What’s in a Swarm? Making Sense of Foursquare’s Split
As has been broadly covered on this site and elsewhere, Foursquare announced last week that it would split its popular smartphone app into two services. First, in a month or so, the company will launch Swarm, an app dedicated to check-ins and showing users which friends are in their vicinity. Next, sometime this summer, the main Foursquare app will lose its check-in button and focus exclusively on local search, discovery, and recommendations.
Somewhat oddly, the company is responding to internal research that suggests the overlap is low between users who like to check in and users who want to search for local places. Concluding that the local searchers are put off by the conspicuousness of the check-in button, Foursquare thinks it can lure check-in lovers to Swarm, where it will provide a more immersive social experience that involves passively projecting your location to a select group of friends. Purportedly, the more obvious strategy — creating stronger and more distinct user paths for both use cases within the same app — was tried and abandoned in favor of separating these user groups entirely.
The gains in engagement Foursquare has in mind must be truly dramatic to be worth the risk of splitting its user base. For the most part, spinoff products like Facebook’s Poke and Pages apps have been unsuccessful at gaining traction.
But the move is consistent with Foursquare’s announced intention to become the location layer of the internet. The company has already quietly built an impressive list of relationships with the likes of Instagram (from whom it may now be splitting), Pinterest, Apple (maybe), and most recently, Bing, which along with a hefty investment from Microsoft has announced plans to bundle Foursquare data into Windows 8 and the Windows phone. Foursquare reportedly owns location data on 60 million venues across the globe, mostly created by users who have checked in 6 billion times and helped the company amass one of the richest repositories in existence of valuable content like location-linked photos.
Now, CEO Dennis Crowley’s gamble is to cut the location database free from the check in and let it begin to perpetuate itself. What may emerge is an entirely new concept of a local database, one that is essentially crowdsourced in nature and built according to user interest.
Any company saying they want to build the location layer of the internet is going to inevitably be compared against the current gold standard, Google Maps. By this measure, Foursquare data is notably insufficient, full of duplication and inconsistencies. Such shortcomings are, one might argue, hallmarks of any service that permits users to create and modify location data freely. Yet the Wikipedia model suggests otherwise. The online encyclopedia has shown a consistent ability to crowdsource accurate information at scale, but Foursquare users seem disincented to correct each other’s errors at a sufficient rate to prevent such glaring and easy to find problems as that of the United Methodist Church in San Luis Obispo, Calif., whose city is listed as Bakersfield (137 miles away).
Even more broadly than issues of duplication and inaccuracy, Foursquare’s orientation toward eating, drinking, and entertainment venues has been so endemic that full coverage of all types of business seemed unlikely even as a long-term possibility. The announcement of Swarm and the reconfiguration of Foursquare proper suggest that the company may be restructuring itself precisely in order to move beyond this limitation.
Furthermore, weighed against these shortcomings are the many benefits of a socially driven local dataset. Like Yelp with reviews, Foursquare has built a database that captures the patterns of interest of actual consumers. Though I argued against this in the case of Apple, the integration of Foursquare data makes a lot more sense for any service (like Pinterest) that wants its location data to be social in nature or that (like Bing) already does a pretty good job of providing canonical data but needs to bring a higher level of social engagement into the mix.
Who knows? It’s hard to envision now, but it could be that a decoupled discovery and recommendation service will be just what Foursquare needs to scale its dataset beyond entertainment and to encourage users to improve the quality and accuracy of venue information. These developments would turn Foursquare into a viable competitor to data aggregators like Infogroup and possibly to Google Maps itself. Actually becoming the Wikipedia of local data would be a great achievement for the company.