To Fight Junk Data, Ad Industry Eyes 'Nutritional Labeling'

The data behind marketing is often wrong. Now the industry wants to fix it. The ugly truth about the data that drives marketing is that it's often simply wrong. Procter & Gamble Co.'s Gillette found that out the hard way with recent publicity about its "Welcome to Manhood" 18th birthday razors sometimes going to women 19 to 50 years old. Everyone can discover the problem in a personal way by visiting Acxiom's AboutTheData.com to review what one of marketing's big data brokers knows about them—information that's often wrong. Acxiom lets people correct the record, on the debatable theory that they actually want big data brokers to have accurate information. But that just isn't cutting it as the industry increasingly gets religion on media transparency, which includes the notion that programmatic media buying can and should be fuled by actual facts. So Acxiom President-Advertiser Solutions Rick Erwin joined heads of the Advertising Research Foundation, the Coalition for Innovation in Media Measurement and the Media Rating Council for a surprisingly well-attended two-part, 10-person panel at Advertising Week that unveiled a solution—or at least a broad vision of a solution. The industry's analytical powers that be are considering a "data nutritional labeling" system that would rate the quality of data sets, according to the panelists. But big questions remain about how this would work and who would pay for it. ARF President-CEO Scott McDonald acknowledged he doesn't have a budget for it, and mentioned GoFundMe. MRC CEO George Ivie said it would be perfectly feasible for his group to help on the auditing and accreditation for such a system, but first there has to be a system to do it. So it could be years before there's really a good way to shine a light on the dark recesses of data quality. Or maybe not. Former P&G and ARF digital executive Ted McConnell is cooking up a data nutrition-labeling plan of his own, as described in recent interviews. He's joined Lucid, which surveys much of the online world continuously with pop-up surveys—to the tune of more than 18 million questions in the past 24 hours—to fuel its business of providing qualified respondents to others for online surveys. McConnell's plan is to link Lucid's endless stream of fresh survey responses to various data sets, rating their "target density," or how many actual new-car buyers or nursing mothers, for example, they contain. It wouldn't remove all the bad data, but it might provide an advertiser-customized rating of how many useful prospects are in a database.

To Fight Junk Data, Ad Industry Eyes 'Nutritional Labeling'

Featured Stories