Map It Out
CSP Exclusive: Heat-map study shows fresh c-store opportunities and frustrations.
Everyone suspected the time consumers shop in a c-store was small, but now we have a number: 1.68 minutes.
“People making a purchase in c-stores are extremely mission-oriented,” says Priya Baboo, executive vice president of shopper insights and strategy for VideoMining Corp., State College, Pa. “You have to think out of the box to capture their attention.”
Baboo has a powerful perspective to support her point. During six weeks in August and September 2010, VideoMining conducted the second in a series of heat-map studies designed to quantify and analyze c-store shopper interactions. The purpose? To reveal new opportunities for retailers to grow incremental sales.
As part of this heat-map study—a method of linking foot traffi c to actual transactions across all day-parts—five of the channel’s largest chains agreed to participate: 7-Eleven, Circle K, Sheetz, Cumberland Farms and BP’s ampm. In all, 48 stores were included in the study, with 12 outfi tted with ceiling-mounted video sensors to track customers’ movement in the store. Wall-mounted servers in the stores’ back rooms processed the video in real time with VideoMining’s proprietary software. Analysts then reviewed the data to develop shopper insights. All of the retail locations provided point-of-sale (POS) data, with some undergoing manual audits and customer intercepts.
Michael Burkenbine, marketing programs specialist for BP’s ampm brand, had seen the results of heat-map studies before yet hadn’t felt compelled to participate. But in this situation, BP was intrigued by the study’s ability to cross-index the heat maps with actual transaction data from stores. “That was motivation for us to say this could be really interesting,” he says. “A heat map is interesting if people pause, but if they don’t buy, what’s the point?”
In the space of an afternoon, VideoMining installed more than 30 cameras into the dropped ceiling of three franchised ampm sites: a traditional West Coast location; a renovated site in the Midwest; and a new, larger store on the East Coast. Meanwhile, BP supplied approximately 5.4 terabytes of transaction data to the company.
For Burkenbine’s team, the heat map showed parts of his stores where foot traffi c was high but sales remained stubbornly low. For example, in the gum section, video confi rmed many customers stopping to review a variety of offerings. Yet a disappointing percentage actually made a purchase. Across all the chains, gum was one of the categories where the conversion from “shopper” to “buyer” was low—or only 18%, according to VideoMining. Based on the video footage, Burkenbine suspects ampm’s gum assortment does not sufficiently reflect the store’s demographic.
All of the retail chains have kept the cameras inside their sites so that they can continue to study customer behavior longitudinally, while vendors are homing in to their particular categories.
“Heat maps help us target … the key locations for certain categories,” says Mark Krull, a former manager of c-store category development for The Hershey Co., Hershey, Pa., who now oversees the grocery channel, citing in particular the best adjacencies for candy. “It’s all about … understanding what the destination and impulse categories are, and making sure the store is laid out in a way to capitalize on different aspects of each category.”
For Hershey, the research confirmed that candy needs to be at the counter, with its home location on the path to the cooler doors. Real estate in the foodservice section—which the study reaffirms as a destination area—is just as valuable.
“Only a handful of categories are really popular in a c-store,” says Baboo of VideoMining. “If that’s the case, it’s all about making categories more productive through understanding what would be the best location, and how can we really make the category more appealing for the shopper?”
Following are key highlights of the 2010 heat-map study.
Study Stats Retail chains: 7-Eleven, Circle K, Sheetz, BP ampm, Cumberland Farms Markets: Seattle, Los Angeles, Phoenix, Dallas, Miami, Atlanta, Boston, Pittsburgh, Chicago, Denver and Richmond, Va. _ No. of stores: 48 _ Total store traffic: 643,724 _ Total buyers: 456,075 _ Conversion rate: 71% _ Average total time in store: 2.35 minutes _ Average shopping time in store: 1.68 minutes _ Average in-store units purchased: 2.4 _ Average in-store $ basket: $5.70 _ $ contribution per in-store customer: $4.00
Gas Customers’ Behavior
While signage on top of the fuel pump has become ubiquitous, results of VideoMining’s 2010 heat-map study suggest that the industry is not maximizing the promotional medium. For example, only 8% of shoppers videotaped as part of the research looked at the pumptop signage.
Interpreting this fi gure—and how to react—is not clear-cut. For Baboo, it suggests that fuel consumers are tuning out the advertisements because they are designed to appeal to mission-oriented store shoppers. “People who are pumping gas should be treated as a separate population,” she says. “They are completely different from people walking into the store to make a purchase.”
Explore the needs and triggers of this particular consumer segment, Baboo advises: “What will appeal to this segment? What do they expect from c-stores? What are their reasons for not walking into the store? If we can fi nd that out, we can customize ads on the pumptops to infl uence people to at least walk in the store.”
For Burkenbine of ampm, the pumptoppers’ low viewing rate simply confi rmed his chain’s plan to pursue TV screens at the pump to make the marketing experience more interactive.
“We have a lot of pumps, and were doing tons of different configurations of the exact same sign,” says Burkenbine. “We suspected that the answer was that pumptop advertising was not impactful—there was too much activity going on.”
On the supplier side, Hershey participates in pumptop advertising. “They raise brand awareness, especially on new items and in-store promotions,” says Krull. Hershey also encourages suggestive selling at the counter, and signage on the store windows and within the store near complementary categories such as beverages and foodservice.
Action item: When designing promotions targeting customers at the pump or those destined for the store, consider their different motivations and needs.
As a second step, consider day-part-specifi c strategies that tie into customers’ natural shopping behavior. For example, the heat-map research reveals a bump in c-store traffi c and buyer conversion for candy during lunch, mid-evening and after work.
Suggestive selling and merchandising targeted at particular day-parts—such as fi xtures offering up pastries in the morning and switched out to candy in the afternoon—could take advantage of this dynamic. “There is a lot of store execution and operations in this,” says Krull, “but I think there’s an opportunity to be merchandising and moving the store around to meet the needs of shoppers as they come in.” In this vein, Baboo suggests a breakfast station next to the coffee and hot-beverage bar, populated by complementary categories such as sweet snacks, granola bars and orange juice. “You will have incrementality vs. placing the cereal bars in a separate place all together,” she says. Action item: Tailor suggestive selling and merchandising to each day-part
Home Base vs. Secondary Locations
Primary locations had vastly higher traffi c-to-shopper and shopper-to-buyer conversion rates than the various secondary locations documented in the heat-map study. Counter displays showed the lowest shopper-to-buyer conversion rates, while endcaps and secondary aisles had the lowest traffi c-to-shopper conversion.
“Endcaps are so valuable, and there are some things we’re just not going to put on them anymore,” says Burkenbine of the results. “Now we saw it was actually very ineffi cient. Something can be more effi cient.”
In its own research, Hershey fi nds that two-thirds of candy purchases are made from the primary or home location, with one-third from a secondary location. That said, “Does that secondary location trigger a purchase thought within the shopper’s mind?” Krull says. “Then they go to the home location to look at the variety of brands they’re looking for. So it’s all about point of interruption.”
Baboo of VideoMining says assortment may be key to boosting conversion rates for secondary displays. “For candy, sales counter displays capture a lot of attention from shoppers,” she says. “But conversion for candy on the sales counter display was extremely low. … The candy category lost me as a buyer because they did not have enough SKUs and a good enough assortment on the counter.”
Two categories with among the lowest conversion rate of shopper to buyer in the heat-map study were gums/mints and cookies/crackers—18% and 17%, respectively. Baboo of VideoMining says it reflects a defective assortment or placement of the secondary displays.
“Don’t assume you’ll get incremental sales just because you have multiple product locations,” she says. “You have so many secondary locations that everybody is shopping at least one or two of them, but they’re not making a purchase because the category is influencing them to shop but not make a purchase. I’m not saying all secondary locations are ineffective … but we need to think more about which display is more effective and what are best locations for the display.”
Nearly one-half of shoppers in the heat-map study traveled by gum and mints at some point in their shopping trip; only 5% actually sprung for a purchase. Heat mapping at ampm stores showed that customers were looking at the gum but walking away without picking anything. After reviewing the data, the company decided assortment was the issue.
“We suspect it wasn’t necessarily that we didn’t have what people wanted,” says Burkenbine. “We didn’t have what these particular people wanted.” In particular, ampm is reexamining its plan-o-gramming for stores that index high for particular ethnicities, such as Hispanics and Asians. Action item: For categories seeing high foot traffic but low sales, consider whether faulty assortment or high prices are turning away customers.
Feel the Heat
Mapping out traffic—areas customers are drawn to—vs. areas where they actually picked out a product for purchase reveals which areas of the store might see a lot of footfalls but trigger few buys, and vice versa. Regardless of a store’s layout, fountain drinks, the beverage cooler and foodservice areas saw a great deal of traffic and shopping activity.
For ampm, the heat-map study highlighted the sizable impact shelving positioning had on shopper traffic. For example, if an aisle runs parallel to the store’s entry, customers might choose either side of an aisle. Most customers head to the cold vault at the West Coast site, while they check out the cigarettes first on the East Coast.
If the aisles are perpendicular to the entry, the heat map revealed “hot spots” where customers encounter facings of products. Behind the aisle, however, the trail grows cold. The company is examining what effect perpendicular might have on particular categories.
“Do you want to encourage what’s already natural, or push for something else?” Burkenbine says. At ampm sites, the item most commonly purchased with a packaged beverage is chips. “If you put chips out there, it increases chip sales, but if you flip it around and put sweet rather than salty there, then would you get an additional sale?” he asks. “That’s what we’ll have to play with to see how that works.”
Action item: Experiment with shelving configuration and category positioning to find the mix that best resonates with customers.