"Google Phone," an HTC-built device called the Nexus One handed out to Google employees last week in what Google describes as a "mobile lab." Confirmed to be running Android 2.1, the Nexus One has once again raised the idea of Google selling unlocked devices directly to consumers. (Google has been selling unlocked HTC Android phones for some time, but only to developers.)
It would be a strange turnabout if Mountain View made this move, directly going in the face of previous assurances that Google had no plans to compete directly with Android hardware manufacturers. What's more, there are a lot of unanswered questions here.
The first question: How would Google bring an unlocked phone to market? There are really only three ways to sell phones. The first is to license spectrum from a carrier and become a mobile virtual network operator or MVNO -- a business model that time has proven to be a failure. The second, of course, is to partner with carriers and offer phones at discounted prices through carrier subsidies, which is more or less the case with every successful device on the US market today. The final model is to sell unlocked devices at full retail price that can be used by consumers on the network of the their choice. This is allegedly the model Google will be using to sell the Nexus One.
Until someone can give me a ten-word answer to how Mountain View can manage to build an ecosystem while trying to compete with it, I will remain skeptical that the Google Phone ever comes to market.
Selling unlocked devices sans carrier is a lousy business model in the States, however. There's no mass market for unlocked phones in the US -- just ask Nokia how hard it is to sell a high end phone with no carrier subsidy or support. Either Google would need to take a huge loss on every device to achieve a consumer-friendly price point, or hope to convince consumers to pay full price for an unsubsidized device -- even though Eric Schmidt in the past has argued phone prices need to trend to zero through full subsidies. What's more, an unlocked device will at least get you onto T-Mobile and AT&T's EDGE networks, but Verizon and Sprint both require phones that are approved for network use and can easily be locked out.
There's something even more fundamental that struck me as I listened to the "Google Phone" chatter, and that's the basic challenge of licensing to competitors. One reason Microsoft is successful in the PC industry is that it's never built PCs. Licensing to folks you compete with doesn't work: either your licensees do better than you do, in which case why bother, or you do better than your licensees, in which case your licensees wise up and go elsewhere. Apple's tried this twice: first with the Newton, where Apple did better than the licensees, and second with Classic Mac OS, where licensees like Power Computing did better than Apple -- eventually driving Cupertino to give up on the licensing idea entirely. Palm tried it, and it eventually had to split up into Palm and Palm Source. Nokia tried it with S60. The whole point of the Open Handset Alliance is to create a partner ecosystem of handsets and other devices, and a "Google Phone" that undercuts both carriers and licenses might well be the death of Android in the marketplace.
There might be a strategy here that allows for this to happen -- I can even think of one or two -- but until someone can give me a ten-word answer to how Mountain View can manage to build an ecosystem while trying to compete with it, I will remain skeptical that the Google Phone ever comes to market.
Michael Gartenberg is vice president of strategy and analysis at Interpret, LLC. His weblog can be found at gartenblog.net. Contact him at gartenberg AT gmail DOT com. Views expressed here are his own.