In this first of two parts on Consumer Direct, we ask Warshaw how the technique has evolved and the varying sensitivities consumer have over tracking offline versus online activity. Next week, we speak with Yahoo about how its own use of the data has evolved into a better targeting tool.

Behavioral Insider: Walk us through the basic of Homescan. How does it relate to the Consumer Direct product?

Warshaw: Homescan is a panel that collects everything they have purchased using a wand scanner on anything with a UPC code. The full panel totals about 120,000. We have a subset of that panel who agree through an opt-in cookie match to allow us to track their surfing behavior on Yahoo. It is done through our panel, so we manage that. About 45,000 from the panel opt-in.

Behavioral Insider: What kind of data do you get from this cross-platform analysis? Is it a raw data dump to Yahoo or is it shaped and segmented on your end?

Warshaw: We shape it. A client will come to us or through Yahoo and say they want to run an ad on Yahoo and target heavy soft-drink buyers, for instance. So we look at the 45,000 households which exhibit the heavy soft-drink buying behavior, and then we look at their Yahoo surfing patterns. We model that and with algorithms provided by Yahoo, we provide them with an algorithm that tells them which are the ones in their universe that would be the heaviest soft-drink buys. Then Yahoo serves the ads to those people.

Behavioral Insider: But on the back end, doesn't it also helps measure results?

Warshaw: Right. Once the program is over. Yahoo sends us a file back that includes those households in our panel that were exposed. And we can do our pre/post analysis looking at exposure and non-exposure.

Behavioral Insider: How has the system changed in the last four years? What has been the key learning about this kind of research?

Warshaw: A few things. We have done a better job over time of targeting the households and getting the right people in the mix. We have done some learning on the backend to find the types of campaigns that work better than others. Rich media campaigns uniformly seem to do better. The product itself has not changed a great deal. The sample size has gotten larger over time, which has enabled us to run and measure campaigns against smaller brands.

Behavioral Insider: Any changes in the product that you anticipate?

Warshaw: We are trying to look at sample size growth to identify new surfing behaviors, to start marrying the results with other general purchase behavior measures, such as brand loyalties, shifting between brands, etc. We want to give the marketers a better feel for the overall impact beyond the ROI.

Behavioral Insider: Privacy is a constant concern in the BT field, but your approach gets closer to consumers than just about any other. What have you learned about consumer attitudes toward intrusiveness?

Warshaw: There are different levels. Out of the 125,000 or so that are in the full panel, 45,000 have opted in. If you take out 30% of [un-wired] people who wouldn't have the opportunity to opt-in, we end up with rates of about 50%. That means half the people already won't participate. If you ask the households why they are participating, you typically [hear] that they are part of a process. They feel they are getting their values and their input heard.

Behavioral Insider: Is there a special level of sensitivity to tracking online behaviors that isn't there for scanning consumables?

Warshaw: I think there is. We have another opt-in panel for a different service, which is not restricted to Yahoo. It is not an ad-effectiveness tool as much as a tracking vehicle: Here is everything they buy, and here is where they surf. It requires a software download from a sister company. And the opt-in rates are less than half of what they are for Consumer Direct. There is a difference between knowing you are being watched per se on Yahoo. But when you tell them you are looking at everything they are doing on the Internet, the opt-in rates go down. It's a little more Big Brother-ish. We are very careful not to overdo it, not to ask for 14 different opt-ins. We feel that Yahoo is relatively safe in terms of opt-in.

Behavioral Insider: How is tracking like Consumer Direct helping CPGs feel more comfortable about using the online channel?

Warshaw: When you go back six seven years, when CPGs first got involved in the Internet, it was all about selling product. And we developed our Homescan online capability, where you just download the meter. We were talking about how people would be able to surf and purchase. The clients were all saying this would be great -- they could sell products online. And it was a disaster. How do you sell bathroom tissue online? And a lot of those companies folded up their Internet groups. It has taken them up until a couple of years ago to start thinking about the Internet for any purpose. So that was a long cut. The CPGs were probably the last medium to embrace it as an advertising media based on this experience eight years ago.

This commentary is insightful. I recommend it to others.