Behavioral Insider: In the addressable TV world so far, in trials and deployments, what kind of targeting is available now?
Mitch Oscar: There are two categories that seem to really know about targeting that we haven't really dealt with much. One is direct mail. Don't they know more than anyone? And direct response.
Two companies have been very active in this from the data mining perspective, Acxiom and Experian. Acxiom has been doing a lot of work with Navic, and I have been working with them. We deployed in Oceanic and Charter where we actually look at the number of digital cable subscribers and look at lifestyle about each one, but not knowing who they are. A client can say they are looking for people who have three heads, people who have four legs and 27 kids, and from that these companies can determine if they have the information from the cable operator. All privacy is protected so they won't say the name but look at life stages. If a hundred people over 20 have three heads, and 40 have 27 kids, then we can target to them. We have been doing this. Experian is doing the same kind of thing so that the advertisers involved can match up a number of their commercials or clients to the right households. [with Comcast, Cablevision and Visible World].
BI: How would addressable TV advertising re-orient the way buyers think about TV inventory?
Oscar: Now we are starting to mix and match. So we are no longer thinking, should we buy by a network -- but we should by a household. An advertising agency could purchase inventory for a bunch of clients and then mix and match brands. Households that have girls maybe they should get lipstick. Invidi and its technology suggests they know who is watching the TV set from some fuzzy mathematics, so it doesn't really matter the program. When the TV is on, that person -- that demographic -- will be there.
You have Navic on the other hand, in their ad auditioning model, saying we can look at historical viewing for the set top box. Rather than buy by network or day part or program on on ESPN, I could say what about how-to shows? I think my product works best in how-to shows and they can say that these set top boxes in these markets actually watch these how-to shows in general and they also seem to watch these programs. So now we are creating a vertical and a horizontal. We make the ad more relevant but also generate efficiencies on buying.
BI Behavioral targeting online always struggles with achieving sufficient scale on its segments. Will we encounter that when BT comes to TV?
Oscar: The TV business is at $70 billion. Online video is $1.3 billion. We are spending a lot of money in TV. If the advertiser is spending that money anyway and we offer more enhancements in interactivity or targeting, it may mean we counter the flow [to Internet] or the slow growth.
In terms of behavioral targeting, on TV we're trying o figure out lifestyle just generally -- all the things that Acxiom and Experian are allowed to know without naming names. We still don't' know who is watching. I don't know if it is the same, and it will never be as granular. There will always be scale in the expenditure towards TV.
When I started my career, I ran Pepsi buying. We would do an up front and spend $10 million or $20 million and have four brands and do an index. Which is the younger-skewing show, or older? That was as granular as we got. But now we can take a look at four brands and look at all the networks we have and figure out which households work best, so every brand might get the same ESPN spot but it will get it in a different household.
So I think that the granularity will never be so miniscule that someone will ask, is it worth it. People still think TV is worth it. It could never be three people. If a million people are watching on cable then that is a million based on contextual, because they are watching 'Monk' or a sport or a how-to. If we divide it by three different targets, you still have over 300,000. That isn't bad since you are spending branding money anyway. I don't think the numbers will be so small.
BI: How soon before we see these techniques at a larger scale?
Oscar: Project Canoe, from what I hear, is going to take a couple of years to connect markets and technology, so for a couple of years out it is interesting. I hear rumors that EchoStar and DirecTV would merge in some form. Then there are 30 million homes that can have some form of interactivity. The cable operators now have 54 million digital cable homes out of the 65 million in total. They are goosed to do it. We have more TV sets coming out with Internet connections, which will have interactivity. And most important to me -- if everyone is learning about interactivity online anyways -- as soon as it is easy for TV, then people will naturally gravitate towards that as well. It won't be a different training process.