Brown Shoe: Linking Supply With Retail Sales
With 90 million pairs of shoes sourced globally, Brown Shoe Co. had a supply chain shortcoming: The leading footwear supplier had no way to link retail sales with incoming supply.
"We were limited to ad-hoc reporting and e-mailing order changes, but we were not reacting to consumer demand," says Ron Boschert, the company's director of demand planning. "We had no systematic way of updating the forecast based on consumer activity."
St. Louis, MO-based Brown Shoe's brands include Naturalizer, Life Stride, Franco Sarto, Via Spiga, Carlos Santana, Etienne Aigner, Dr. Scholl's, Nickels, and Buster Brown, with footwear sold at more than 2,000 department stores, mass merchandisers, independent and specialty stores worldwide.
Orders were placed in overseas factories with no systematic process to adjust based on consumer demand. "Basically, orders were delivered as we originally placed them, without making adjustments based on consumer activity," says Boschert.
Systems At A Glance
Assortment Planning & Allocation: JDA
Demand Planning: Logility
Financials: Oracle/PeopleSoft
Price Optimization: ProfitLogic (Oracle)
Transportation & Logistics: Oracle/PeopleSoft
In order to become more consumer-driven, a forecasting process was needed to synchronize customer demand with factory operations, to service retail customers more effectively. "We needed to capture information about our customers' purchases, and then apply that information to our sourcing and inventory management processes," says Brian Minnix, Brown Shoe's vice president of retail planning & analysis.
About three years ago, Brown Shoe began its quest by manually feeding pointof- sale (POS) information from its 1000 company- owned Famous Footwear stores into a single PC. "We wanted to see what kind of forecast we could turn out and what we could do with the information," Boschert says. "After we did a proof of concept, we went through a formal systems search and software demos."
Logility Voyager Solutions was chosen, in part because of its ability to handle highly seasonal, short-lifecycle merchandise, as 75 percent of Brown Shoe's product offerings are new each season. The company also was intent on using POS information to improve forecast accuracy.
During the four-month implementation, the biggest challenge was internal - stemming from the need for a huge amount of data to be organized and loaded into the new system. "We had a lot of rich EDI point-of-sale history from our customer base and we had to get that organized," Boschert says.
After the new system went live, Brown Shoe had an accurate weekly forecast to predict consumer sales and improve inventory flow to retailers. "When a shoe is in high demand by consumers, the chance to sell more in season requires a quick reaction," says Boschert.
New collaboration between retailers and suppliers
Brown Shoe now closely collaborates with its Famous Footwear company-owned retail store chain on inventory planning by obtaining weekly POS data, getting retail replenishment as close to true market demand as possible. The retail stores provide POS data by size, width and store, so it is clear which shoes are selling fastest.
Brown Shoe now closely collaborates with its Famous Footwear company-owned retail store chain on inventory planning by obtaining weekly POS data, getting retail replenishment as close to true market demand as possible. The retail stores provide POS data by size, width and store, so it is clear which shoes are selling fastest.
"We are much smarter about managing our inventory based on the consumer popularity of the footwear," says Boschert. "Brown Shoe is becoming more consumerdriven in terms of how we forecast and manage our inventory." The weekly forecasting has resulted in improved inventory turns, with deliveries made every two weeks instead of every two months for some items.
In addition, working with its factories, the company established new parameters regarding time frames during which it could make adjustments to orders. As such, Brown Shoe's communication with its suppliers has become much more frequent, with inseason adjustments routinely made to incoming factory orders.
Better forecasting = shorter lead times
Forecasts can now be created based on just a few weeks of selling at retail. "For new products, we can quickly make adjustments to the original expectation based on consumer preferences," says Boschert.
Forecasts can now be created based on just a few weeks of selling at retail. "For new products, we can quickly make adjustments to the original expectation based on consumer preferences," says Boschert.
Brown Shoe has had success with forecasting sizes as well, with lead time reduced by about 50 days for core items selling at Famous Footwear. "For every style and color of footwear we have, we typically have 12 to18 sizes, so forecasting is imposed on the sizes," Boschert says. Forecast accuracy is running around 80 percent 13 weeks into the future, a critical timeframe because that is when the factory needs to cut sizes.
"We use the forecast to stage materials for some core items. By doing so, we are able to delay the point in time when the factory cuts sizes, so we have a better idea of what sizes are being consumed at retail," says Boschert.
The solution is also used to forecast consumer demand for department store customers including Macy's and Dillard's, but currently this information is only used to provide guidance to Brown Shoe's internal sales force. "Also, for department stores, we are forecasting at the item/color level, and have not begun to forecast by size as we are doing for Famous Footwear," says Boschert.
System is tweaked
Shortly after the new system went live, the number of periods of demand used to calculate forecasts was adjusted. "We noticed that the first week or two of selling at retail can be misleading, and that you need three to four weeks of selling in the number of stores you expect to be in to calculate a reasonable forecast," says Boschert. "Otherwise, you are not calculating the forecast based on the correct information."
Shortly after the new system went live, the number of periods of demand used to calculate forecasts was adjusted. "We noticed that the first week or two of selling at retail can be misleading, and that you need three to four weeks of selling in the number of stores you expect to be in to calculate a reasonable forecast," says Boschert. "Otherwise, you are not calculating the forecast based on the correct information."
Adjustments can also be made to baseline demand. By using attribute-based forecasting -grouping items by similar traits - predictable seasonal profiles are created which are then used to forecast individual items and sizes.
"If we see a trend, we can get the item back on track based on the most recent consumer demand," says Boschert. For example, last year Famous Footwear enjoyed significant growth in its women's casual clog category, and this data was used to forecast consumer demand for shoes in this category.
The firm plans to upgrade to Logility's web-based version 7.5 within the next six months, with the end goal of increasing productivity and viewing information faster and more completely. Forecasts for specialty retailers and the company's e-commerce site, shoes.com, are not currently
The firm plans to upgrade to Logility's web-based version 7.5 within the next six months, with the end goal of increasing productivity and viewing information faster and more completely. Forecasts for specialty retailers and the company's e-commerce site, shoes.com, are not currently being utilized, but this is a possibility for the future.
Continued increased collaboration between retailer and supplier will ultimately benefit Brown Shoe's consumer, says Minnix. "A more streamlined supply chain enables us to deliver trend-right shoes closer to season, so shoppers can always find the latest fashion," he says. Apparel