Measuring retail foot traffic has come a long way. Traffic is one of the oldest forms of data to track a storefront’s popularity. It’s important because, after all, if no one visits a store, nothing is going to get sold. Foot traffic data is used in several ways and it’s been more helpful than ever in recent years as the pandemic shifted shopping online. But despite all the frenzy about traffic data from many in retail, not all experts think it’s the holy grail. Tenants and landlords increasingly look at the data together, part of an evolving relationship between the two parties. But some sources told me that while footfall traffic is an important data point, it’s not the only one, and real estate owners should be careful not to put all their faith into it.
Foot traffic refers to people who enter the store, but it’s not necessarily indicative of sales. People go to stores for many reasons, some of which have nothing to do with buying anything. They may search for products, compare prices, use the restroom, or meet friends. Some retailers make conversion rate calculations, which is the relationship between the number of visitors and purchases. The conversion rate is a better indicator of profits and losses, plus it enables retailers to make more accurate business plans.
Regardless of the debate about it, foot traffic data collection in retail has rapidly increased in the past decade. E-commerce has always had the edge over brick-and-mortar when it comes to online shopping since patterns are more easily tracked and measured. But now, new tech enables brick-and-mortar to gain better insights into both how many people visit a store and what they do when they are inside.
Traffic data always has a certain amount of noise. Interestingly enough, a concern for retailers and vendors is counts generated by children. Some vendors claim to provide height settings on sensors to distinguish kids from adults or eliminate them from the data collection altogether. Foot traffic data can be helpful, but actually collecting the data, it turns out, is a challenging process.
From old school to new school
Many old-school methods for capturing footfall traffic are still used. For example, some retailers still use beam counters to track foot traffic, but they can be wildly inaccurate. Beam counters use sensors to record how many people cross an established entry or exit point threshold, so they’re often mounted on a ceiling in a doorway. The problem with these people-counting devices is that they can’t tell the difference between shoppers and employees, so numbers can be inflated by too much ‘noise’ in the data. Some retailers even use more low-tech ways to count traffic, such as manually tracking shoppers on surveillance footage or having door greeters do a manual count. These methods can work, but they’re labor-intensive and not the best use of staffing resources.
Beam counters and manual data collection methods can work fine, but it makes it hard to trust the data since it can be incomplete and inaccurate, making measurements of store performance less reliable. Some retailers with deeper pockets opt for higher-tech cameras, but they can get expensive and complicated. Even with the best device money can buy, the cameras and sensors must be mounted correctly, properly calibrated, and constantly monitored. Some vendors claim that counters are 95 percent accurate, but that’s sometimes a theoretical number based on lab tests, not how they perform in a store.
Newer traffic-gathering tech goes well beyond old-school methods. Computer vision is an AI-enabled tech that’s helping some retailers gather foot traffic data and customer behavior analytics. Computer vision lets digital camera systems understand the content of images, similar to how people use their eyes to perceive things around them. Cameras and sensors that use computer vision can analyze the average shop times and shop times of individual customers. The technology can also drill down various hot spots and dead zones within a store, and customer ‘dwell times,’ which refers to the length of time a customer looks at a merchandise display. Computer vision enables retailers to gather more advanced and real-time data that can help in various ways, including determining the proper associate-to-customer ratios and how to optimize how merchandise is laid out in a store. During the height of COVID, some retailers also used computer vision tech to track real-time store occupancy data that could help them meet social distancing requirements.
Traffic isn’t sales, though
Retail tenants and landlords are increasingly looking at foot traffic data together, according to Rachel Elias Wein, CEO and Founder of WeinPlus, a real estate consultancy specializing in how the change in consumer behavior impacts commercial real estate. Wein said that there has been a more ‘symbiotic relationship’ between retail tenants and landlords over the last several years and that landlords are doing more to help retailers be successful. Some landlords and mall owners may even offer data services to their tenants. Sergio Gutierrez, Head of Revenue at RetailNext, a retail analytics provider, told me, “The landlord-retail tenant relationship is evolving. It’s become more of a partnership lately.”
Retail landlord and tenant relationships have traditionally been transactional. But with the growth of e-commerce, landlords and tenants have had to think of better ways to draw shoppers to brick-and-mortar stores. Landlords and tenants are experimenting with lease terms, trying shorter leases to allow stores to new concepts with less risk. Landlords are also using concierge services to entice shoppers, generate buzz, and improve foot traffic. Services like valet parking, curbside picking, and daycare have become much more common. In the battle against e-commerce, these in-store conveniences are much more expected among the average shopper.
While foot traffic data collection and analysis has become increasingly popular for retail owners and tenants, Wein said the data should be taken with a grain of salt. “The data is helpful, but it’s not an indicator of sales, and sales aren’t an indicator of a stores’ profitability,” she said. Traffic is a data point, but it may not be the most important, and without measuring conversion, traffic data has limited use for retailers. For example, Wein said a grocery store next to a college may have higher traffic patterns but lower ticket sales than a suburban store with families buying in bulk. And even if sales were more closely linked to foot traffic, a retailers’ profitability depends on much more, including overhead costs like staffing. Another example would be relying solely on traffic counts such as page visits on a website. It’s great if a web page gets many visits, but that data doesn’t tell you how long the visitor was engaged, what links they clicked, or if they bought anything.
Wein isn’t keen on shelling out big bucks for expensive in-store sensors and other tech when retailers could just use point of sales receipts. Sensor and camera technology also have downstream effects. Wein gave the example of cleaning robots in stores, they sound flashy, but an associate may not know how to fix one if it breaks. An associate with a mop is more efficient. The same goes for the newer sensors and cameras that measure footfall. Vendors may promise the world, but this equipment requires close attention and maintenance, and the accuracy of measurements may still get called into question. “Those are the downstream effects of new technology you have to think about,” Wein said. “Even if the tech sounds exciting, it doesn’t mean it’s ready to be used, and even when it can be beneficial to deploy, there are always downstream impacts to consider.”
Depending on traffic data too heavily causes other problems, too. Some retail tenants will tie staff manager bonuses to traffic data, prompting employees to fudge numbers to make them look better. Nikki Baird, VP of retail innovation at Aptos, wrote an op ed about how this happens. She wrote about her experience with a fashion retailer in a tourist area that complained that their conversion rate should be held to a different standard than a low-traffic suburban store. The tourist area store got more traffic but fewer sales, but the store manager argued the ‘entertainment’ traffic should count for more so her bonus would be higher. Ultimately, Baird said raw traffic numbers don’t measure levels of engagement. Measuring engagement requires different metrics such as dwell time and bounce rate. Technology like video analytics and Wi-Fi counting makes this more possible to track. With Wi-Fi counting, shoppers’ smartphones are connected to the network and send information about movement patterns. This new technology is exciting, but not all retailers use it.
There are several new ways to capture foot traffic data to inform better decisions for retailers. Tracking footfall certainly won’t hurt tenants or landlords, and looking at the macro-level numbers can indicate how well the industry is performing nationwide and in different regions. For example, traffic has been bouncing back lately at U.S. brick-and-mortar locations, increasing by 32 percent in February 2022 compared to last year, according to a recent report.
Foot traffic should not be regarded as the ultimate source of truth. For it to be useful for decision making it should be pieced together with other more meaningful metrics like conversions to give a more accurate picture. Retailers and landlords that use footfall data should also look closely at how the data is collected and its quality, which is more complicated than it sounds. Whether retailers use manual methods or new tech like sensors and cameras, collecting in-store traffic can be challenging. Many retailers must dig deeper with their traffic data analytics, depending on how much they can allocate to the efforts. Footfall data collection will continue to evolve, but let’s remember it’s one data point among many, not the holy grail.