Demand Visibility: Supply-chain transparency reduces inventory costs

AMR Research defines Demand-Driven Retailing as the blending of an atomic-level understanding of consumer demand with effective retail execution to create a superior, profitable shopping experience. But it's one thing to understand the concept, and another to get there.

Visibility into demand is a key ingredient to attaining this vision. While retailers have long collected their POS data, only in recent years have they begun truly tapping its ability to sense and respond to what's happening in their stores. When combined with other data about their customers, they gain even more insight into assortment and inventory needs.

Checking the Backlist

Barnes & Noble is among those seeing the profitability that can result from better visibility into demand. Improved insights into the chain's backlist titles were one result of a project that boosted forecast accuracy to more than 85 percent, reduced inventory 30 percent to 40 percent and garnered $4 million in annual savings.

"We needed a much more scientific, forward-looking view of our 70,000 fastest-moving older titles," explains Joseph Gonnella, vice president of inventory management and vendor relations for Barnes & Noble. "New books do drive sales, but 65 percent of our business is basic backlist, with automated replenishment." The bookseller maintains 1.5 million active titles.

Barnes & Noble deployed a suite of i2 Technologies modules, including Demand Planner and Demand Fulfillment, and began using them to monitor POS signals and generate a 12-month forecast and a 16-week purchase plan.

"Every month about five percent of the backlist bubbles up," varying from forecast, Gonnella says. "Old books become new because authors publish new books," or become newsworthy. That spurs review by a human, who can tag the data as repeatable or non-repeatable for use in future forecasts. An author's death, for example, is presumably non-repeatable.

Changes might include a new re-order trigger point for a title. Local popularity also impacts replenishment. "The book business has a lot of variability by store, by title," says Ash Bakre, director of supply chain planning for Barnes & Noble. "That variability expresses itself as incremental safety stock. The more variables, the more safety stock." The better the data about demand, the better that stock can be fine-tuned.

The system is also beginning to recognize seasonality of stock, such as for Mother's Day or the back-to-school season. "Every time we improve the forecast, we reduce safety stock," notes Gonnella.

Daily and trickle polling helps merchandisers quickly detect surges in demand. Analysis of just a few hours' sales enabled the retailer to quickly place reorders for Hillary Clinton's book, for example.

Using this system, Barnes & Noble now boasts a 92 percent distribution center fill rate. About 98 percent of the time, the retailer has the book somewhere in its system, and delivers a customer order within three days.

Barnes & Noble is beginning to move to weekly, rather than monthly forecasts for its backlist, and is also starting to use the tools for its frontlist. While booksellers, unlike most other retailers, are able to return unsold merchandise to the manufacturer, more aggressive forecasting will minimize the considerable expense involved in returns. The bookseller has also begun collaborative planning, forecasting and replenishment (CPFR) with some trading partners, which is helping to improve forecasts, get quicker turns and reduce safety stock.

The retailer is already seeing results. Comparable store sales have been beating competitors and per-store inventory is down one percent. "We're getting more productivity out of our inventory, and delivering higher service levels," Gonnella says.

Collaborative Visibility

Demand visibility is also a key ingredient in the CPFR relationship between Sears, Roebuck and Co. and Michelin. Both partners were maintaining high levels of inventory despite collaborating on sales forecasts. Sears, with help from Michelin, launched a CPFR program for the tire category and later expanded to other partners, using a Manugistics GNX CPFR solution.

Sears POS and forecast data now fuel collaboration among Sears and its tire partners, using it to optimize replenishment and inventory, minimize stock, maximize customer service and optimize transportation. Alerts highlight issues to be discussed in weekly collaboration calls, where any required adjustments are made.

As a result of the program, Sears' in-store stock levels improved by 4.3 percent, distribution center inventory was reduced by 26 percent, store fill rate went up 10.7 percent, and proactive follow-up to exception reporting generated $140,000 in margin dollars.

According to AMR, retailers and CPG suppliers alike need to tackle demand visibility on three levels. The first is replenishment demand, such as the systems in use for Sears' tires and Barnes & Noble's backlist. Next is surge demand, such as Barnes & Noble's analysis of trickle polling to re-order hot new books, blended with seasonality and local preference data to fine-tune inventory. The third is future demand, a long-term perspective that incorporates effective product portfolio management for future product creation. At a July, 2004 supply chain forum hosted by AMR, 94 percent of respondents called demand visibility extremely or very relevant to their businesses.