Forecasting for Success
Apr 2, 2007
Dori Saltzman, RIS News, April 2007
Forecasting is a tricky business, retailers lose money every time a demand forecast is inaccurate. Sophisticated software is necessary to ensure that customer demand and in-stock products are a perfect match.
The Guitar Center chose Quantum Retail's Q forecasting software to help it get a handle on the "fashion-based, choppy demand" of its higher-priced merchandise. Early indications are positive: though Q is currently forecasting only 50 percent of the inventory -- representing the vast majority of sales dollars -- the forecasts it generates are 20 percent better than before, according to Irene Messier, executive over inventory management.
Prior to implementing Q, the Guitar Center used two separate forecasting systems. One, an automated replenishment system, was used solely on the less expensive, high velocity items -- the so-called "sticks and strings." A manual, excel-based system was used on the more expensive, slow moving items -- such as $700 guitars.
The Guitar Center selected Quantum Retail to provide a system to handle both types of inventory. "We were looking for something that gave us better forecast outcome (on the slow moving items) and also hoping to find a solution that could be utilized in updating our auto-replenishment product as well," says Messier. The Guitar Center upgraded its forecasting methods after identifying a need for greater demand accuracy for the slow-moving items. "One of the factors in a situation where you're dealing with Excel-based forecasting is that the level of accuracy, due to the amount of information that needs to be processed, isn't going to be nearly as accurate as using a statistics-based forecasting model," says John Zavada, executive vice president and CIO.
The $700 guitar is the quintessential high-priced, slow moving product for which accurate forecasts are vital. Guitar Center stores risk unbalanced assortments and lost sales if forecasts are off.
In order to determine that Q could generate accurate forecasts for both the high priced and high velocity items, the Guitar Center ran a six-month pilot before rolling out the system live on the higher priced inventory. The pilot was run on a sub-section of the products. "We observed consistent month after month, better forecasting coming out of the Q application," says Messier.