What are a Refrigerator's Contents Worth?

Linking marketing research and zip code data `truly amazing'

Location, location, location goes the old real estate mantra. Well, geography matters in the market research business, too. Whether you are introducing new products, growing existing brands or trying to understand market trends and consumer habits, knowing where your customers live can help your company to become more profitable.

That's why businesses are increasingly turning to Geographic Information System (GIS) applications to more effectively link geographic and descriptive information on the customers in their databases. In our case, we have geographically coded our large and diverse consumer panel. We have also coded each household with the popular Prizm system, giving marketers a powerful problem-solving tool to pinpoint, among other things, hard-to-reach customers.

Prizm (Potential Rating Index for Zip Markets) codes were developed by the Claritas Corporation to group zip codes into 62 socio-economic clusters. Using address and zip-code information, we now assign Prizm codes to each of the 400,000 households in our home-testing panel. Linking socio-economic and geographic information makes using our panel-based consumer surveys even more efficient, and the results produced more powerful.

Example: A national department store is considering closing one of its locations, which it believes to be too upscale for its particular brand. By studying the Prizm codes--and demographics of the home-testing panel members who live in this area and shop at this store--we can develop a detailed shopper profile for this particular location. And compare it to other locations nationwide.

The information available at the zip code level is truly amazing. For instance, because we know what people purchase, we know what's in their refrigerators. How useful can that be, you ask?

Example: Con Edison, the public utility company that supplies New York City, was faced with a dilemma. The State of New York requires public utilities to compensate customers for losses caused when electricity is out and the customer loses the contents of his or her refrigerator. In the past, claims had been limited to $100. But a proposed increase to $400 would have cost Con Edison millions of dollars annually. Con Edison turned to us with a simple question: What is the value of the contents of refrigerators across New York City? A GIS-based model was developed to evaluate randomly selected zip codes of all socio-economic areas across the city. Our methodology was evaluated by outside reviewers and presented to the NYC's Public Utility Commission. Based on our findings, the commission established $125 as the new maximum customer claim, saving Con Edison millions of dollars each year.

What else can we do? Using a geographically coded (GIS-enabled) database of 11 million U.S. businesses and 14,000 shopping centers, we can examine local retail competition in enormous detail.

• Are there other store locations with similar shopper profiles?
• Do those locations perform better or worse than the location about to be closed?
• Does store performance depend primarily on shopper profile, or on local competition (e.g., no other department stores are within close driving distance)?

Recently, we were able to tell a pharmaceutical industry client how socio-economic conditions are linked to the type of ailments people suffer, and how people in different groups seek treatment.

Using geographic information can also create research efficiency. When we know where your customers live, we also know how close they live to your restaurant, grocery store, mall, golf course, or bank. In fact, we can map out the driving route of a consumer to a particular location, and measure the driving time. This makes it possible to pinpoint customers that are hard to reach and invite them to stop by an area mall and evaluate your product. By selecting, in advance, customers within reasonable distances, we can improve response rates significantly, and reduce research costs.

Now that geographic information can be integrated into marketing research, the possibilities are truly boundless.