'It Can’t Really Predict Things': Post-Pandemic, Retailers Question Utility Of Data For Outpost Selection
After two years of operating during a pandemic, chain retail stores and eateries are finding that data collection has become a less reliable predictor of customer behavior — forcing retailers to readjust their balance between science and art as they determine what spaces to open and where.
Retail chains are used to collecting data such as foot traffic and sales volume around spaces they lease, helping them to build an understanding of how customers behave in their stores or restaurants. But the pandemic has changed the advantage that data previously offered retail tenants, experts say. While they can still understand which locations are doing well and for what kind of customer, they are less able to predict how customers may behave in the future.
Meridian Capital Director Vince Sweeney, a New York-based tenant representative, said Tuesday at Bisnow's National Retail event that data can help retailers learn how their customers behave by offering information on how and where they live, work and spend their leisure time. But when that data collection becomes oversimplified, he said, retailers lose out.
“I've seen some of these self-tracking data programs where it's like, ‘I see a lot of activity over the bridges.' And it's because there's traffic, so everybody’s stopped there. And that data, that’s not really useful,” Sweeney said. “All this data — which is amazing to have — it's just information at this point. It can't really predict things.”
Starbucks, with 175 stores in Manhattan alone, is able to see easily which locations are getting business in the expected volumes and which areas are lagging, Starbucks Director of Store Development Dan Shallit said.
“I definitely have a little bit of a cheat,” Shallit said at the event.
“We're seeing higher volumes in the Upper East and West Side, where people live and are working from home, and we're still getting crushed in the Financial District,” Shallit added. “There are still Starbucks stores that have not opened since the pandemic closed the store.”
Starbucks’ data can tell it not just which neighborhoods are seeing foot traffic, but how they’re behaving. The UES and UWS have historically had an active lunch market — something that’s translating into foot traffic at Starbucks stores in those neighborhoods — whereas FiDi relies on workers being in offices to get lunchtime coffee runs back to their pre-pandemic pace.
Successful data collection has helped companies adapt to what markets demand, Jokr Senior Real Estate Manager for North America Alex Solomon said. But while the rapid delivery app company has a large amount of data on its customers’ behavior, that information doesn’t tell it everything it needs to know for its business strategy.
“You understand where we need to be in certain markets. Then we need to be able to reach those people,” Solomon told Bisnow’s audience. "So it's definitely scientific and how we get to where we need to be in a market, and then that's more of an art form."
Shallit said Starbucks has long relied on a blend of art and science to figure out what products its customers want, where, and what store environment they’re seeking.
“When we first started, it was 75% art and 25% science. We got a demographic report, but it was more about what you felt, how you saw market fits,” Shallit said. Then came the age of data, where traffic patterns and sales volumes largely dictated Starbucks’ retail strategy.
Now, Starbucks is back to balancing that data-driven approach with feeling to determine what type of store works best for each location.
“We may want to build a big store with a lot of seats, but in other markets we may just want to build a grab-and-go store. That's where the science doesn't help us,” Shallit said. “Today it’s back to more art.”