When people at Food Lion set out in 2002 to reinvent the supermarket, one principle above all guided their work: Design it with the convenience of the customer in mind.The customer who's busy. Who wants to find jars and bottles at easy reach. Who doesn't want to be forced to go down extra aisles to find what she wants. Who expects to find products organized by how people look for them, not by what's

When people at Food Lion set out in 2002 to reinvent the supermarket, one principle above all guided their work: Design it with the convenience of the customer in mind.

The customer who's busy. Who wants to find jars and bottles at easy reach. Who doesn't want to be forced to go down extra aisles to find what she wants. Who expects to find products organized by how people look for them, not by what's easiest for direct-store-delivery vendors to stock.

The resulting Bloom, a Food Lion banner, operates five stores in North Carolina that opened between May and October 2004. It represented a radical approach to store design, though one whose elements other retailers are emulating.

In the center of the store, that meant a number of things: lowering shelves, tossing conventional ways of organizing categories and grouping associated products in "universes."

Using shorter aisles meant less space to work with. New rules had to be devised to guide product placement, and products didn't always easily conform to new organizing principles.

"Every planogram, it was literally like throwing a thousand-piece jigsaw puzzle up in the air and doing a redo," said Robin Johnson, director of marketing and brand development at Bloom. She was a member of the concept renewal team that spent two years researching and exploring new concepts that resulted in Bloom.

A year and a half after launching the concept, she talked about the challenges in executing Bloom and how they were resolved.

An easy decision Bloom made was to separate food from nonfood. That came straight from customers, who said, "If I only buy laundry detergent once a month, why should I have to walk through the aisle to get to the things I buy weekly?" They also preferred having chemicals kept separate from food.

In addition to separating food and nonfood merchandise, Bloom put nonfood in shorter aisles that ran perpendicular to the food aisles. The idea was to make them stand out and further underscore the separation. Shelves stretch 66 inches tall rather than the standard 72. Fitting categories in smaller gondola runs meant eliminating some less popular package sizes and lower-tier brands, but ease of shopping dictated it.

"The average female is 5-foot-4. And remember, our brand is all about convenience and comfort," Johnson said of the decision to use the 66-inch-tall aisles.

Bloom also sought to organize products according to how people look for them. Take baby food. In a Bloom store, it's ordered by baby stage rather than brand, as conventional stores do.

"Most retailers are going to have a brand set. All of Gerber is together, whether it's infant, stage 1, 2, 3, 4," Johnson said. Shoppers confirmed what Johnson remembers feeling in her days as a young mom, when, she said, "I was less brand loyal than I was wanting to find the right item at the right price."

That rule didn't extend to diapers, though, which stayed with their brands. "Diapers, people tend to be more brand loyal," she said.

In grocery, a recurring challenge was putting products together in a way that made sense for shoppers.

Cookies, crackers, chips and the like were centralized in a snack "universe." Yet what to do with products that blurred category lines, like Nestle Pretzel Flipz, chocolate- and yogurt-coated pretzels; and Hershey's Swoops, the potato chip-shaped chocolates? Some wondered if raisins belonged with snacks or dried fruit.

In several cases, the primary-ingredient rule applied. Since Flipz are essentially a pretzel, it was decided, they stayed in snacks. All-chocolate Swoops went to the candy aisle. Raisins, the consensus was, belonged in dried fruit.

Bloom took a similar approach to canned vegetables, which also were grouped by type. If a product had multiple ingredients, such as corn salsa, it was placed with other products that shared its main ingredient. When it came to salad dressing, however, there were so many flavors that defied the conventional ranch, French and Italian categories. The solution was to create a specialty category just for them.

Customer feedback showed that while a new format may be logical, they still have to be retrained.

Bloom first tried grouping sodas by flavor, a dramatic departure from the standard DSD model where brand rules.

"We've heard back loud and clear from customers, they really are brand loyal, particularly in the cola category," Johnson said. "The decision is brand first. They're looking for that color." So Bloom went back to a brand set, while tertiary flavors like orange, root beer and cherry are still grouped by flavor (Bloom says the set is still a work in progress).

Bloom wanted to make pasta a commodity set that organized items by type, then by brand. The varying package sizes and types made it difficult to keep the shelves stocked properly, though. "It really became a bit of a train wreck in how it looked," Johnson recalled. Customers preferred the old familiar way of shopping the aisle anyway, which made it easy to go back to a brand-based set.

Such examples illustrate the challenge of carrying out a radical supermarket redesign, said Neil Stern, partner at McMillan-Doolittle, Chicago, a retail consulting firm.

"You're doing it with the customer's best intentions in mind," he said. "Even though it's more customer friendly, you've been training people for 50 years, 'This is how you do it.' Whenever you try to change that, there's re-education of the customer base."

Adding to the risk, Bloom adopted an upscale image that aimed far from the price-focused Food Lion shopper base, he said.

Customers gave unprompted feedback that they loved the store layout, one commenting that it was so intuitive, a woman must have designed it. But Johnson and company aren't satisfied. Starting Aug. 31, Bloom introduced a series of refinements, based on a year's worth of customer feedback. Those included new wine displays for in-and-out promotions and broader selections of perishables, beer and specialty grocery products. It continues to solicit shopper feedback via focus groups and phone and e-mail surveys, as well as executives' own observations.

Bloom's experience is instructive as more retailers design stores based less on logistics and more on category adjacencies.

"We're learning as we go," Johnson said. "Intuitively, it's a nine. It's not a 10 because I don't believe we're ever going to be there. We've got to constantly be out in front of that. And constantly remain open to feedback."

When Less Is More

Bloom features extensive use of gadgets, but Susie McIntosh-Hinson, concept creator of information technology for Bloom, is quick to say that was never the end goal.

So more than a year after opening the first of its new concept stores, Bloom is simplifying its technology to make sure it's meeting its goal of helping customers get in and out faster.

Bloom's wine kiosks help customers with everything from pairing wines with food to planning parties based on the number of guests and their drinking preferences. Yet the retailer found that many of these functions were going unused by customers. Accordingly, the retailer decided to look at changing content providers for its wine kiosks.

"People want to know what wine goes with this food, what cheese goes with this wine," McIntosh-Hinson said. "They need something they can walk up to and get an immediate answer."

That also explains the popularity of Bloom's so-called Information Stations, kiosks that help people locate products and check item prices. Each store has seven or eight.

When Bloom relaunched its stores this fall, it piloted a new self-checkout that speeds the payment process by cutting the number of steps required for checkout.

Its self-scanners that let customers add up their purchases as they shop could be loyalty marketing goldmines, but McIntosh-Hinson was emphatic that the scanners won't be used for data collection. "Our primary filter is, what does a customer want. What they didn't ask for is, 'I want you to take my name and address and send me stuff."'

Its self-checkouts and scanners exist to give people options, not force them to forgo personal service. Self-checkout users, Bloom found, tend to have small baskets, while people who use scanners tend to have full carts and are looking to shave minutes off the trip. "Sometimes you want full service, sometimes you want self-service," McIntosh-Hinson said.

"Bloom was never an initiative about technology," she said. "But we learned technology is important, and will remain a part going forward. Customers like it."