Family Dollar Tests Basket Analysis

In a project last year, Family Dollar worked with Applied Predictive Technologies (APT), a leading provider of test and learn management systems to retailers and consumer products companies, to understand what's happening at the basket level and found implications for in-store promotions, merchandising, and even operations.


In one test, circulars were mailed to customers of roughly one-half of Family Dollar's over 6,600 stores with roughly the other half only including in-store circulars. A small subset of the network did not receive a circular at all. The circular test ran for about 10 days beginning at month end.

Speaking at a session at the 2010 NRF's convention in January, Brian Strickland, divisional vice president of market strategy and business development research at Family Dollar, said the off-pricer suspected that circulars lifted overall basket size and the test confirmed it. The bigger benefit was learning what items particularly drove basket size. Using laundry and underwear as generic examples, it found that both generated similar sales when those items were featured in circulars but laundry had a significantly higher lift on overall basket size than underwear.

"In our fictional example, featuring laundry turned out to be twice as important as featuring underwear," said Mr. Strickland.

"What this has been able to allow us to do is really take a hard look at those baskets and understand what's happening in these situations," said Mr. Strickland. "We have customers coming in response to promotions and we're better understanding the impact on margins and obviously what is the impact on margin on that promotion. What that's allowing us to do is make more informed circular distribution decisions."

On the merchandising side, basket analysis has proven to be particularly helpful in new product introductions, particularly in understanding which products drive incremental lift across categories. Overall, it's helping Family Dollar merchants better understand which items are good sellers on their own but provide limited basket-size impact; versus good sellers that also tend to lead to larger baskets. Said Mr. Strickland, "The key is trying to get more analytics into the hands of the merchants to make better decisions."

Basket analysis also aided in optimizing productivity during extended store hours. Basket-store analysis gave Family Dollar the ability to look at what products are being sold at what time, be it day-of-the-week or time-of-day.

That helped merchants better tailor promotions but it also aided its operations team. "We were able to tell our operations team, 'Hey guys, during the weekend we need to make sure we cover these areas of the store better than we have in the past' because we know what those transactions look like on weekends versus weekdays. With the extended store hours program, one of the things that we were also able to answer was what were we selling at those early morning hours and at those evening hours and what were the margin rates on those baskets to help us make overall better decisions."

Finally, a better understanding of baskets helped Family Dollar with affinity analysis, or being able to go in and identify items that sell well together. For instance, the test found that buyers of laundry were 14 times more likely to buy fabric softener. "That's an item that wasn't featured in the promotion but it's obviously had a huge impact."