Mobile location data has become an increasingly important tool for retailers making site selection decisions. During its time in commercial real estate, the data has evolved to encompass more detail, pull in from more sources, and increase its reliability. It’s used most commonly for descriptive analysis on a site-by-site basis.
According to Paul Sill, head of JLL's Visionary Insights Group, predictive forecasting may hold a new era of mobile data, but warns that users should approach it with some caution. "We use mobile data in predictive analytics, but it doesn't have as dramatic an upside effect, or as dramatic a benefit to that process as maybe a lot of people expect," Sill says.
The evolution of mobile data
Recommended For You
With nine in ten Americans owning a smartphone, retail site selection has drastically transformed in recent years. Early methods used basic radius rings, that assumed every customer within a certain distance was a potential customer. The introduction of mobile location data in the late 2010s was a major breakthrough, allowing retailers to observe real user behavior rather than relying on assumptions. With Americans now checking their phones 205 times a day on average, the volume of data has exploded.
"Retailers are using it to say, 'I'm going to consider this mall with this spot in it, I'd love to have a better sense of what kind of foot traffic exists,'" Sill says. This data helps retailers understand who's shopping at a location and how long they're staying, providing valuable demographic insights.
Mobile data still has limitations, Sill explains. It provides an incomplete picture, often drawing from a limited pool of apps and vendors: “It’s not a perfect tapestry of data by a mile,” Sill says, noting that revenue projection based solely on foot traffic can be difficult.
That said, today’s mobile data is far more precise than earlier iterations, which relied on cell tower triangulation. Effective forecasting now demands a deep analysis, and new mobile insights can help. Sill focuses on four key data buckets – demographics, competition, site attributes and operational factors – that can help retailers assess if current visitors match the retailer’s target customers.
Building a balance model
While there’s significant excitement surrounding artificial intelligence in retail analytics, Sill suggests the technology is still evolving. "We don’t have the data to really generate a benefit with AI-enabled algorithms," he explains. Even retailers with thousands of locations still struggle to get enough site-level data points to use AI effectively.
Instead, JLL relies on traditional statistical methods, like multivariate regression modeling, which Sill calls the gold standard. "The techniques for building good, predictable models are the same today as they were 30 years ago," he notes.
His advice? Keep it simple. Distill sophisticated models down to their core elements by identifying the handful of elements that truly matter for each retailer.
Those critical variables vary– "For some retailers, site attributes might be really important. For other retailers, they're destinations, and it's really not that important," Sill says.
Mobile location data may have changed how retailers evaluate potential sites, but its greatest value lies in combining high-quality mobile insights with proven modeling.
Visit JLL at ICSC Las Vegas, North Hall booth 3018G.
© 2025 ALM Global, LLC, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to [email protected]. For more information visit Asset & Logo Licensing.
