Market basket analysis and its offshoot, affinity marketing, continue to follow the usual c-store industry adoption pattern: cautious expansion, mostly among the bigger chains.
But the potential payoffs could be enormous.
That said, like most new technologies market basket requires commitment at the top, beginning with more expensive software matched with an appropriate point-of-sale (POS) system as well as an employee to sift through and analyze the mountains of collected data and both the willingness and expertise to use the findings effectively.
First, some definitions: Market basket analysis is the capture and study of detailed data about a purchase transaction. According to Casey McKenzie, senior partnering consultant of the Impact 21 Group in Lexington, Ky., and Todd Roberge, pricebook manager for WilcoHess LLC in Winston-Salem, N.C., both of whom spoke on the subject of using scan data wisely at May’s NACStech confab, employing market basket analysis helps operators answer questions such as:
• Do some items generate a greater market basket transaction than others?
• Do some items, when properly promoted, generate significantly higher sales and gross profit?
• Are consumers "cherry picking" promoted items?
• Do promoted items perform better in different locations due to improved adjacencies?
Affinity analysis is a method for analyzing item relationships within a purchase transaction, with the objectives of influencing customer behavior and improving merchandising and promotions.
Together, they promise to focus marketing, product and promotional strategies as never before and strengthen bottom lines. The question, as always, is what it takes to get there—and, once there, will it have been worth the trip?
"Definitely a Discipline"
Market basket analysis "is definitely a discipline in and of itself," said Lesley Douglas Saitta, CEO of the Impact 21 Group, an alliance of retail consultants with broad convenience store management experience.
Operators sometimes err in thinking that market basket analysis can be their "be all and end all," said John Russo, general manager for RedPrairie Performance Management in Waukesha, Wis. "They may have an expectation like, ‘As soon as I know that then that’s half the battle.’ But actually coming up with a promotion that is interesting to that customer is something you have to work on, and it’s going to be based on the way your company operates and the kind of product offering you have."
Business Intelligence
"We have (market basket analysis) as part of our business intelligence suite," said Brad Buckmaster, IT manager for 99-unit Plaid Pantry in Beaverton, Ore. The question, he said, is "How much are we using it? I know we look at the (product) movement, but we also have a priori analysis" which lets the operator prioritize product categories such as beer, cigarettes or chips. "Then you can get more specific and say, ‘If it’s a beer item, what’s the item most likely to sell with it?’"
Plaid Pantry put its system together two and a half years ago with a software firm known primarily for its work in the movie industry. Operating its own data warehouse prior to that "took a lot of maintenance on my part," Buckmaster said. The system he’d been using had proved "pretty unwieldy. It took forever to run any sort of analysis on it at all. Finding them to maintain a lot of it was very helpful for me."
Plaid Pantry’s affinity marketing efforts have proven helpful "for every department in the company," Buckmaster insisted. "In fact, we’ve found that our operations people no longer have to do a lot of the data collecting they’d been doing. Now they can spend their time managing however many stores they have rather than gathering data from those stores and not really analyzing it."
Coffee with Pizza?
"People have been doing market basket analysis solutions for years," RedPrairie’s Russo pointed out. "I have been in the IT industry around retail since 1985, and I recall as long ago as the early 1990s people marketing data-warehouse solutions that would analyze market basket details."
In recent years, however, Russo explained, advancements in database technology "have allowed for better analysis for c-store companies that are doing massive volumes of transactional data."
The RedPrairie Performance Management solution delivers a holistic view of operational performance across the entire enterprise via an easy-to-understand dashboard, empowering users to identify trends in sales, product costs and inventory costs, compare similar stores and evaluate performance against established budgets and targets. It allows operators to select any category of products and chart how the products sell.
This allows retailers to analyze how many of the different transactions are single-item or single-category transactions. From there, chains can slice sales data down and begin looking at it geographically or by daypart. One fairly counter-intuitive connection Red Prairie turned up for one client was the fact that coffee was its top-selling beverage when consumers bought pizza at dinnertime.
Doing More with Scan Data
Pat Zelechoski, pricebook manager for NOCO Express in Tonawanda, N.Y., which operates 32 c-stores in Buffalo and Rochester, said her company isn’t yet doing market basket analysis because its POS system is "not able to get that information together."
Executives have, however, "been looking at a few different POS companies that do that type of analysis. It’s one of those things where, what we have now works but it doesn’t give us everything we’d like to have."
NOCO management is doing "more and more" with its scan data, according to Zelechoski. "Right now, we’re trying to standardize our products more. We’re taking items that are in the planogram for a section and trying to measure when something has run its course. If a product was hot, when is it time to take it out? We haven’t really done that before. Now, what’s new, what’s coming and what’s better is our focus."
While she acknowledged the value in market basket analysis, Zelechoski is quick to add that its true value "depends, like anything, what you do with it. We get a lot of scan data now but the question is, how much of it are you using and what are you using it for?"
That answer, she said, "would depend on what areas you’re working on. If you’re struggling with your coffee sales, let’s say, that might be something you focus on to see where you can improve, where you can get that extra foodservice sale. But again, you have to use it. That’s the big thing."
The Aha! Factor
"We evaluate our promotions based on whether that product or group of products sold more than they did before, or more than we expected," said Impact 21’s Saitta, a former vice president of marketing for Speedway SuperAmerica LLC. "If we are looking at how valuable a promotion really is you can’t just look at the item being promoted. You need to look at products that were adjacent to it, products that are complementary to it. If we have a cigarette customer, and we know that cigarettes are declining, what else are cigarette buyers purchasing that I could promote?
"When operators are actually getting more per transaction by offering complementary items I know are selling, that’s what I call the icing on the cake."
What Saitta described as "the real challenge to my retailer clients is having someone to analyze it, which is always the problem. Even getting the data itself in a form that can be used is a challenge. You still have to determine, for example, what stores are going to pull data from. If it’s all my stores, do I have a capacity? Can I query it in a way that brings it to me in a way that’s usable?"
"Market basket yields a lot of data. If you thought scan data was a lot, just wait until you bring up every transaction throughout the day for 250 stores."
As always, it comes down to commitment.
"If it’s something they’re going to invest money in they definitely have to use it," Zelechoski warns. "Everybody is scanning, but what are they doing with the information? I’ve gone to a lot of classes and people say (about scan data) things like, ‘Wow, we never thought of doing that,’ or ‘we don’t have time to do that.’ There is some value in (market basket analysis), but if they’re going to invest the money they really need to use the tool and work at it and get out of it what they’ve put into it."