Home Depot Relies on Predictive Analytics for Reliable Location Intelligence

Home improvement retailer Home Depot is finding that many markets are becoming saturated for its “orange box” stores, and yet it opens a new one every 48 hours. With more than 2,000 existing stores yielding 260 million square feet of retail space, this $81 billion company relies heavily on location intelligence for its site selection strategy. In order to continue its expansion, Home Depot must innovate with different store format concepts and find locations where those concepts can yield a high rate of sales per square foot, in addition to squeezing sales from existing locations.

Mike LaFerle, vice president of real estate, says, "Location is everything ... convenience is the driving factor for most customers." So, to do the heavy lifting of market analysis, Home Depot relies on the advice of MapInfo. This long-standing relationship has led to exponential growth during the mid 1990’s. Today, Home Depot looks toward executing its expansion goals into new markets, urban environments, and internationally.

Since the company’s inception in 1978, Home Depot has utilized the expertise of MapInfo's retail consultants (then Thompson Associates) to provide a site selection strategy for its "mega store" format: large aisles; varied inventory; comprehensive in-store centers, such as the garden and home decorating centers; all under one roof. As such, Home Depot has become a "category killer" in home improvement and is the second largest retailer in the United States, after Wal-Mart. Prior to the big box concept, "mom and pop" hardware shops were the mainstay of this retail category. Most of those have disappeared and some of the niche hardware retailers that had franchised are now being acquired by the big box companies.

Still, LaFerle faces the daunting task of keeping the pressure on his competitors, growing the business, and making certain that the existing stores increase their average sales receipts - a metric most retailers use to understand customer loyalty and to verify that the store is offering the correct merchandise mix. Even with the home improvement market estimated at over $1 trillion each year, Home Depot is looking at ways to uniquely position itself for growth. LaFerle’s strategy is to enhance the core business with distinct merchandise, modernize stores, improve the customer experience and extend the business into areas that serve the professional home builder, as well as offering an online catalog of products.

To do all of this, MapInfo works closely with LaFerle’s group in just about all aspects of its business. "We do just about everything for Depot," says Bob Buckner, vice president of MapInfo and author of the book, Site Selection. Key projects include looking at how many stores a market can support, evaluating point of sale data, and supporting the merchandising function. "We’re not quite at the SKU [stock keeping unit] level yet but we’re getting there," says Buckner. Buckner explains that Home Depot is very much engaged in the store location modeling process and often challenges the forecasting methods, resulting in an excellent dialog between the two companies. "We like when a customer understands what we’re doing," comments Buckner. "In fact, we’re most fearful when a customer just wants the answer."

For example, sales forecasting models showed Home Depot that the garden center section of its stores could be expanded. For 2006 and beyond, it is looking to add 400 to 500 new stores and expand internationally into China. The company just opened an office in Shanghai. In addition, every year all existing stores go through a rigorous process of evaluating their positions in the market. However, LaFerle is keenly aware of the need for data with more currency. In high growth areas such as the southwest and Florida, building contractors are putting up homes in a 90 to 120 day cycle. For Home Depot, the process of siting a new store, building it and merchandising the store floor is a 42 month engagement. That means the sales forecasting, based on the most current demographics and sound spatial interaction models, must be able to closely predict when the market can support the investment in another store unit. Ideally, that will be well before residents start moving into new homes and planning their first trip to the "orange box."