Over the past two months, we’ve examined Foursquare’s ongoing data on local-commerce foot traffic patterns — specifically in light of shifting behavior due to quarantining and other lockdown measures. We also promised we’d stay on top of that moving target as new data become available.
True to that promise, we’re back with the fourth batch of data covering the week ending May 1. The background is that Foursquare’s core technology tracks large-scale opted-in mobile devices for location-marketing insights. In this context, the data track “new normal” foot-traffic patterns versus baseline activity.
The first two batches covered the beginnings of regional shelter-in-place orders, while the third batch took us into the thick of the lockdown period. The theme with this fourth and latest data release is what we can expect as many states reopen while relaxing stay at home orders and business closures.
For example, a few highlights:
— Visits to health and beauty services (hair salons, barbershops, nail salons, tattoo parlors, and spas) quickly jumped up to 80% of pre-pandemic levels.
— Restaurants returned to more than 75% of “normal levels” once dine-in service was allowed.
— Fast food traffic in Georgia has returned to pre-pandemic levels
— Casual dining restaurants that typically rely more on dine-in traffic are down about 50% from normal levels.
— Though still below pre-pandemic levels, gym visits were down less than 60% in the week ending May 1, versus 65-75% declines in the weeks prior.
— Notably, foot traffic to all places in Atlanta remained lower than GA as a whole, though not by much.
These are mostly expected outcomes in terms of re-assimilation and consumer confidence levels. But the degree of activity is notable. The underlying point is to view this data as an indicator of what we can expect as other states reopen, and therefore hopefully a valuable tool to inform public health decisions.
For example, like the leading indicators above, we can expect wellness and self-care (hair salons and gyms) to spike first. Trailing those could be entertainment and restaurants, though all of the above will be tempered by varying levels of consumer confidence, despite government-approved reopenings.
With that backdrop, here’s the deep dive:
Taking A Closer Look At Dining
While many restaurants have remained open for pickup, drive through and delivery since the pandemic began, dine-in foot traffic is an excellent indicator of consumers’ comfort level in returning to their daily activities. Dine-in service was permitted in Georgia as of Monday, April 27, subject to specific social distancing and sanitation mandates.
Restaurant visitation in Georgia did not immediately spike once restrictions were lifted, but our data showed significant upicks in traffic on April 30 and May 1, with visits reaching roughly pre-pandemic levels on Friday May 1. This may indicate that restaurants in Georgia did not choose to open immediately once allowed to do so, and instead may have taken a few days to prepare themselves for new protocols and procedures.
Analyzing specific types of restaurants, we see that fast food traffic in Georgia has returned to pre-pandemic levels, whereas casual dining restaurants that typically rely more on dine-in traffic are significantly lower.
Upticks in Georgia restaurants’ traffic have been largely driven by rural and suburban areas, with much smaller changes observed in densely populated zip codes. Restaurant visits in rural areas reached 88% of pre-pandemic levels, whereas restaurant visitation in urban areas of Georgia was still at 50% as of May 1.
While restaurants in Atlanta saw upticks in foot traffic once dine-in service was permitted starting April 24, traffic rose slightly less in the city versus the statewide trend.
Older audiences over 65 years old have been much slower to start visiting restaurants in Georgia again compared to other age groups.
Men have been quicker to start visiting restaurants in Georgia again compared to women.
More to come as this develops, and see Foursquare’s methodology and additional context here.