Skip to main content

Focus on: Business Analytics

4/19/2010

When it comes to information technology, Family Dollar Stores is always open to testing new projects and systems that can add efficiencies and improve its operations. The Matthews, N.C.-based discounter is helped in this regard by an internal process, called “Test and Learn,” that it deploys to understand which projects are the most viable to pursue long term and enterprise-wide. The process helps keep the chain open to innovation, while ensuring its capital spending is used to maximum advantage.

“Family Dollar’s Test and Learn process is all about asking questions,” explained Brian Strickland, division VP market strategy and business development research, Family Dollar Stores, which operates some 6,690 stores and reported sales of $7.4 billion for fiscal 2009. “The answers help us determine if a specific program or project is financially viable for us to look at.”

The Test and Learn process has two important factors. The first is Family Dollar’s relationship with its five-year partner, Applied Predictive Technologies. The chain relies on the Arlington, Va.-based business intelligence technology provider to help it expertly analyze and evaluate the potential and viability of initiatives across multiple divisions within its vast enterprise. Positive results have helped the chain launch successful projects company-wide.

The second component is Family Dollar’s use of the APT technology across test and control stores, which are locations where the chain conducts and measures all concepts. The amount of stores grouped into each category tends to vary depending on projects, but it is not about the number of stores.

“We do not compare five stores in one group with five stores in the other group,” Strickland explained. “Rather, we weigh results of one or two stores to ‘X’ number of control stores. This methodology allows us to better understand financial performance prior to an event and segment results following the test.”

The chain is aware that many companies conduct tests by comparing group A’s performance with group B, “but this doesn’t provide segmented results,” according to Strickland.

“Unless you can compare one store with many,” he said, “it is difficult to determine how well a new process is performing in a specific market or situation.”

Family Dollar has successfully applied its Test and Learn methodology in such departments as loss prevention, facilities, merchandising and marketing. When it comes to marketing programs however, performance testing gets a little dicey. It is not uncommon for some chains to struggle when pursuing and evaluating marketing campaigns since they can often be difficult to measure. APT is helping to ease this process with its market basket analysis tool.

Putting an emphasis on transaction log data, the tool’s algorithms uncover SKU adjacencies, product bundling and the influence of promotions. The solution also helps Family Dollar measure the impact of basket size, related sales margins, basket composition, shopper visits and purchase frequency.

With thousands of stores located across the nation and no loyalty program, Family Dollar uses the APT tool, combined with the Test and Learn process, to understand whether it can target promotions on a market-, store-, even customer segment-level.

“Overall, we are able to determine sales lift and learn what products are supporting that lift,” Strickland said. “It puts data in the hands of more people, and this helps us make better decisions on promotions, and even learn how to introduce new products into our assortments.”

Although Family Dollar had been measuring marketing initiatives prior to deploying market basket analysis, the data set was not as robust, and it didn’t encompass results from all stores. Instead, the retailer focused purely on store groups.

Since adding the solution approximately six months ago, “we are using a more robust, extensive data set that encompasses data for our entire enterprise,” Strickland reported. “It is great to have access to so much data at our fingertips.”

The solution came in handy, for example, when the chain was interested in learning the impact of its monthly store circulars on sales. During one test, the chain monitored sales information based on circulars distributed to 50% of the test stores, which were scattered across various markets. The other half of the group did not receive the promotional materials. “We were able to analyze individual baskets and uncover purchase patterns, as well as specific items that provided significant sales lifts and drove larger basket sizes,” Strickland said. “We were also able to correlate when a promoted item drove additional sales of items that may not have been promoted that week.”

Not surprisingly, the process is being applied to new areas. Among the newest projects it is supporting is the viability of the e-mail distribution of the chain’s monthly circulars and store-level promotions. Family Dollar has been inviting shoppers to visit its Web site, familydollar.com , to opt-in to receive e-mail messages. The reception has been positive.

“Our e-mail list continues to grow daily,” said Josh Braverman, public relations manager, Family Dollar. “We are also sending out messages twice a month, on average.”

Family Dollar’s e-mail promotions run the gamut from holiday-based campaigns to special events, such as the Super Bowl. The messages are also segmented to specific customer groups who receive special promotions and opportunities to take advantage of at store-level. The analysis tool is helping the chain determine which promotions are providing the most value. It can determine this by analyzing specific promoted events.

For example, during the 2010 Super Bowl, Family Dollar distributed e-mail messages highlighting specific snacks. The tool helped the chain determine the value of these items versus those only promoted in the traditional circular.

“This proves that we are using Test and Learn to ask good questions,” Strickland said. “The answers are clearly helping the chain make better business decisions—ones that are improving our business as a whole.”

X
This ad will auto-close in 10 seconds