Retail Location Intelligence Grows in Sophistication and Usage

In the late 1960s, science fiction writer Arthur C. Clarke posited a now-famous law of the universe: “Any sufficiently advanced technology is indistinguishable from magic.”

As the rapid advancement of technology, especially given the ongoing progress of artificial intelligence, often makes this seem truer by the day, its implications in commercial real estate are often of a type that would have seemed like pure voodoo to CRE professionals when Clarke put forth this statement.

Here’s one example: 

A restaurant chain with over 3,500 locations in the U.S. is researching potential sites for additional stores in a certain part of Los Angeles. The chain has very specific parameters in mind, including a cash-on-cash return of at least 20 percent, at least $1.5 million in annual revenue, and a maximum cannibalization factor of the company’s other area stores of 10 percent.

Punch this information into a computer program, press a button, and moments later the computer produces a detailed map of the neighborhood with a series of misshapen blobs representing the exact areas — down to specific streets and even addresses — where a new store would produce these results or better.

In an ever-evolving boon for retailers, location intelligence has ascended to a level where platforms can predict retail success, including revenue and investment return figures, down to a specific site. These predictions are based on hundreds of demographic and psychographic factors as well as information such as competitive intelligence, complementary local businesses, local crime patterns, and the retailer in question’s own history.

As one example, consultancy Bain & Company markets a location intelligence program called Vantage that evaluates current and prospective physical sites for retailers, driven by global data sets and machine learning analytics that translate its findings into a user-friendly location visualization dashboard.

Bain, in a demonstration of the platform for Commercial Observer, produced the hypothetical example cited above, which also offered further predictive detail for the areas represented by the misshapen blobs on the map.   

Clicking on one such blob showed that potential new stores for the franchise in that area could be expected to produce annual revenue of between $1.8 million and $1.9 million and cash-on-cash returns of 25 percent, with the highest potential cannibalization of a previously existing franchise location of just 2.3 percent. It also noted, for a franchise with high levels of local saturation, exactly which stores will face cannibalization, and by what percentage.

The program works the other way as well, allowing retailers and brokers to drop a pin on any nearby location to adjust predictions accordingly based on information already provided. Predictions also can be based on additional details that can be as granular as whether there’s a traffic light nearby or on which side of the road the site would be located, all of which can impact foot traffic and sales.

“This is where it becomes really powerful,” said Francois Vayleux, vice president of global retail practice at Bain, “because you potentially have tens or hundreds of areas that could have high potential for a site, but this allows you to select an area and really assess how much money you can make based on the information available.”

Given the vast amount of location intelligence data and analytical tools available, including significant advances in these categories in just the past few years, being able to analyze locations down to the finest detail is quickly becoming both a valuable tool and an essential ability for any retailer or retail broker.

CBRE Executive Vice President Cassie Durand, who specializes in tenant representation for retail clients, recalled that, as the industry was attempting to carry on in the wake of the COVID-19 pandemic, any recently compiled data became quickly irrelevant. The manner in which people worked, shopped and lived had changed faster than it previously had at any point in our lifetimes.

Durand noted that, as she sought to determine the best way forward for her retail clients, her first new hire — which she characterized as an out-of-the-box hire at the time — was not a broker but a location intelligence specialist.

“I hired someone who could leverage data to help me answer questions about how the market was evolving day to day, because I knew that my clients needed a sense of confidence for themselves, their boards and their stakeholders to be able to make not even big decisions, but just the next decision — what do we do in the next six months, let alone the next three years,” Durand said.

Client confidence in company data, especially at a time when confusion swirled about the nature of even the most basic human behavioral patterns and interactions, was paramount both for servicing clients and as a competitive advantage, Durand said.

“What started as a brokerage business has evolved into a consulting business, and that is absolutely the goal,” said Durand, echoing a similar repositioning for brokers throughout the industry. “We want the client to feel like they’re dealing with a consultant who really cares about the data that is driving decision-making.”

This clarity and certainty regarding data and how to best analyze it is essential today, especially given the oversaturation of available data.

“In the past, it used to be digitally native brands that had all this data about their consumer at their fingertips,” said Alanna Loeffler, a senior managing director at Cushman & Wakefield, who works with digitally native brands branching into physical locations as well as traditional retailers. “Now, all brands have all sorts of data, and they have so much more on the consumer than they ever had before. I think that’s something that’s really shifted in the last few years.”

Sebastien Pavy, a partner in Bain’s Los Angeles office, also noted how deeply the data environment has changed of late.

“The big difference between 10 years ago and today is how we moved from ‘You should look for a site in this kind of very large, undefined trade area with a large radius,’ to ‘This is the corner, or the three great locations [you should look at],’ and much more specific and targeted information from types of store formats to specific store locations,” said Pavy.

Loeffler noted that the growing range of options for how consumers can spend their time and money, paired with tough economic times, has made possession of this knowledge essential for retailers.

“Given the macroeconomic environment, consumers are being so much more selective than ever before about where and how they shop,” said Loeffler. “It’s become leaps and bounds more important for brands to understand everything about that consumer — not just the demographics or household income, but who they are, where they live, where they work, and what other brands they shop for.”

While nearby competitors are an essential data point in site selection, this last point makes complementary retailers just as important in helping select locations that offer cross-shopping opportunities more likely to attract consumers.

Vayleux mentions the integration of detailed factors about other nearby retailers, both competitive and complementary, as significant data points in a Vantage analysis.

“At the end of the day, we want to use Vantage to identify the market around a specific store, identifying the dynamics in terms of competition and the demographics around a specific store that could explain its performance,” said Vayleux.

On the competitive front, Vayleux demonstrated how Vantage can select any store or potential location and overlay a slew of competitive information onto the site, including local malls, department stores, grocery stores and other restaurants.

“Here, for instance, I have a Subway, a Taco Bell, a McDonald’s and a Wendy’s,” said Vayleux, pinpointing a location on the visualization map and layering on other demographic characteristics such as population density. “That’s really important as we think about identifying the competitive intensity of a given area.”

One much-relied-on source of data and analytics is Placer.ai, which provides foot traffic information and interpretive analytics to clients throughout the CRE industry to reach conclusions about ideal sites.

“When retailers are evaluating new sites, there are a variety of ways our data could help them navigate that process,” said Elizabeth Lafontaine, director of research at Placer.ai. “One is really looking at the trade area of a new market as well as the trade area of their individual stores so we can help them home in on exactly where their visitors are coming from, because sometimes retailers have sort of a general idea — ‘This is my audience, this is how far they’re driving, this is where they’re coming from and these are the ZIP codes they live in.’ But we can help them home in on exactly who their core customer is, then they can apply that logic and reasoning to any potential new sites.”

The catch

While all this data, as well as the ability to deeply process it all, is important for retailers today, it hasn’t completely replaced the need for human experience, particularly the kind you get from having walked local streets to familiarize yourself with an area, and a history of working with a wide variety of retail clients.

Ben Weiner is an executive vice president at Ripco Real Estate who works with retailers such as Target, Best Buy, Chipotle, Bob’s Discount Furniture and Shake Shack.

When it comes to location intelligence, Weiner describes the pooling of data as a combined effort that includes each clients’ own proprietary location intelligence data and programs, foot traffic data from Placer.ai, and Ripco’s own information based on its vast experience working in New York City retail.

“While retailers are using their own programming to guide decision-making, we can then come in and say, ‘On this street, here are the sales that X, Y and Z tenants are doing. Here’s where they rank among their stores in the outer boroughs. Here is the subway ridership for the subway stop on their corner and how that has grown or declined over the last five years,’ ” said Weiner.

For all the ways technology has advanced in the area of data site selection, Weiner portrays this as the advantage for a retailer of having a broker who has engaged in old-fashioned shoe leather information gathering.

Weiner notes that while data points like ridership and vehicular data are public, the “gold” a good local broker provides for a retailer is often in their information about local sales and leasing.

“That information is the most important thing we have as brokers,” said Weiner. “When we’re gathering sales data, like what a restaurant is doing in a market or what they paid to be in a certain space, that’s us out there getting that information through relationships, and that’s extremely valuable.”

Durand noted that there could be many reasons why trusting data alone could risk missing crucial information.

“Data is wonderful for gut checking, helping you make sure you’re not missing anything and validating decision-making, but it should never stand alone,” said Durand. “It should be married with the art of what we offer, which is our local market knowledge and deep understanding of historical performance, including who’s coming, who’s going and why.”

One key metric data providers don’t measure, according to CBRE’s Durand and Placer.ai’s Lafontaine, is traffic around international tourism, the collection of which is restricted due to privacy laws in other countries. 

“There is a world in which we would have much better data around tourism. That would be my No. 1 wish,” said Durand. “That’s the biggest hole, because tourism is such an incredible driver, especially in urban markets where retailers are making much bigger decisions because there are larger rents. There’s more risk in making those decisions, [which are often being made] because of the perception that tourism is there.” 

Durand also noted that demos can be misleading in areas that may not follow predictable patterns or trends.

“If you look at demos of South Florida, you wouldn’t think of the Greater Miami area as dense or as affluent as it is, because you have a lot of transient snowbird-type behavior in that market,” said Durand. “So you really do need to marry the art with the science.” 

Peter Braus, president of commercial real estate services firm Lee & Associates, cited this sort of detailed local information as the one area where technology is often still lacking, especially in a sprawling market like New York City. He also noted that such details are often some of the most useful information in selecting a site. 

“The one thing everyone asks for, the holy grail that we haven’t been able to get in a meaningful way, is the ability to track individual store sales,” said Braus. “That’s something every tenant wants, [but that you can’t obtain] unless you’re dealing in a mall setting where tenants have to report sales to their landlords. Here in New York City, that almost never happens, so you’re at the mercy of guesstimating what a store is going to do.”

Braus also noted other limitations in using technology to evaluate retail possibilities throughout New York City. 

“In markets outside of Manhattan, where I do a lot of business, technology tends to come in half-mile, mile or two-mile swaths,” said Braus. “But that’s not really applicable here, because a half-mile in New York City puts you in a totally different neighborhood, and oftentimes things look very different one block away.”

Braus said that he fills data gaps with information from Placer.ai and other sources that help isolate shopping patterns.

But Weiner noted that even the universally utilized Placer.ai data does have one weakness in urban environments in that it can’t distinguish between different floors for high-rise buildings.

“When you’re in an urban environment and there’s 50 floors to a building, but you’re looking at just the ground-floor space, you have to keep in mind the data you’re seeing from Placer might be for the entire building or something on an upper floor, not just the ground floor,” said Weiner, who added that this reinforces the importance of on-the-ground knowledge. “Where we add value is that we know the markets on a granular level.”

(In response, a spokesperson for Placer.ai noted that in dense urban situations, “the platform provides additional context and notes that certain properties in ‘dense areas, multistory buildings, or similar settings’ are handled differently from standard venues. For those properties, we offer a separate report type that measures activity within a radius of the property instead.”)

All of this reinforces the idea that, for all the incredible advances in data and technology, sometimes the most important tools are more traditional.

Durand, stating that the plethora of data can sometimes raise more questions than it answers, believes that adding a retailer’s own long-gathered information on their customers to the rest of the available data is the optimal blend for finding the perfect site.

“The special sauce is when your client has data you can leverage,” said Durand. “It makes the output clearer in that you’re layering in existing customer behavior with the broader market, and there’s a lot of power in that. For companies that are direct to consumer, you could have one that’s 20 years old that’s never opened a store, so they have a ton of information [about their customers] but don’t have information about how their customer interacts with retail. 

“That’s where the data can be that much more impactful, because you don’t have the comp set around store performance to look at alongside your e-commerce performance. That’s where we get super deep into the weeds with our clients to help them build out their thinking around how their customer is ultimately going to behave in the market.”

Larry Getlen can be reached at lgetlen@commercialobserver.com.