Building the Location Layer: A Conversation with Foursquare
Location technology company Foursquare recently announced its new Pinpoint audience segments product. Building from its large corpus of data on places, spatial movements and behavioral patterns, Pinpoint represents the latest in Foursquare’s evolution as the “location layer,” for the internet. We got the chance to sit down with Foursquare CEO Jeff Glueck in San Francisco to find out more. Here is the full interview.
Street Fight: There are lots of players in location data and a fair amount of confusion about differentiation. Foursquare has always been advantaged by independence for data integrity and collection methods for (opted-in) background app data on things like spatial movement and dwell times. You also recently announced a 300% boost in accuracy. How did you do that, and what other technical differentiators does Foursquare have?
Jeff Glueck: There are a few differentiators for our location technology. It was built from the ground up, so we really see it as a phone’s-eye view of the world, and it’s understanding when a phone really stops at a place—a “true stop” as we call it. That’s quite distinct from most of the industry. Secondly, GPS doesn’t work well in dense urban areas and in buildings; it doesn’t work vertically. So we really spent years and billions of human-powered confirmations to build this digital fingerprint that is essentially a map of businesses.
It’s validated by things like triangulating the billions of permutations of 100 million wi-fi access points. We started with 13 billion human-powered confirmations, which started as check-ins in on our consumer-facing apps, and we still do more than a billion check-ins per year. But now when people are tagging a tweet on Twitter or redeeming a coupon on SnipSnap or other apps, we are learning that, ‘yes, this is the Starbucks and not the Juice Generation next door.’ The system gets better over time through machine learning and a real-time feedback system that is different than other ad tech companies.
SF: You’re announcing Pinpoint, a new audience segments product built from Foursquare’s behavioral and spatial data. Tell us more about the new product, its inception, and rationale.
JG: The places you go are the best indicator of who you are. So we are launching a segments product, so you’ll be able to get to hundreds of generalized, privacy-friendly location behavior profiles, kind of like a traditional Axiom profile, across about 150 million U.S. devices. We’re excited about this to help brands identify fitness junkies or fast food frequenters or car shoppers and things like that.
You’ll be able to access it through trade desks and other distribution because a lot of agencies and brands are moving towards a more self-serve data-centric approach. But they want partners to help them with optimization and planning, and we’ll offer them scalable audiences. It’s all privacy-friendly because we abstract to very general characteristics.
SF: Speaking of privacy, data collection is under lots of scrutiny these days, considering GDPR and Facebook’s missteps. What’s Foursquare’s stance on regulating data collection, particularly location data given the sensitivities attached to anything involving location?
JG: Not all location tech is created equal. Ours has always been opt-in, permission-based and consumer-control based. We support GDPR features like Right to Be Forgotten and portability, globally. We’re in favor of smart regulation and think it will lift the industry up to the high bar that we advocate. But it’s important that the industry get out there and communicate because there could be bad regulation. The analogy that I always use internally is self-driving cars. They’re going to save millions of lives. Humans are very bad at driving. But it’s an evolving set of norms, and society needs to go slow and figure it out, but you wouldn’t want to shut down the opportunity to save millions of lives.
It’s the same thing with location: if applied badly, location tech would be a problem. If we apply it well with the right norms, it’s going to make people’s lives better, solve congestion in cities, eliminate spam and help guide you to amazing experiences. So let’s not throw away the baby with the bathwater, and that’s the key for us to communicate the benefits of the technology and get smart regulation.
SF: What are the next areas of growth and applicability for your location data? Any use cases or even verticals that are underserved today? Any evolution beyond using location data for ads and attribution?
JG: I think over the next five years, an enormous amount of opportunity will open up in industries like health and real estate and financial decision-making, logistics, and urban planning. We’re just starting to develop partners with things like augmented reality and IoT and AI contextualization. So we think all these parallel technologies like AR and AI are going to need a location intelligence component—even just to trigger or alert you when you’re in a place where an AR experience is available or relevant because you’re not constantly going to be in AR mode. So we are interested in all of these growth areas.
Near term, we’re thinking of things like the advertising industry and retail. The way big companies are organized today, they can’t take advantage of this technology because of legacy organizational structures. The digital teams that are doing digital marketing have this legacy set of metrics for driving web traffic or e-commerce, even though for most of these companies, 90-plus percent of their customers are in store. And that leads to an inability to take advantage of a much more omnichannel view of the consumer, and that’s got to change. We think more and more brands are going to be fusing these capabilities.