PRINT THIS STORY:

CHANGE TEXT SIZE:

DOES IT COUNT?

SOME LANDLORDS QUESTION THE ACCURACY OF MALL TRAFFIC-COUNTING SYSTEMS

By Curt Hazlett

Bill Martin, co-founder of ShopperTrak, is blunt when asked why some shopping center owners shun the traffic-count technology his company offers as a property-management tool. When malls suffer traffic declines, dropping the tracker data gives the owners “some plausible deniability,” he says.

“Back in the late 1990s, total property counts were good, but in the early 2000s we started to see declining mall traffic numbers,” Martin said. “The last thing [an owner] wanted to do was suggest that a portfolio was declining because of a traffic-counting number, so in some cases mall developers who had used traffic counters for years have fallen away from using them.”

Simon Property Group, a former ShopperTrak customer, offers a different yet equally blunt explanation for why it no longer uses the technology. “I would have to tell you that our experience [with traffic counting] was as deep as anyone’s, and we walked away from it,” said Scott Mumphrey, president of Simon Management Group and executive vice president of the parent company. ShopperTrak “can say that we stepped away from it because it was saying something we didn’t want to hear, but I would tell you that I don’t have any idea whether what it was saying was anywhere near accurate.”

Others have a similar take. Among them is David Goldberg, Taubman Centers’ vice president of marketing and sponsorship. “We don’t use it,” Goldberg said. He is uncertain whether Taubman has ever even tested the product, in fact. “I wouldn’t go as far as saying it’s not valuable, but I would say from our perspective that we’ve never been convinced of its reliability or its total accuracy. That’s been a big hurdle for us to get over.”

Traffic-counting technology in one form or another has been around since the 1960s, and it has found varying amounts of acceptance from centers owners and retailers. Among the large retailers that use it are American Eagle Outfitters, JCPenney and Zale Corp., and all of them are customers of ShopperTrak.

The applications vary between stand-alone retailers and shopping centers. Retailers typically use the tool to determine conversion, the rate at which customer traffic translates into sales. “Without traffic counts, retailers simply can’t calculate their conversion rates,” said Mark Ryski, CEO of Edmonton, Alberta-based HeadCount Corp., whose customers include Home Depot and Sears. “And without customer conversion rates, retailers simply can’t understand how well they are performing.”

Shopping centers, on the other hand, usually employ the device as a way to determine staffing levels, to judge the value of promotions and to set rates for in-center advertising, Martin says.

Despite its promise, traffic-counting technology “most certainly can be a bit of a tough sell,” said Ryski. He estimates that only 25 percent of North America’s retailers use it. Martin says ShopperTrak is used in some 300 enclosed malls in the United States, among them Macerich and Westfield Group properties.

Simon, the country’s largest owner of shopping centers, once saw traffic counting as a valuable tool. In 1997 the firm signed a seven-year deal with RCT Systems for traffic-counting technology. “Our intent at that time was to see if we could develop some reliable information that could be used in managing our business,” said Mumphrey. RCT merged with ShopperTrak in 2001, and ShopperTrak is now the largest provider of traffic-counting data in the industry.

As Simon began looking at the data, weaknesses began to emerge, Mumphrey says. One area of concern was the reliability of the overhead cameras that determine counts. “Movement of a planter or a trash receptacle, or shadows that might come in as the daylight changed, could affect the count,” he said. “In some cases it caused a constant count, in some cases no count. When you have as many as 40 or 50 cameras in one shopping center and you start to get gaps in the data, you have to go back and re-create what you think was happening while that camera was not operating.”

In the end, the firm opted out. “There were too many gaps for us to reach any kind of comfort level that what we were getting was meaningful,” Mumphrey said. Simon declined to renew its agreement with ShopperTrak upon expiration in 2004.

“No technology is 100 percent capable of capturing everything, and no store is a completely closed loop,” Martin conceded. And he acknowledges that such problems have moved some to drop the systems. “But there are others who maintain them and use them and understand them, and use the information to make business decisions,” he said.

Mall operators in Europe have been more willing to accept tracking, says David Smyth, vice president of sales operations at London-based FootFall, which also does business in the U.S. Because European shopping centers typically have long, fixed-term leases rather than the revenue-based leases common in the U.S., mall owners there were able to gather less data about their tenants’ performance, says Smyth. “Once they had a retailer in they were kind of stuck,” he said. “They didn’t get much information unless a retailer wanted to give it.”

As a result, European mall owners have been more willing to employ extensive tracking in an effort to determine how well the properties were operating, he says. “A lot of the systems we’ve put in the U.K. and Europe have actually measured where people moved along the ground floor, second floor or third floor, not just where they entered,” Smyth said.

With U.S. shopping centers increasingly pushing for in-center advertising revenue, the issue of traffic counts may take on greater importance, Martin says. A center’s advertising rate depends on the number of people who see it, and yet there is no common way of determining that viewership, he says. “Mall developers have not united on a common approach on how to value their ad space, but the currency is going to be traffic no matter what happens,” Martin said.

Taubman’s Goldberg says there is a better approach. “When it comes to traffic, we are more supportive of some of the industrywide initiatives,” he said. Goldberg cites the work of the Scarborough Research and Arbitron media research firms, whom he says are working to create a rating system “that would allow sponsors and potential advertisers to look at a mall in a manner that’s more familiar to them, similar to media rankings. Between that, which would be of interest to a lot of advertisers, and the sales per square foot figure that would be of more interest to traditional tenants, we haven’t seen the compelling need for traffic counters.”

Nevertheless, Martin maintains that shopper tracking can indeed provide valuable information. “If your fundamental purpose for being there is to bring people to your tenant base and provide services and you’re not measuring how many people are coming in,” he put forward, “then how do you know if you’re successful and if the efforts and dollars you spent yesterday are having any effect today?”

But Mumphrey says Simon prefers to focus less on traffic as a measure of performance and more on the fundamentals of occupancy, rents, sales and operating expenses. “Traffic counting is somewhat of an indirect measure of performance,” Mumphrey said. “We’re a public company, and our results speak for themselves. If you look at our track record, you can see that we shouldn’t be afraid of any data that’s going to come out about our company or industry. We much prefer to put our focus on direct business fundamentals as opposed to those that are indirect and haven’t had a clearly established correlation in terms of what happens with traffic and sales.”

Anyone considering traffic counting should ask a few basic questions, Mumphrey says: “What specifically is being included in the sample and what isn’t? Will the data be audited and verified by an independent third party? Are there any blind spots in the data caused by hardware or software? What percentage of the data is simulated or re-created?”

Such questions may be tough to answer, Mumphrey says, but “when you say this is going to predict behavior and performance, you should be held to a high standard in proving that information is accurate.”