Targeting and ROI

The Future of TV Series

This is part two of a three-part series where Andy Fisher of Merkle sits down with Bill Harvey, cofounder of TiVo Research (formerly TRA) to discuss the state of the evolving television industry.

Andy: We have discussed two things – one is the ability to do better targeting, and a lot of times that's custom targeting based on the individual advertiser. The second thing is to be able to do a guarantee against ROI. So my question is how should that work from a business perspective? Who should be measuring and validating that that better targeting happened? And who should be measuring and validating that the lift in ROI happened? Is there a role for the MRC here? Or is there a role for a third party here? How should that all work to make the business flow most efficiently?

Bill: It’s happened without coordination of course. The state of the art right now is TiVo Research; for a number of advertisers it’s the fact checker who comes back and says, "This is the ROI you got." Prior to TiVo Research, it was marketing mix modeling. And even to this day, marketing mix modeling in 99 percent of cases is the final arbiter of “did you get a higher ROI?” Then who's doing the marketing mix modeling? For some advertisers, they have in-house people doing it. Other advertisers are using, for example, anyone from Marketshare Partners to Nielsen, to dozens of other companies who are having their agency do it.

And then you get in to the fox guarding the hen house situation. What I always liked to do at TiVo Research was to have a double check where we would show based on our own data that we got you a lift in ROI. But I would always want them to independently verify it with their own data set. I called that "bank account validation." Tell us if by your own metrics the sales went up and how much they went up, and let's compare that with what our data shows. Mars-Wrigley did that with us at the ARF. Kraft did that with us at the ARF. A number of companies stood up and they said based on their own data, "You did get us an ROI increase."

There was another interesting case where a company used Nielsen marketing mix modeling to measure whether TiVo Research got an ROI. We were nervous about that, being judged by Nielsen who was already starting to build their own Nielsen Catalina emulation of TiVo Research. But nevertheless, Nielsen's marketing mix modeling people showed that the ROI did go up.

Andy: This world that you're describing of mathematical modeling, big data, and first-party customer data on the advertising side, and set top box data – do you think the advertisers, agencies, and publishers are really prepared for this world? And if not, what do they need to do?

Bill: They're trying to get prepared. The advertisers in particular are trying to get prepared. Some of them are taking functions in-house, trading desks for example. Many advertisers have already built their own trading desks or acquired trading desks from third-party tech companies. Some advertisers have taken some creative functions in-house. There's a general hovering dark cloud of threat over the agency business that advertisers are looking at this in-house option.

Other advertisers are taking more of a classical approach where they're building their own DMP and coordinating what they build with their agencies to make sure that A: the systems talk to each other, and B: there's an API connection between the systems (or a plan for one): Machine-to-machine communication. And a clear division of who's going to do what so that there isn't duplication, and there's the best end result. So I think they're trying to get ready. Everybody's trying to get ready. Everyone I talk to has got something going on in the way of system building.

Andy: You mentioned a dark cloud hanging over the agencies. So what do you think the agencies need to do to move through that dark cloud?

Bill: I just learned about 20 minutes ago from you that you have a way of putting it, which resonates with my feeling – agencies should become the integrators of the data from the advertisers and the publishers. I think that's a beautiful way of putting it. Prior to hearing you say that my thought was I'd like to see a return to full-service agencies. I think that the idea of splitting the media and creative functions or splitting into 12 different siloed functions is so ironically and obviously the opposite of integrated marketing communications.

It makes me laugh to think that in an era where we talk about integrating, we're disintegrated. So I have a personal preference for the agencies going back to full service. I also think the business model should have an ROI bonus component, that the agencies should have an up-side. Maybe MRC or ARF or some impeccable third-party measure of ROI. That would be a firm basis for agencies to be able to collect on those ROI bonuses.

Andy: Gotcha. And so I'll question about the silos that you just brought up. One of the things that I've seen in this industry is a little bit of a debate about whether or not something like addressable TV should be bought, and executed, and optimized by TV people who understand television, or it should be bought and optimized by people who are digital and understand addressability. So which silo do you think that should go in, assuming of course breaking? You certainly can answer that by saying, "All the silos need to be broken."

Bill: Today we have addressable TV in 42 million US homes. About 35% of the population can do addressable commercials. And I think the buying, selling, and optimizing of addressable commercials shouldn't be done by TV people alone. I don't think you can get a maximized ROI result by having some people buy addressable TV, some people buy non-addressable TV, and some people buy digital, social, mobile. The same folks that'll be planning and buying all of this stuff and the systems should support the integration of all of this stuff.

Right now I've got a new company RMT that's building the kind of optimizer that can put together addressable, non-addressable, digital, social, mobile, etc., and all the data, and capabilities of a TiVo Research with other capabilities of a Merkle, let's say, and maximize ROI or if it's a tune-in client, maximize rating results considering all of the nine factors including creative.

For more insights on the state of the television industry view the next video post in the Future of TV Series, “Big Bold Predictions For the Television Industry.”

Join the Discussion