Kmart

KMART CORP.

A SEASONED

PERFORMER

The venerable retailer's bid to recoup its fortunes includes a system that overhauls the management of highly profitable seasonal merchandise

BY MEGAN SANTOSUS

For retailers, the effective management of seasonal merchandise is a delicate proposition. The selling window is of limited duration and the goal is not to get caught with too much-or too little-inventory at the wrong point in the seasonal cycle.

Take Christmas items. If there were no tinsel or wrapping paper or spare tree lights on the shelves the week before Christmas, customers would head for the exits. But if, on Dec. 26, the shelves were still well-stocked with these items, that too would be egregious.

And, because of the shorter selling season, the profit margins are kept higher on seasonal items than on regular stock; so, when it comes to maximizing profits, the balancing act between supply and demand is all the more crucial.

If at first blush all that tinsel doesn't seem like such a big deal, it can definitely add up. For ESPRIT Award honoree Kmart Corp., holiday merchandise totals some $2 billion-and Kmart's Seasonal Merchandise Management System (SMMS) currently handles only a fraction of the seasonal goods Kmart sells. (The inaugural project was limited to holiday items; it was recently extended to include back-to-school merchandise, and there are plans for further expansion.)

The universe of seasonal retail merchandise is potentially quite broad and might include such items as snow blowers, lawn mowers, bathing suits and bags of peat moss. Kmart, however, declines to reveal the total percentage of revenues for which its seasonal inventory accounts.

According to David M. Carlson, Kmart's senior vice president of corporate information systems, managing seasonal inventory perfectly would mean "the last customer who wants a particular item will find the last one in stock."

[Editor's note: After this article was reported, Carlson was shifted out of Kmart's top IS spot to other unspecified duties, reportedly because of dissatisfaction over problems plaguing Kmart's inventory replenishment systems.]

While perfection is elusive, Kmart's use of client/server-based technology to fine-tune seasonal merchandise management and centralize decision making at its Troy, Mich., headquarters is paying off big. Through fiscal 1996, Kmart projects a total return from SMMS in the neighborhood of $240 million. Not bad for an investment of roughly $2.5 million.

Kmart could certainly use the money. Earnings at the $34 billion-a-year discount department store chain have been plunging lately. The company recently announced a restructuring effort that calls for closing 110 stores and eliminating 6,000 jobs. But many retail analysts contend that shedding unprofitable stores alone will not cure all of the company's ills. If Kmart hopes to regain healthy earnings, it must reduce overhead in existing stores and further exploit the strategic use of information technology, according to Seth Kranz, a management consultant with CSC in Newton, Mass.

SMMS is a step in the right direction. By providing buyers and planners at headquarters with daily sales and inventory data, the system has transformed the way Kmart manages the seasonal goods in 2,350 stores in the United States. Inventory and pricing decisions are now based on sales performance at the individual stores. As a result, Kmart can maximize profits by preserving full price-and gross margins-in stores where seasonal items are selling at or above plan. In stores where sales are lagging, the company can tailor an appropriate merchandising strategy to eliminate inventory by season's end.

Since implementation in fiscal 1992, the system has improved seasonal merchandise sell-through-the total number of items sold divided by the total received-by 5 percent. While that may not seem like much, selling more seasonal merchandise at the optimum profit margin, rather than at a loss or at break-even, translates into millions of dollars. In 1992 and 1993, higher sell-through rates on seasonal merchandise accounted for incremental profits of $28.9 million and $37.2 million respectively.

Historically, Kmart managed seasonal merchandise chainwide, measuring total sales against total purchase orders. Chainwide markdowns were mandated when sales fell behind plan. This blanket approach meant there was no rational link between local prices and local demand. Thus, the needed stimulus of a price promotion in stores where sales of an item were slow became a drain of profits out of stores where sales were brisk. Worse, in stores with high demand, the discounted price led to premature depletion of inventories, causing shelves to empty before season's end. And moving merchandise between understocked and overstocked stores in time to do any good was both costly and logistically difficult.

"District managers manually took inventory at each store," explains David A. Wegscheid, the divisional vice president of merchandising. "There was no practical way to figure out which items could be transferred by eyeballing the stuff." Even if it was known that a store was running low on an item, it took too long to scour the reports to locate stores with surpluses.

It fell to Joseph G. Lichocki, a project manager in corporate systems development, to work with Kmart's merchandising group to link price to demand on a store-by-store basis. Lichocki was familiar with the frustration felt by planners and buyers. In 1987, Kmart had acquired a massively parallel relational-database server from Teradata (now an NCR division of AT&T Global Information Systems). It was used to process sales, purchase-order and inventory data. But there had been no easy way to view the data. While hefty weekly reports were produced, merchants had to flip through hundreds of pages to find what they wanted. What merchants needed was daily sales information that could be easily manipulated so that action could be taken in time.

"Our challenge was managing the volume of data while providing our end users with ultimate flexibility," says Lichocki, the architect of the SMMS. "We had the Teradata, but we weren't sure how to make the data 'fingertip ready' and useful." According to Lichocki, the system had to recommend, execute and track action on behalf of merchants. "Any system with those capabilities," he asserts, "will provide a return."