The data from several unrelated retail sensor projects—all of which track consumers electronically during their shopping experience—will be hitting IT desks soon, as no fewer than three such efforts are going to be announced by September.
Retail marketers are notorious for arranging for a ton of data to be collected, only to have the task of making sense of it fall on IT. In this case, marketers and their love of dollars and sensors have outdone themselves, with today's CRM data of what consumers actually purchase becoming a rounding error when compared with the sensor data tracking every move and every potential move.
The most public effort—announced almost two years ago with the support of Wal-Mart, Kroger's, Walgreen's, Procter & Gamble, Coca Cola, Disney, Kellogg's and Miller Brewing—has been publicly quiet, despite running a major trial last year.
That trial—which ran from April to December 2007—ran in select locations of A&P, Ahold USA, Catapult Marketing, Clorox, Coca-Cola, ConAgra Foods, General Mills, Group M, Hewlett-Packard, Hy-Vee, Integer, Kraft, Kroger, MARS Advertising, Mars Snackfood, Mattel, Meijer, Miller Brewing, Nintendo, OMD, Procter & Gamble, Pathmark, Price Chopper, Rite Aid, Safeway, Schnucks, Sears/Kmart, Starcom MediaVest, Stop & Shop, Supervalu, Target, Unilever, Walgreens, Wal-Mart and Winn-Dixie.
Even though the trial closed last year, the company in charge (The Nielsen Company) doesn't plan on releasing the results for several weeks, targeting August or September, said Peter Hoyt, executive director of In-Store Marketing, a company coordinating that trial. (See a humorously unintended impact of the trial, involving mysterious chicken babies.)
Hoyt described the way that trial examined purchasers and their purchases: "At the entry-way, (customers) would have walked through infrared beams and been counted on the way in and counted on the way out," Hoyt said. "And at different times of the day, Nielsen auditors swooped into these stores all across the country and counted the number of people in every section of the store and did correlations of those counts with the infrared beams so they could verify the audit information."
Nielsen is far from alone. On July 14, retail research firm TNS rolled out its Insight Dashboard, which uses similar sensor techniques to track shoppers before they "enter a store, as they walk around a store, at the point of purchase and after they have selected a product for purchase."
And early next month, another company will introduce a similar effort, except this one won't rely on stationary sensors to watch consumers as they walk past.
That August effort will use sensor-laden mats—to track product movement on the shelf—as well as sensors embedded in shopping carts and baskets.
As these sensors become widely populated, the potential for a sharp increase in the amount of data available for analysis—both in terms of CRM plus anonymous aggregated data—is enormous.
The open questions are how the output will be formatted and whether the formats will be consistent and interchangeable.
But there is a bigger question, which is a CRM holdover. Many retailers don't use—in the sense of analyzing, distributing and having managers use the information to make decisions—much of the data that is already being collected.
With the potential to deliver far more TBytes of data than many CRM projects, will this new data be leveraged any better?
The knowledge potential is certainly huge, with an intelligent approach to such data potentially able to yield a stunningly detailed and accurate picture of what consumers are truly doing.
But beyond the logistical obstacle to making use of such data lies the consumer privacy concerns. Will such efforts alienate customers more than anything else?
Information last week about a Japanese government-sponsored effort to extract consumer data from music players, cell phones and televisions has already raised consumer suspicions.
But whether retail IT execs are ready or not, the sensor data is about to start coming. One can only hope that retail CIOs won't make like the sensor beams and start bouncing off the walls. |