, but at the end of the day, you need to negotiate with DISH and Direct to get that inventory. It may be over email and phone, but they’re not just going to let you buy that stuff yet.
How long, until linear TV can be bought in a 100% automated fashion, then?
It’s not “programmatic linear feed buying the way we do online video right now.” And that’s just [monetizing] two minutes of the 16. The remaining 14 are another few years away. TV in an hour has 44 minutes of content, 16 minutes of ads. Fourteen of those 16 goes to the people who own the content, like Warner Bros. made Friends, but the distribution rights is owned by NBC. NBC gets 14 minutes. That 14 minutes is the big cherry. The two minutes go to Comcast, TNT, Dish, Direct, etc. That two minutes is like the most liquid, easy to sell business… it’s a multi-billion dollar business. Everyone’s looking to monetize those two minutes and we’re not even close to programmatic at the first two minutes. The next 14 minutes is 3-5 years.