JD.com and Alibaba create indexes to identify Chinese shoppers’ spending trends, which help retailers gain insight.
The web is transforming the setting of prices from an art to a science—with notable boosts in sales and profits.
Maintaining the best prices on the retail floor to produce the best margins is as fickle a process as choosing the hottest teen apparel fashions. In any given week, a merchant may have to make thousands of pricing decisions based on factors like changing demand, special pricing offers from vendors, or competition from the new retailer who just opened down the street.
Changing prices at the right time and for the right amount can produce a windfall of sales and profit margins. But setting prices too high or too low risks upsetting the fine balance between moving products according to a goal and keeping customers both profitable and happy. “It’s critical for us to predict how consumers will react to the pricing moves we make,” says Gary Charboneau, senior vice president of sales and marketing for Duane Reade Inc., a chain of 250 New York-area drugstores.
Getting the nuances
Often, setting prices is as much art as science. But now several vendors of price-optimization systems are trying to make it more science. And they’re basing their efforts on the web. They argue that data that move over the Internet can be both gathered and disseminated in a more timely fashion, allowing retailers to exploit nuances in pricing that could make the difference between profitable and unprofitable.
Analysts agree, citing web-based systems’ ability to integrate with legacy systems and the universal access throughout an organization that a browser-based system affords. “Retailers should make sure that any new pricing system they buy is web-based,” says Hung LeHong, analyst with researchers/consultants Gartner G2. “That’s the wave of the future.”
Duane Reade is rolling out DemandTec Inc.’s DemandTec 3 pricing optimization software platform chainwide following a test last year that showed the system produced increases in unit sales without sacrificing margins, Charboneau says. “The pilot worked very well, close to what we had expected,” he says.
Duane Reade increased unit sales of baby formula by 14% and of disposable diapers by 10% after accepting the DemandTec system’s recommendations for initial prices, Charboneau says. Often, boosting unit sales requires lower prices, which, of course, results in declining margins. But not in this case. “The forecasting turned out quite accurate,” Charboneau says. “While unit sales rose, gross margins remained the same for both baby-care product categories.”
DemandTec hosts the DemandTec 3 system on a web server that Duane Reade pricing managers access via a browser. The system contains historical pricing, sales and inventory data on items and gets a regular feed via the Internet of ongoing sales, pricing and inventory data as the stores sell down the products.
Charboneau says it will take to the middle of this year to get the initial pricing system working in all 250 stores. “There’s quite a bit of work involved to implement this for initial pricing,” he says.
Pricing managers can test a series of what-ifs by accessing the appropriate product page and inputting hypothetical prices, based on, say, margin or competitors’ prices, fed into the system by QRS Corp.’s price-tracking service; amount of inventory; time period during which they want the merchandise to sell and other parameters. The system responds with recommendations.
Once the recommendations are finalized, the information can be distributed throughout the enterprise via the web to all applications that need the data, including servers and POS terminals in stores.
A major ingredient of making this all work is that Duane Reade’s managers trust the system. “The trust merchandisers have with an optimization system is a big deal,” says Gartner’s LeHong. “Retail managers have been relying on instincts, their feel of the market, to price things, and all of a sudden they have a computer coming up with recommendations. That requires trust.”
Charboneau says several retail managers spent two weeks of training in how to use the DemandTec system, followed by two more weeks of on-the-job learning before they became comfortable using the system and trusting its recommendations, but that they continue to learn how to better use the system.
Managers can easily override the system’s recommendations. But for the most part, Duane Reade prefers its managers to avoid overriding recommendations, Charboneau says. “The more you override the system, the more you handicap its ability to meet the goals set for it,” he says.
Duane Reade tested the DemandTec 3 price optimization software suite in 20 stores, using another 20 as control sites for comparison, from December 2002 to April 2003. Charboneau declines to say what Duane Reade invested in the DemandTec 3 system. But Dan Fishback, CEO of DemandTec, notes that it can cost a retailer $1-$8 million in startup fees depending on a company’s size, including number of stores and SKUs, plus an ongoing annual maintenance fee of 15-20% of the license fee.
Before using the DemandTec system, Charboneau says, it wasn’t possible to come up with effective price strategies for thousands of products in each store. Duane Reade stores carry an average of 18,000-20,000 SKUs. In an effort to set pricing levels, it would gather information on past item sales from its own system as well as prices set by its competitors. “We’d make pricing moves and track what happened, but we didn’t have a real forecasting ability,” he says.
The new system is more effective, though not without its challenges, Charboneau says. Because it factors in more details of pricing and expected customer demand based on historical data and competitors’ pricing, the DemandTec system takes longer to put together a price modeling. “But it gives a map of what we’re looking for to happen, then within four to eight weeks we see the customer response to our pricing moves,” he says. “Then we track how our business is changing in relation to the way we anticipated it would.”
For now, Duane Reade is using the DemandTec system only to set initial pricing on about 70% of its non-pharmaceutical sales. The other 30% include products such as greeting cards whose retail prices are pre-set by suppliers.