Bend it like Beckham: Using Geolocation Data to Measure Economic Activity

Stephen Buehler
Stephen Buehler

Creator of Astra Insights & Founder of Astra Ventures

Over the past few months, the phrase “flattening the curve” has become mainstream vernacular. We have also been flooded with references to other types of curves…namely the potential shape of economic recovery. Will the recovery be a “V” shape, “U” shape, “Nike Swoosh” or something else? We all have opinions. But no one knows for sure. I thought it would be interesting to look at some other curves to provide some potential clues. In doing so, there are two important concepts to point out.

First, when we refer to a macroeconomic recovery curve, it is important to remember that macro curves are the summation of many, many micro curves from all corners of our economy. For example, the medical sector may recover at a different pace than the entertainment (theaters, concerts, etc.). So what this means is details matter. The curves I’ll refer to show that level of granularity (compliments of great data from Foursquare).

Second, the macro curves are important on just that, a macro level. However, to drive change (and to bend curves to reduce or increase the slope), it is more effective to do so at the micro level. Because broad, blunt stimulus is a least common denominator approach which, by definition, is not optimal for anyone or any industry.

The following charts are the work of Foursquare. Simply stated, Foursquare is a social networking company that focuses on physical location data obtained from your mobile phone (e.g. when you click yes to “will you allow XXX to track your location”). This geolocation data provides some fascinating insights around consumer behavior trends. I provide links to their work below so you can read more on their methodology. For brevity, I’m providing a handful of the many interesting charts (FYI, data is through mid-April). These are the curves we want to bend upward.

As a guide for reading this post, I have updated (and will continue to do so) Foursquare’s data graphics as time passes. I first posted back in April. I later updated the post with Foursquare’s data through May 1. It is interesting to review both sections below to see how activity has evolved. In fact, they just published a fantastic tool to show week over week changes by state. You can check that out here.

Data as of April 17

The first two charts are indicators of business and leisure travel. We can see that activity levels dropped off dramatically for both airports and hotels. I see these as two of the most critical micro curves to monitor to get an indication of trends at the macro level. This is because business and leisure travel have so many downstream impacts economically speaking (rougly 10% of GDP is attributable to tourism – which doesn’t even capture the business impacts).

The next few charts are good barometers of consumer behavior trends. What will be interesting to see is which of these recover more than others. I suspect that curves related to clothing stores and movie theaters will be victim to secular behavior shifts to online and streaming.

Next, I’m including a graph that may be interesting to watch as a gauge for long term impacts on commercial real estate valuations. The forced hand to working from home has shown that productivity can be maintained without incurring expensive real estate leases. I would not be surprised to see many companies move to a hybrid model where they take ~30-40% less office space and then have workers rotate days where they work in and out of the office.

Next are a few counter-trend graphs of activities that have either not been impacted or that where activity has increased. You can see from the below charts, people need to eat, drink booze and do yardwork.

Data from May 1

Next, I’m including a number of charts from Foursquare’s data from Georgia as the state was one of the early movers to slowly reopen its economy. While Georgia is only one data point, human behaviors in the US will likely be similar across similar areas (urban, rural, etc.).

Foursquare created a new tool that allows you to toggle by state and industry. That tool and some overlaid analysis is here.

Admittedly, I’ve been of the mind that the recovery would be very slow and could take quarters if not years. But these charts are somewhat encouraging. However, we are not yet out of the woods. As the economic contraction continues on, layoffs that began as temporary risk becoming permanent. And this vicious cycle begins. This is why I’ve always found the following saying insightful (unsure of attribution).

Unemployment takes the elevator down but the stairs back up.


Leave a Reply