Removing Inventory Guesswork

Although the blizzard of '03 wreaked havoc on most stores throughout the Southeast, not every retailer considered it bad news. Some, like Movie Gallery Inc., actually welcomed the onslaught of snow, considering it an opportunity to rent and sell more videos and DVDs than usual.

Movie Gallery was prepared with extra copies of movies it knew would turn over quickly, thanks to a recently installed inventory optimization system that measures the effects of future weather patterns on consumer demand by product, location and time.

To optimize inventory based on weather patterns, Movie Gallery turned to Impact LR, an Internet-based application from Planalytics Inc. of Wayne, Pennsylvania. The system analyzes two years of Movie Gallery's rental history as well as expected weather patterns in each region where the company operates stores.

"Rain, snow and low temperatures — things that would keep you indoors — are good for our business," notes Ted Innes, senior vice president of marketing at the Dothan, Alabama-based company, which has 1,800 stores in the U.S. and Canada. "Before, we'd just look at USA Today and assume that the reason our business in the Northeast was down was because it's been too sunny. This was a way of quantifying what we had always talked about."

Movie Gallery also uses the system to better plan promotions. "If we know what the weather will be, we can adjust our promotional calendar to maximize that prediction. We may move a promotion back or up a few weeks depending on what the system recommends," Innes says.

Removing the Guesswork

More and more retailers are turning to inventory optimization software as a way of improving forecasting for seasonal and short lifecycle items, markdowns and promotions, store assortment and key item planning. Many believe that optimizing inventory to increase turnover and reduce excess inventory is particularly important in today's sluggish economy.

According to Steve Schwartz, senior vice president of planning and allocation at men's clothing retailer Casual Male, "When the economy is good and the company is growing, sales rise, and if you control your expenses, profits go up proportionally. But when sales are no longer rising because the country is in a recession or waiting for a war, you still have to be able to increase your profits somehow."

In many ways, inventory optimization software transforms an educated guessing game into a game of scientific probabilities. Once you have science on your side, the guesswork — and the mistakes those guesses generate — are reduced dramatically.

"The mathematical algorithms behind it are amazing," Schwartz says. "It can tell you what kind of acceleration you'll get in unit sales if you take a 10 percent markdown versus a 20 percent markdown. And if we tell the system we want to be out of this product by July 1st, it looks at our history and current sales patterns, taking our current on-hand and on-order into account, and tells us how much we'll have left at the end of the season. If it tells us we'll end with 5,000 pieces more than we need, it can recommend what kind of lift we need and how and when we should mark down the product."

A recent emergence from bankruptcy was one of the factors that led Casual Male to engage ProfitLogic Inc. of Boston to run its Pricing4Profit software on behalf of the 480-store company. The software recommends optimization methods based on analysis of several years of sales history by SKU within each store.

For Casual Male, the biggest benefit has come in the area of markdowns. Executives traditionally have marked down like items across every store in the country. That system was inefficient, Schwartz says, because different areas have different buying patterns at different times of the year.

Schwartz explains, "This system recommends what makes sense for different areas. It might tell us we have to mark down swimwear in the north in July, in the middle of the country in August, and after Labor Day or not at all in the South."

Demanding Growth

For O'Reilly Auto Parts of Springfield, Missouri, accelerated growth in the number of automobiles being manufactured and the tendency of Americans to keep their cars for an average of nine years has greatly increased the number of SKUs per store to as many as 40,000. That aggressive growth led the company to abandon its aging homegrown inventory optimization system in favor of Evant Demand Planning and Replenishment Planning from California-based Evant Inc., which uses sophisticated formulas to calculate item demand and safety stock.

The ability of the software to handle exception reporting via customer-defined alerts has been key to its success, says Michael Williams, O'Reilly's vice president of information systems.

Williams says, "We can monitor the sales of short lifecycle products because the system analyzes the trends as they happen. It allows you to monitor short lifecycle and seasonal items much more closely so you know when to start stocking less or transferring those items to another area once sales slow in the first area."

The alerts are so efficient, according to Williams, that inventory management personnel can manage about 90,000 SKUs, giving the company room for more growth.

O'Reilly also has integrated the Evant software with planogram software from JDA Software Group of Scottsdale, Arizona. Combining the capabilities of both helps Williams' staff merge the stock assortment for displays with sales for a particular item. "Say we want to stock 12 cans of WD-40 in a planogram because it looks better. By combining it with this system, we'll be alerted that when we fall below six, we should order an additional six to make the stock look good again," he explains.

The system has really paid off. During the initial two-year period in which O'Reilly's 10 distribution centers began using the software, the company reduced inventory by $66 million. Once all 1,000 locations throughout the United States and Canada began using it, inventory was reduced by another $20 million.

Williams attributes the additional $20 million savings to increased visibility into store inventories. "We were able to identify those items in specific stores that we could bring back and replenish our distribution centers with rather than ordering that product through the supplier," he says. In addition, Williams notes that the company's turnover rate for its distribution centers increased from 4.5 to an average of 6.5.

Williams plans to expand the software's use. Next up, he says, is incorporating vehicle registration data in regions where stores are located to help determine the items to stock in particular stores.