Breaking Down Big Data
NACS Show: Retailers give information-analysis concept meaning
ATLANTA -- While the concept of "big data" may need to find more widespread acceptance within the c-store channel, some retailers are figuring out ways to use huge amounts of transactional and sales information to predict consumer behavior and to plan future projects.
Speaking before about 200 attendees at the 2013 NACS Show in Atlanta, two retailers and a data-analytics consultant fleshed out the promise and the practicality of so-called "big data" uses within the channel.
Initially setting up retailer priorities, session moderator Patrick O'Reilly, president and COO for Applied Predictive Technologies (APT), Arlington, Va., suggested retailers set out to conduct "rapid experiments" designed to utilize large amounts of data to prove the validity of certain hypothesis--ranging from the positive effects of raising coffee prices to the repositioning of merchandise on the sales floor.
"The problem is that c-stores are a noisy environment," O'Reilly said. "You're trying to find a 1% comp gain--which is good for an industry [comparison]--but that's amid a 20% plus or minus [average shift]."
That said, O'Reilly believed that analytical steps using the proper number of control and test stores could mitigate the shifts and lead to concrete results, answers that can validate a retailer's gut instinct or persuade him or her to pursue another direction.
The "first priority" types of data O'Reilly suggested retailers focus on were the following:
- Weekly sales data.
- Transactional logs.
- Customer loyalty data.
- Competitive fuel pricing.
When working with large amounts of data and the ability to create formulas that would provide meaningful insight, Joe Venezia, senior vice president of operations for The Pantry, Cary, N.C., said the results provided insight in at least three areas: the total impact of the program, the development of a predictive model and how to use transactional data.
Ultimately, Venezia described two ways his chain recently used "big data." One was with determining where to build new quick-serve restaurant (QSR). He said a good number of the chain's 1,550 c-stores already had proprietary QSRs in them, but executives saw opportunity for the rest of the network. Inevitably, the company came up with the type of projections that would identify potential opportunities for new QSRs.
Explaining yet another endeavor, Venezia spoke of deciding which store in which to extend hours. While new profits were a motivator, some stores would not produce the added revenue to make up for additional labor and operating costs. "Big data" helped The Pantry address that issue as well.
Speaking from another perspective, Scot Tomlinson, systems engineer, Mid-Atlantic Convenience Stores (MACS), Richmond, Va., said they used big data to analyze market-basket data to grow sales.