Google has abandoned  its U.S. map-data provider, Tele-Atlas, in order to begin making its own maps. Earlier this month, Apple traveled a similar avenue, buying  a small maps competitor called Placebase. Is all this movement towards in-house mapping because of bad blood, or is it business savvy?
The Google-Apple relationship has been framed as melodrama; some outlets have suggested  that growing competition between the companies' smartphones and browsers is leading each to make preparations in case of a corporate breakup . Those rumors were sparked by Steve Jobs' explanation of the departure  of Google CEO Eric Schmidt from the Apple board. "Unfortunately, as Google enters more of Apple's core businesses...Eric's effectiveness as an Apple Board member will be significantly diminished," Jobs said . And now Apple director Arthur Levinson just announced  he's leaving Google's board.
The break-up theory is also bolstered by the recent tension  over Google Voice, a telephony iPhone app that Apple mysteriously rejected this summer. As for Google and Tele-Atlas, the situation is less clear. This is Google's second map breakup; it also ditched Navteq last fall after it was acquired by Nokia. Sounds like more melodrama, right?
Wrong. As fun as it is to think of CEOs bickering in the boardroom, it's more likely that both these schisms are about two other factors: hyper-local data, and how to scale it.
When a company like Google licenses map data from Tele Atlas, Navteq, or other primary data collectors, they're not paying flat fees for the privilege. According to one navigation-industry source I spoke to, each level of map zoom and each layer of data (street names, freeway exits, topography) has a different copyright, and each copyright requires a license. That means more detailed maps come with more fees.
Planning for a world of hyper-local data usage--and increasingly popular augmented reality--means lots of very detailed mapping. And every time Apple and Google think of a new way to let developers use their mapping frameworks and APIs, they open themselves up to new kinds of licensing fees and restrictions. To really be flexible with hyper-local and do it cheap, the only option is to invest in-house.
Not to mention that providers like Tele-Atlas don't have much hyper-local data to sell, and what data they do have is poor. No wonder Apple wants to bail from Google--the search giant feeds Tele Atlas data, with its dearth of street-level layers, to iPhone maps. Cupertino wants to make sure its app developers and iPhone customers have quality granular map data. And no wonder Google wants out from Tele-Atlas; with Street View, Google has more of that neighborhood-level moxie than any mapping company around. What they lack--big-picture, low-detail maps--can be crowdsourced or borrowed freely from the U.S. Census Bureau's data contained in the OpenStreetMap  project.
More proof? Google is still using Tele-Atlas data for non-U.S. locations, where hyper-local isn't yet a priority.
If all this hustle over maps is any indication, we, the end users, can look forward to smartphone geo-services we never imagined.