Taming Fresh

MICHAEL GARRY

Investments in technology appear to be paying off for Sobeys.

Beginning in 2005, Sobeys, Stellarton, Nova Scotia, Canada's second-largest food retailer with 1,332 stores, upgraded its information processing and decision-support capabilities, focusing on the “back-shop,” according to its 2009 annual report. The company introduced the Club Sobeys loyalty card program in 2008 and opened a state-of-the-art distribution center in Vaughan, Ontario, last year.

Meanwhile, over its past three full fiscal years, Sobeys' operating income grew from under $300 million to $401 million in fiscal 2009, which ended last May.

Another technology system that has contributed to Sobeys' operating efficiency and income during this period is a fresh-item management (FIM) system from Invatron, Mississauga, Ontario. Speaking earlier this month at the National Retail Federation's 99th Annual Convention & Expo in New York, Clinton Keay, Sobeys' senior vice president and chief information officer, described the retailer's adoption of this system, which its stores use to track shrink and plan production for perishables and, more recently, to order fresh products.

Sobeys, which operates supermarkets under such banners as Sobeys, IGA, IGA extra, Foodland, Price Chopper and Thrifty Foods, began working in earnest on the FIM system in 2007 after an initial pilot in 2006. “There was pressure from the business to help with the fresh-item offering,” said Keay, a former finance executive who has been with Sobeys for two decades.

The FIM implementation has been completed in Sobeys' Atlantic region, with its other three regions across Canada in different stages of deployment. The project coincided with the company's expansion of its higher-margin fresh departments, including prepared meals, deli and bakery.

Keay said the FIM system has enabled Sobeys to achieve better in-stock position, improve sales and margins, and reduce shrink in its perishables departments. “The business pushed us to move [the implementation] faster because of the benefits received at the store level,” he said, adding “the ROI is quite significant.”

“It was nice to see a company's profit in its fresh departments so impacted by a technology project,” said Frank Urbaniak, consulting principal, C-Core Consulting Group, Sussex, N.J., who worked on the implementation for about two years. Urbaniak participated in the NRF presentation with Keay.

Sobeys is not alone in using a FIM system. Among other retailers with this application is Winn-Dixie Stores, Jacksonville, Fla., which also began deploying Invatron's system in 2007 to gain visibility into perishables losses. United Supermarkets, Lubbock, Texas, uses Web-based FIM applications from another supplier, ADC, Tampa, Fla.

“We are seeing an increase in inquiries in this area from food retailers,” said Mike Griswold, vice president, research, AMR Research, Boston. “I think the realization that FIM can provide significant shrink and margin improvements is finally sinking in.”

Urbaniak pointed out that FIM technology may help supermarkets to fend off competitive pressures on fresh sales from convenience stores, drug stores and other alternative formats.

THREE-PHASE ROLLOUT

The first phase of Sobeys' FIM deployment involved capturing and tracking daily fresh-item shrink information — generally on discarded product — with handheld scanners. This information populates shrink reports that are disseminated regionally. Shrink data “opened our eyes regarding the opportunity for fresh-item management,” said Keay.

The next use of FIM was in production planning, a very “data-intensive” operation, said Keay. The production planning function is designed to help stores produce the right amount of fresh product at the right place and the right time of day and week. It includes spot checks of inventory levels using handheld scanners.

“We found that we had been producing to historic norms,” said Keay. “We would bake for the entire day in the morning, yet by the end of the day we would have out-of-stocks or stale product.” But the FIM system helped to prevent stores from having “just one piece of bread left at 9 p.m.”

The production plan, which can be set by day-part if multiple cycles are needed, is based on a forecast of sales, reduced by the current on-hand inventory, and incremented by the minimum amount needed at the end of the day (or until the next production cycle). “Every store determines what the critical [minimum amount] is for them,” said Keay.

Keay noted that the on-hand, or perpetual, inventory calculations require tracking discards, rewraps, bad labels and transfers. “Perpetual inventory is something not a lot of retailers do in fresh,” he said. “If those processes aren't disciplined, you're going to run into challenges with the system.” It requires change management and training, he added.

The change management called for involvement from top executives, noted Urbaniak. “Department managers doing this for 20 or 30 years said, ‘Computers can't tell me anything better than what I know.’ But a region president was a firm believer in the system and made this a No. 1 focus.”

“Some of the change management around production planning scares some people,” observed AMR's Griswold.

On the other hand, some store personnel were pleased that the system figured out production plans for them, Urbaniak noted.

The third phase of the project is suggested ordering of fresh products, which wasn't in the original implementation plan. “As a result of doing [production] forecasting, we were able to do automated or suggested ordering,” Keay said. “We fell upon it.” Sobeys began suggested ordering for produce during its rollout of production planning. “We improved profits significantly when we did the deployment of ordering,” he noted.

The system is able to order products like Macintosh apples “so you have significant decreases in back stock,” said Urbaniak. “It's one of the greatest benefits of the whole system.”

Future phases of FIM deployment will include full ordering of recipes and ingredients, full physical inventory capture, Web integration, PC-based scales, alerts and business intelligence analytic tools.

While Keay declined to provide specific financial results of the FIM implementation to date, he declared that “it is possible to realize substantial benefits per week per store,” including improved in-stock position (and therefore higher sales), reduced shrink (and thus better margins) and lower inventories. “As we deploy the solution, stores get an immediate benefit.” He also cited “an improved company fresh image.”

SINGLE DATA SOURCE

What has been very important to Sobeys' FIM implementation is having easy access to sales and inventory data in what Keay called a “single source of the truth.” Sobeys' data is “generated once and not housed in multiple areas,” he said. Moreover, it “comes from the same t-log [transaction log]” and is fed into the FIM system “on a timely basis.” Data requirements “are something you have to pay close attention to,” he noted. This includes defining “master data” such as quantities of production and presentation stock, recipes and units of measure.

The FIM system's forecasting system does quite a bit of the data processing, providing forecasts for inventory, shelf requirements, production orders and suggested orders.

One of Sobeys' advantages in implementing FIM is that, despite its multiple banners, it has common POS systems and in-store processors. It also has an SAP enterprise resource planning (ERP) system supporting finance, human resources and merchandising, and a robust communications network to ensure that information is available to all stores. The FIM setup is “standard across Canada,” he said.

Sobeys followed a “building blocks” approach to FIM. It began with setting up foundation systems such as scales, scale management and a POS interface for t-log sales. That was followed by identifying and improving manual store processes that support FIM, such as shrink capture and perpetual inventory, and establishing the master data requirements.

“Everybody wants to start at the top with a nice dashboard, but we built this in a sequenced way,” Keay said. “Process improvement is just as critical as the technology.”