DirectBuy takes different approach to business intelligence

Membership buying club DirectBuy Inc. takes an unusual approach to retailing. It offers paid members the chance to directly purchase goods from manufacturers and wholesalers, without paying retail markup



So perhaps it is to be expected that 50-plus-store DirectBuy, which also offers online and in-home shopping options to its 250,000-plus members, approaches business intelligence (BI) a little differently, as well.



“The traditional way of looking at BI is as a descriptive and diagnostic tool for operational decision-making,” said Armin Roeseler, CIO of DirectBuy, Merrillville, Indiana. “Year-over-year reports are generated.”



However, DirectBuy applies BI in a predictive and prescriptive manner that has enabled the company to transform how it operates key aspects of its customer engagement strategy.



“We didn’t know how to spend marketing dollars,” said Roeseler. “We had six marketing channels, but didn’t know in what channels adding or removing funding would have maximum positive impact on the full-line business.”



DirectBuy applied BI to analyze its main marketing channels of web display, search engine optimization (SEO), TV, direct mail, email and pay per click. Advanced BI analysis uncovered some surprising insights the retailer would not likely have discovered with traditional analytical tools.



“We identified SEO as having a negative impact, which ran counter to conventional wisdom,” explained Roeseler. “It created a disturbance that was felt throughout the other channels and led to fewer leads.”



Also flying in the face of current mainstream business thinking was DirectBuy’s discovery that web display ads such as banners, which most marketers will tell you have devolved into background noise, were actually far and away its most successful customer engagement channel.



“Display ads were 10 times more effective than any other channel,” said Roeseler. “We also found our pay-per-click channels weren’t negative, but were not very effective. BI brings management science, such as using industry best practices and what you know, together with technical science to fine-tune campaigns.”



DirectBuy also applied BI to analyze its membership retention strategy. The company typically renews memberships on an annual basis, and has about 18,000 renewal dates every month. While the ultimate goal is to have no lapses, DirectBuy lacks the resources to perform personalized outreach to every member up for renewal.



“The marketing team suspected some members were more likely to renew and wouldn’t need a personalized call,” said Roeseler. “Using a decision tree model with BI, we ran 15 years’ worth of renewal data on more than one million renewals.”



DirectBuy found that members who had made at least one transaction in the past year or had at least a three-year history of consecutive renewals were almost certain to renew. This left a pool of about 3,000 customers per month on average that marketers could target with personalized renewal offers and promotions.



DirectBuy runs a Microsoft technology stack and BI tools. While Roeseler said the Microsoft platform works well, he advised the specific solutions a retailer uses are less important than the underlying data architecture.



“We have a single source of truth for all operational data,” said Roeseler. “All stakeholders have a universal data architecture. It’s critical to have a good data dictionary.”