Retailers that integrate customer data from multiple sources often misinterpret it and produce faulty analysis, causing confusion in subsequent attempts to serve customers, says analyst Jim Okamura, an author of a new multi-channel retail study, "Retail Details: Best Practices in Multi-Channel Integration."
"Retailers are increasingly integrating external data (like census figures) with their internal databases to get richer insight," the study says. "However, when not used correctly, the proliferation of data sources available to a retailer can lead to data overload with important findings getting lost amid the sea of numbers." The study was produced by consultants J.C. Williams Group, where Okamura is a senior partner specializing in multi-channel retailing, for DoubleClick Inc. and its consumer transaction data division, Abacus. The study was designed to gain insight on multi-channel integration strategies from 36 executives, analysts and academics involved with the retail industry.
"Trying to put together a reasonable plan to get smarter use of data is often a challenge," he says, noting that several executives involved in the study noted that retailers are known to take on more data than they can handle for effective analysis.
Okamura notes that retailers may combine external data such as government information on consumer expenditures with their own data on their customers’ purchasing history, but without making sure that the two sources of data can be successfully integrated. "In the case of looking at consumer expenditure data from the government, for example, how the government classifies certain expenditures may not match up with a retailer’s internal definitions. But having that match up correctly is a starting point for apples-to-apples comparisons."
In addition, he says, retailers often need to do more preparation in how they plan to use multiple sources of data. "You have to take into account what all the data mean to the business issues you’re dealing with," he says, adding that retailers also need to make sure that data analysis projects involve the heads of all pertinent departments, including marketing and merchandising.