“When I was a boy and I would see scary things in the news, my mother would say to me, “Look for the helpers. You will always find people who are helping.” – Mr. Rogers
Like all of you, we at Foursquare are directly impacted by COVID-19. After publishing our COVID Insights Report, which detailed foot traffic trends across large cities, we received dozens of requests from crisis response teams around the country asking how our data might help them better understand how people are behaving, know what’s happening, and help these teams make smarter decisions for their communities.
Foursquare can be one of the helpers. We are committed to sharing data that can help to save lives and doing so in ways that do not violate individual privacy. We have shared our aggregated and anonymized foot traffic data with a distributed team of volunteer technologists. Together we’ve worked to make that data available to both local government response teams and local public health officials using simple filters and data visualizations.
You can find the tools powered by our data at http://www.visitdata.org.
We encourage you to explore the data so you can understand the trends in public behavior as well as confirm that no personally sensitive data is being exposed. The tools allow aggregate Foursquare data to be filtered county-by-county, state-by-state, day-by-day, and category-by-category without exposing personal information. It’s important to note that these tools CAN NOT be used as “location trails” to identify the places where people infected with COVID-19 have visited (as this would go against Foursquare’s privacy principles). This data doesn’t reveal the identity of anyone, nor the trails of any individual. It is completely anonymous, and aggregated. These tools are designed to provide aggregate data on where large crowds are congregating on a consistent basis, and how those trends may be changing over time.
These tools are currently available for anyone to access for free and include data from all 50 states, updated daily and dating back to March 1, 2020. If you are a crisis response team who needs help working with this data, please contact firstname.lastname@example.org. If you are a volunteer programmer, data analyst, or crisis team liaison who wants to help, please contact email@example.com.
I wear a number of different hats at Foursquare these days, and I’m writing this post wearing my Executive Chairman hat. It is personally important to me that Foursquare is willing and able to put the technology we’ve built over the last 10+ years to use in ways that reach far beyond advertising and retail analytics, and it is my job for the foreseeable future to make sure that Foursquare is here to help however we can. We believe every company can do its part, and we are working to do ours.
Notes On Our Methodology
The visitdata.org website is powered by Foursquare’s US foot traffic data, derived from our panel of 13 million always-on opted-in users. The data is represented as an index. Foot traffic for a given county on a given day is indexed to average foot traffic for that county in the month of February 2020. Increases or decreases relative to that average are corroborated with a measure of statistical significance. In this case, we are simply testing for a difference in means, given “n” size, effect size, and variance. Please note that we are constantly iterating on this model, so don’t be alarmed if you see slight changes in numbers day to day.
To learn more about Pilgrim – Foursquare’s “snap-to-place” technology which powers our visit data – click here.