Supermarket executives have adopted unified market knowledge systems to improve the murky and difficult process of store site selection
My hometown, Dumont, New Jersey, a solidly middle-class town of 20,000 people, is currently at the center of a vicious, years-long turf war. The combatants? Pharmacies.
CVS struck first with a store on Washington Avenue, and then expanded west into New Milford. Walgreens volleyed with stores to the east, and further south in Bergenfield. Then a dark horse emerged, Rite Aid, establishing positions to the heavily trafficked south. Walgreens countered with another location on Washington Avenue, smack in the middle of CVS and Rite Aid territory. It's become Game of Thrones, with pill bottles.
Dumont and its neighboring towns happen to be a gold mine for prescription-fillings. For every square half-mile in the area, residents are filling more than 80,000 prescriptions per year. In Dumont alone, that means there are at least 320,000 prescriptions filled annually. Eighty thousand is a magic number for national pharmacies -- it's the bare minimum required before entering a market. There are eight pharmacies within two miles of Dumont, but the area's annual prescription demand suggests it could support four or five more. This battle will rage on.
I came to learn all this while test-driving TAS Unity, the flagship product of a tech company called Trade Area Systems. TAS Unity is an example of a unified market knowledge system (UMKS), a relatively new technology which large retailers and commercial real estate developers are using to improve the murky process of site selection. Say CVS wants to open yet another store near Dumont. It could use UMKS to help choose its next location.
UMKS is not an overtly sexy product -- it's simply a package of hardware, software, data and Web/mobile apps. But it comes with a compelling value proposition: UMKS can help retailers make better, faster decisions about where to build. Based on the conversations I've had with developers, retailers and researchers in the commercial real estate business, this makes UMKS a very exciting idea.
* * *
Site selection is difficult. It can require thousands of dollars and hundreds of man-hours to simply evaluate a site. Companies take the process very seriously, because poor site selections wind up becoming very costly mistakes. Consider all the expenses involved with opening a store: research, broker fees, property, permits, construction, staff, promotions, marketing. The threat of failure looms over everything: How is a brand affected when a store opens and then closes?
- Supermarkets assume any customer more than 8 minutes away will find their store inconvenient.
- 55% of the products in Super Centers (combo discount department and grocery stores) are supermarket items.
- 60% of home improvement superstore customers are women.
- Supermarkets assume each person spends about $50/week on groceries.
Things weren't always this way. Back in the 1950s, when retailers first started fanning across the United States, site selection was kind of easy. Commercial real estate was a personality-driven business back then, more art than science, filled with men (they were almost all men) who could pick a good spot by the look of the land or the smell of the soil. Dolores Hayden writes in Building Suburbia that Ray Kroc of McDonald's used to cherry-pick locations for his next burger stand by flying over suburbanizing areas in a Cessna.
Thanks to a number of factors, the personal approach proved surprisingly durable. Suburbanization certainly made things easier: new, wide-open markets spawned as urban downtowns were abandoned. The Interstate Highway Act of 1956 was key, as well. It accelerated the development of shopping centers and malls, and created the need for thousands of new roadside businesses.
According to Hayden, a third and less-known factor was perhaps the most important. In 1954, Congress reduced the straight-line depreciation period of greenfield income-producing property from 40 years to seven years. This meant that new commercial properties built on virgin land -- the kind of land readily available in new suburbs -- suddenly became near-instant tax write-offs. (The depreciation rate stands at 39½ years today.) The loophole created huge subsidies for developers and turned commercial real estate development into something like a license to print money.
The gold rush ended after the real estate crash of the late 1980s. The easy money was gone, and the landscape was saturated with commercial properties. The advent of real estate investment trusts, or REITs, made things more complex. Basically conceived as straight-up investment portfolios that contain buildings instead of stocks, REITs forced developers and retailers to more carefully consider return on investment. The situation's even tougher in today's post-recessionary climate. Expense, market saturation, risk aversion and a slowed residential construction industry (new homes spawn new stores) all complicate site selection.
Then there is the problem of data overload. According to Peter Patnaude, head of research at SRS Real Estate Partners, the largest retail real estate services company in the United States, there are "thousands" of potential variables that can be taken into account when choosing a site: age, race, gender, income, credit scores, crime statistics, regional voting records, divorce rates, traffic patterns, loyalty card information and weather just being the teeniest tip of the iceberg.
But the would-be customer is only one part of the equation. Retailers must also have intimate knowledge of their own locations -- no small task when you're talking about thousands of properties -- as well as the locations of their competitors. Some of this information lives in databases, and some of it lives in the heads of people within the companies. This presents another problem. When employees retire or, worse, move to a competitor, they walk away with a tremendous amount of knowledge.
As its name suggests, a unified market knowledge system is designed to aggregate all of this information. Everything a retailer might use to make a site selection decision -- all the data, market research and competitive intelligence -- is loaded onto a server, which can be accessed by anyone involved in the site selection process. Of course, servers have been around forever, and companies like Trade Area Systems and the GIS services provider ESRI have long offered technical solutions to help retailers manage their data. So why is UMKS catching on now?
Eighty thousand is a magic number for national pharmacies -- it's the bare minimum required before entering a market.
According to Mary Lou Fiala, former Chairman of the International Council of Shopping Centers, the advent of mobile technologies, especially the iPad, has been a major catalyst. "It's just changing everything," she says.
Using an iPad, individuals out in the field can not only access information, they can also update incorrect data, add notes about properties or competitor locations and use the iPad's built-in camera and GPS to upload geo-tagged images to the system. The combination of better, more accessible data allows companies to speed up the site selection process. Peter Patnaude says that analysts at SRS are completing research projects that once took days in a matter of hours.
Mapping is another key feature of UMKS. Companies have long used mashups of maps and data to help inform their decisions. One of the most enduring examples of this idea is something known as a ring study. To make a ring study, you start by dropping a pin at a point where you might like to build. Using the pin as your center, you then plot data in relation to the pin. A basic ring study, for instance, might show the number of potential customers living within five, 10 and 15 miles of a potential site. Instead of miles, drive times can be used similarly.
Ring studies are useful tools, but they're only as good as the data upon which they are built. UMKS, in turn, helps companies produce far more accurate and dynamic maps, providing a better picture of the world from the retailer's point of view.
* * *
One thing becomes clear when you look at all the maps and data associated with a site selection: commercial real estate is all about perspective. The same factors that make an area poisonous to one retailer make it ideal for another. Discount retailers, for instance, have been among the few to aggressively expand following the recent recession. These businesses sometimes see opportunity in surprising data, such as low credit scores and high household debt rates.
In this way, it's possible to see retail locations as representations of the communities in which they are built. They're visualizations of demand, manifestations of data. In the case of Dumont, my hometown, the glut of pharmacies paints a certain kind of picture: With an annual prescription demand of 16 per capita, Dumont residents fill far more prescriptions than the average American (11.9). Is this because of an aging population, high rates of access to insurance and medical coverage or simply the fact that people in the area are ... not well? I do know this: there are no retail health clubs in Dumont.
In helping retailers manage data and map demand, UMKS has significant implications for how the business of commercial real estate is conducted. Decisions can be made faster, new markets can be revealed. The question, then, is what do these decisions and markets say about us?
Images: 1. A screenshot of the software shows a ring study; 2. Desire lines, which show where a store has the strongest pull, for Walmarts in the Las Vegas area.