It's Time For An Open Database Of Places
With last week's declaration by Twitter that it intends to start identifying places based on the coordinates of geo-coded Tweets, the location land rush is in full swing. A long list of companies including Twitter, Google, Foursquare, Gowalla …
Context & Ripple Effects
This call for an open database of places lands at the end of a two-year build-up: the iPhone app wave of late 2008 turned location-based social networking from novelty into a category, and by November 2009 the framing was already consolidation-versus-fragmentation — Facebook was flagged as poised to dominate geo, squeezing startups like Foursquare before they could own the layer themselves.
Twitter's declaration that it will identify places from the coordinates of geo-coded tweets is what changes the equation. A service with Twitter's firehose generating its own place records means multiple companies — Twitter, Google, Foursquare, Gowalla — are now each minting competing definitions of the same coffee shop, which is exactly the duplication the author argues an open database would eliminate.
First-order effects
- Every geo-coded tweet now doubles as a place-assertion, putting Twitter in direct competition with Foursquare and Gowalla whose check-in databases are currently their core asset.
- Google's existing place index gives it a head start the newcomers lack, forcing Foursquare and Gowalla to treat their hand-curated venue data as defensible IP rather than infrastructure to share.
Second-order effects
- With each major player maintaining its own place registry, developers building on top of these platforms face fragmentation tax — reconciling conflicting venue IDs across APIs — which creates real demand for the neutral open database this article proposes.
- Facebook's widely anticipated entry into geo, foreshadowed in the coverage since late 2009, raises the cost of sitting out a shared standard: the bigger the number of proprietary place databases, the harder any one startup's data becomes to monetize.
Third-order effects
- If the land rush resolves into walled-garden place registries, whoever ends up holding the canonical database of places becomes the chokepoint between the physical world and every location-aware app — the same platform-control dynamic playing out across the social web.
- Conversely, if an open database gains traction, place data commoditizes and competition shifts up the stack to the services layered on top of it, echoing how shared protocols have historically undercut proprietary data moats.
The trend: Location is repeating the classic infrastructure pattern where competing companies mint overlapping records of the same real-world entities until either an open standard or a single dominant owner settles the layer.