At this morning’s SXSW session, FourSquare CEO Dennis Crowley discussed growing the company on data, maps and recommendations, and the future of location data as a utility…
With DodgeBall (a location-based social network acquired by Google in 2005) Crowley came to understand the value of geo-located check-ins, and conceived FourSquare using game mechanics (badges, leader boards etc) to motivate users to check-in. In four years, FourSquare has attracted over 30m users, growing at 1.5m users per month, with more than 1m merchants on the platform, and 60% of users now non-US. The ‘place database’ has over 50m entries. Interestingly, giant tech companies – from Twitter’s Vine, to Path, to Facebook’s Instagram – see FourSquare as “neutral space” and deliver location services through FourSquare’s API. And this is becoming a closed loop, with user tagging of videos or photos on these platforms driving FourSquare’s location intelligence. Despite all this, FourSquare is just 160 people, shipping new or updated software products every few weeks.
Crowley commented that there is a lack of insight into the massive opportunity and future impact of FourSquare transitioning from “cute” points, badges and gamification to mining the 3 billion check-in data points created by their user base. The “small human actions” that fuelled Google’s growth was clicking of links, and Facebook was ‘likes’: FourSquare can now mine the aggregate value of their user’s small human actions – which now amount to billions of data points – to draw conclusions about community, location, consumerism and sophisticated recommendations. “People are out there crawling the real world the way Google spiders crawl the web”.
Crowley commented that the opportunity lies in “understanding the context of use”: phones passively track where the user is and then “perking up” when a user deviates from their usual routine or place to actively make suggestions and drive behaviours (FourSquare Radar). For merchants, FourSquare has enough data to begin joining-the-dots around community and user profiles to more actively grow local businesses. Crowley hinted that in the future, users may no longer have access to all deals, with the apps rewarding the best users (by frequency of interaction and possibly some socio-economuc measure of value as a customer) with varying value and quantity of offers.
Asked about his mistakes, Crowley said one error was not putting search functionality front-and-centre sooner, as it drives people to think about the service as default local search. Talking about the future, Crowley echoed other keynote speakers from SXSW: don’t lose focus, just fix one thing well (“location is the thing that we do”), don’t get distracted by trying to do everything (e.g. the short-run “follow mode” that FourSquare tried, mimicking Twitter). In terms of future functionality, these will include: context data (queues, waiting times etc), including using data visualisation to “map the shape of places” based upon activity (e.g. roads); social-graphs and interest-graphs will have joint impact to construct dynamic recommendations; as above, increasing emphasis on passive geo-location (FourSquare Radar to “own the local discovery of places”) to drive active recommendations; and while they don’t have ambitions to be a transactional provider, Crowley believes FourSquare has a huge opportunity to make “point-of-sale transactions smarter”. Crowley is ultimately massively passionate about making it easier for people to consume data on what’s going on around them. “Local is going to be huge, maps need to be reinvented, FourSquare gets to invent the future of this stuff…”.