Since the first case of Covid-19 was identified one year ago, scientists have come a long way in understanding how the virus passes from person to person. Location data has helped researchers study stay-at-home orders, which states have implemented them best, and the places people go that present the biggest risk.
Many of these notable, splashy studies incorporate location data from the same source: a company called SafeGraph. It’s one of a growing number of companies basing their business on consumer location data. All sorts of apps and APIs collect users’ location or other personal data, and users often don’t often know where that data ends up.
One answer is in research used to understand and stave off the pandemic. And another, increasingly, is in SafeGraph’s reports to companies, which are using location data to adapt to dramatically changing consumer needs.
Here’s how the system works: SafeGraph has 6 million points of interest, which it defines as “a place you spend time or money,” across the US and Canada. Right now, that’s mostly limited to businesses. Using machine learning (and sometimes coding by hand), the company draws digital borders around retail spaces and other places of interest, and then watches how many people visit.
The raw GPS data comes from apps in which users have consented to having their location tracked; SafeGraph then “ingests” that data, anonymizes it, and aggregates it to blocks of people who move about in a particular area. The company says it gets mobility data from providers like its spin-off Veraset, which own the relationships with the apps that gather its data (Veraset doesn’t share the names of the apps with SafeGraph).
That aggregated location data is proving to be useful for businesses. “We don’t need to provide analytics dashboards or anything with our data. There are a growing number of companies using data across industries,” says Evan Barry, VP of marketing at Safegraph. It’s already taking off in marketing tech and advertising, he says, because the data can reveal whether someone who visited, say, a clothing store did so shortly after seeing a billboard advertising that store.
Retailers have also found it useful. As Covid-19 has reduced visits to brick-and-mortar shops, location data can help them set shorter store hours that result in the most business for a particular site, or whether a location is faltering because of overall reduced traffic to the neighborhood, or if a competitor’s nearby location was performing better.
During the pandemic, SafeGraph’s business has really taken off, growing about 10 times, Barry estimates. Financial services, particularly private equity, has seen the most growth, because firms are looking to location data as an indication of the performance of companies they’re considering investing in.
Those business clients have made it possible for SafeGraph to offer aggregated data for free to scientists and journalists (Quartz has used SafeGraph’s data for one story). The company created an online community made up of 5,000 people using its data to make it easier for folks to build on each other’s work. Barry estimates that 500 academic publications have cited SafeGraph’s data since April—so many that he got tired of listing them all on SafeGraph’s web site. City governments and hospitals have used the data to predict the neighborhood where the next Covid-19 spike will be; the US Centers for Disease Control and Prevention (CDC) even used it to assess the efficacy of mitigation efforts like social distancing.
Public health and national security issues are some of the most valuable applications of aggregated location data, says Kirsten Martin, a professor of technology ethics at the University of Notre Dame. “This is the good side of [data collection]…you don’t get more goldilocks than that,” she says. Aggregated, census block-level data doesn’t worry her from a privacy perspective, either, especially when it’s used by scientists, whose studies have oversight on how such data is used via an institutional review board (IRB).
It’s when we get down to the individual, identifiable level that things get a little more ethically cloudy—something that Barry says SafeGraph doesn’t do. To ensure the accuracy of a visit count, the company has a lot of checks in place, such as spot inspections to ensure a ping indicating a visit to a location wasn’t out of line with the rest of that person’s travel path. But that information won’t make it to clients. “It’s great when companies allow [their data to be used for] that kind of work, and it’s rare when they do,” Martin says.
SafeGraph hopes to lead the way for others to facilitate this kind of work. “Our hope is that, as one of these data providers, we can encourage others to be willing to open up data, and we can have the biggest impact together,” Barry says. “Data is not helpful, almost dangerous, when it’s hoarded.”