Adecade ago, the field staff at Dunkin' Brands would find promising new locations for their Dunkin' Donuts, Baskin Robbins and Togo's stores by counting vehicles and pedestrians while standing on a street corner. Back then, the company typically opened 300 stores annually across all brands.

By comparison, the Canton, Mass.-based company today opens more than double that amount of stores annually. In fiscal year 2005 that ended Aug. 31, Dunkin' Brands opened 772 locations, and the firm plans to open 800 sites in fiscal year 2006.

Why the spike? In the late '90s, Dunkin' Brands hired site selection specialist geoVue of Woburn, Mass., just as technology began changing how demographic data was disseminated and analyzed.

Using geoVue's technology, Dunkin' Brands is able to quickly plot the information that is most critical to site selection: morning traffic counts, the number of nearby residential communities, shopping centers and offices, and other pertinent data. The franchisor ultimately aims to open 1,000 operations annually.

“Retailers have always tried to figure out the best sites, but it's such a cutthroat industry right now that any advantage you can have in market and site analysis is an advantage you're going to take,” says John Dawson, chief development officer for Dunkin' Brands, which has 12,000 locations worldwide. “If you don't, you're at a distinct disadvantage.”

Industry comes of age

The site selection industry's technological advances over the past decade are analogous to the evolution of flight, from propeller airplanes to rocket ships. While demographic data has been available for years, retrieving and converting it to a workable form for analysis, such as spreadsheets, was generally cumbersome and tedious. In fact, site selection experts paint a picture of information technicians controlling the entire process behind closed doors.

But technological breakthroughs have given site selection companies such as geoVue, MapInfo, SRC and Claritas speedier and easier access to the information. The firms that supply data to site selection companies are building bigger and more sophisticated databases, focusing on once-vague variables such as consumer spending habits.

Today's site selection providers are feeding large amounts of detailed information into powerful computer software programs. The software slices and dices the data in a matter of minutes to spit out predictions on optimal locations and potential retail sales.

The software can even provide analysis on which non-performing stores should be shuttered, or which stores will be cannibalized by store expansions.

Retailers also are using site selection firms to identify the characteristics of their core customers through surveys. Integrating that information into the software helps retailers pinpoint geographic areas that match their customer profile.

“We've seen tremendous lift in a user's ability to make better decisions, avoid mistakes, and open up the right number of stores in a market — and open them faster — by using these tools,” says Michael Kesselman, executive vice president of corporate development for geoVue. “They allow the decision makers to screen opportunities much more effectively.”

Wireless technology now gives laptop computer users the ability to access and analyze real-time data while out in the field. That's a far cry from a decade ago, when a company's real estate staff or other personnel evaluating locations generally had to wait while statisticians in the corporate office retrieved and analyzed data.

Olivia Duane-Adams, a founding partner of SRC, based in Orange, Calif., recalls that 10 years ago when she submitted official request forms for demographic information on a particular location, it often took days before she received the results from analysts. “Now I can produce those reports in a matter of mouse clicks from home, a hotel room or a client's office,” she says.

Confirming gut instincts

Hiring the services of site selection companies is no guarantee that a retailer or restaurateur will succeed at a particular location, of course. Poor service or inept management are recipes for failure even at gold-plated sites.

During the first half of 2005, national and regional retail chains announced 1,908 store closings, according to the International Council of Shopping Centers (ICSC), a shopping center industry trade association based in New York. While that was nearly a 50% drop in store closings compared with the same period in 2004, the failures still illustrate how identifying the best potential locations can give expansion-minded chains a running start.

But site selection is still part science and part art even with advances in technology, experts say. There is no substitute for walking a site to make sure it meets the user's needs.

Practical applications

Charlotte, N.C.-based Meineke Car Care Centers, a franchisor that has some 900 locations across North America and South America, for example, provides its real estate field staff with real-time demographic data, such as population density and cars per household in the surrounding area. The field staff evaluates and compares several different sites a day. Previously, workers on the ground would request demographic information on the sites from analysts in the corporate office, and then wait for the data to be relayed before they could begin to make comparisons, says Jim Alderson, director of new store development for Meineke.

“It improves our decision-making process and our speed to market because we don't have to wait for somebody in a remote location to do the analysis,” he adds.

In July, Meineke hired geoVue to better define trade areas at potential locations. Historically, retailers typically have drawn one-, three- and five-mile rings around potential sites to determine trade areas. But those concentric rings are giving way to drive times, says Kesselman of geoVue, because drive times account for barriers, such as lakes.

Site selection companies also are building models that determine how far customers are willing to drive to a retail location in a given geographic area.

In addition to embracing the concept of drive times and moving away from concentric mileage rings, Meineke has added components to geoVue's standard iSite site selection package to determine what similarities its most loyal customers share with each other in terms of income, education, proximity to an operation and other characteristics.

Once Meineke establishes who its core customer is, Alderson hopes the technology helps the company find so-called “sweet spots,” or locations where the surrounding population share similar attributes.

By zeroing in on those sites, the thinking goes, Meineke is likely to tap into areas where dwellers and workers are inclined to become good customers, which ultimately will improve a new operation's chance for success.

Cost benefit hard to quantify

Dawson of Dunkin' Brands and Alderson of Meineke both say that it's difficult to calculate a return on investment or the savings site selection technology generates — and they also declined to say how much they pay geoVue. But the potential reward for avoiding a bad site and making a million-dollar mistake is obvious, they emphatically agree.

Plus, in a market where real estate is in high demand, the latest site selection technology gives the franchisors the ability to find and lock up locations quickly. That's particularly crucial in the automotive repair business, where it's tough to find sites with the right zoning, size and cost, Alderson adds.

“Sites that fill our needs, especially if they're existing facilities, go pretty quick,” he says. “So when you find a good location, you want to be able to react very quickly and get it under contract. To have that information is power.”

How much users pay for site selection technology ultimately depends on their needs, and providers offer a broad range of tools and services with just as broad of a price range. For instance, a user of SRC's basic package, which provides mapping capabilities and updated demographic estimates based on 2000 U.S. Census figures, costs about $1,000 a year, according to Duane-Adams.

Companies that use SRC's customized and most advanced modeling programs delivered over the Internet, on the other hand, can expect to pay upwards of $1 million a year, depending on the number of people using the system, content and functionality, adds Duane-Adams.

Other site selection companies have similar cost structures. Integras, the site selection division of San Diego-based Claritas, was scheduled in November to officially launch its new Prime Location product, a solution that provides a whole host of analytical modeling capabilities, according to the company.

A basic implementation of Prime Location could cost about $40,000 a year, while a more in-depth modeling and analytical system may cost upwards of $1 million annually, depending on the software, data and analytical packages deployed, says Ray Major, senior vice president of Integras.

“We like to think of this (technology) as an objective person sitting at the decision table with real estate guys who may have very different opinions,” Major says. “This tool also has an opinion, but one that's very objective. So you can reject it or not, but the bottom line is that it's already done a tremendous amount of work for you.”

Why customer surveys matter

Users are expressing demand for high-end systems. In 2003, for example, the Glendale, Calif.-based family restaurant chain IHOP Corp. shifted away from a centralized real estate strategy to a franchisee-financed development model.

The switch reduced management wrangling over future sites and sales forecasts. To support the new strategy, IHOP hired MapInfo of Troy, N.Y., to provide objective and analytical ways to evaluate opportunities for new and existing restaurants.

MapInfo surveyed more than 60,000 customers across 100 of the company's 1,200 restaurants in the U.S. and Canada. MapInfo then plugged its findings into a model that ultimately gave IHOP the ability to determine the sales potential at prospective sites and whether existing restaurants were properly positioned to capture their full market potential.

The results also are helping IHOP figure out the ideal number of operations for a geographical area, the competitive landscape, and which locations the company should close, renovate or relocate.

Rick Celio, vice president of franchise and development at IHOP, in a prepared statement said that IHOP now better understands what its customers like about the restaurant from a market and site-location perspective. Meanwhile, Mark Zygmontowicz, managing director of sales for MapInfo, acknowledges that not all users will benefit from powerful systems that are heavy on analytical modeling, particularly companies such as small regional retailers.

Still, utilizing even modest doses of site selection technology should pay off for those firms, given the alternative of going on gut instinct and failing. “A lot of retailers come to the conclusion that they need site selection technology after the heartache of opening stores that five years down the road they wish they hadn't,” Zygmontowicz says. “That's always a triggering event for them to say, ‘We've got to get smarter at this.’”

Joe Gose is a Kansas City-based writer.