The deal, which is not exclusive, is the latest in progression of online targeting firms that are incorporating offline databases and segmentation systems that correlate online users to offline consumer behavior.
"We've built a bridge that will start a new phase of online targeting," boasts Justin Evans, senior vice president-marketing and strategy at Nielsen Co., who struck the PRIZM deal with DataLogix. "The way marketers manage consumers in the offline world can be ported online."
The progression is ironic, because endemic online advertisers and media planners have always prided themselves on the ability to target users based on their actual behavior, albeit those exhibited online, via behavioral and contextual targeting techniques. By adopting systems like PRIZM, they are effectively utilizing surrogate data to define the behavior of online users, clustering them into cute-sounding lifestyle descriptors such as "Blue Blood Estates," "Young Digerati," and "Country Squires" that might make them the ideal targets for certain products or categories.
The method used by PRIZM, known as geo-demographic segmentation, has been popular among offline marketers and agencies for decades, because it enabled them to fine-tune their targeting based on the likelihood that specific households would fall into desirable consumer lifestyle groupings based on their geography. The PRIZM system divides American households into 66 segments based on their neighborhoods and can get as granular as zip codes-plus six, or even household level.
Nielsen's Evans says this is an important "bridge" between the online and offline worlds, because it will enable marketers and agencies that might be less familiar and comfortable with online targeting techniques to utilize the same ones they use to plan offline media such as TV, radio, print, outdoor and direct mail media.
Ideally, he said, media planners will utilize the PRIZM data to "complement" behaviorally-based online targeting data to gain more of a "360-degree view of the consumer."
Bringing PRIZM to Online Consumer Targeting sounds a little bit like bringing "a pair of tin cans attached to a string" to modern telecommunications, in that PRIZM assumes homogeneity of CBGs (census block groups are clusters of continuous block faces like the last four digits in a 9-digit zipcode). That's not always the case, by a long shot. I'll wait with baited breath as we see how they assign cluster codes to individual online consumers, though I am pre-conditioned to be skeptical.
Reverse IP Mapping is anywhere from 25-50 miles accurate. Geo targeting below the DMA level is less than 20% accurate and registration data doesn't scale. Someone needs to take a closer look at how companies are delivering accuracy at the neighborhood or, as Nielsen claims, the household level.
Using an IP Address for geolocaiton is increasingly more suspect for privacy reasons...and already illegal in some counties. Applying PRIZM clusters or any targeting using Reverse IP Mapping means that entire cities (or in my case several towns) are lumped together as one. Something doesn't seem right.