Friday, January 14, 2011

Microsoft vs. Open Kinect

This week’s dead tree Business Week has an extensive article on Microsoft’s war against the hackers trying to repurpose its pathbreaking Kinect hardware for other purposes.

It turns out Microsoft’s cool new hardware is the ideal platform for all sorts of robotics and machine vision applications, at consumer off-the-shelf prices. Primesense, original Kinect manufacturer, is planning on supporting developers with a special developer version.

It goes without saying that firms need to leverage their extremely loyal customers and that it’s risky (if not stupid) to tell them they can’t use a product the way they want.

Reporter Ashlee Vance quotes Utah’s most famous and perhaps America’s most oft-quoted) open source executive, my friend Matt:

Microsoft has succeeded despite themselves in creating something really cool," says Matt Asay, a prominent open-source blogger and executive at software startup Strobe. Yet a number of critics say the relationship between Microsoft and Kinect-loving geeks is already strained, and that Microsoft's early reactions to their playful tinkering suggest it could squander a once-in-a-generation opportunity. "Here was a chance to throw themselves deep into the bowels of the open-source hippy movement," Asay says. "They are kind of trying to do it, but they don't want to touch anyone in the mosh pit."
Adds another well-known expert (and friend):
"Companies should make it easy for people to hack," says Karim Lakhani, an assistant professor at Harvard Business School who studies open-source projects. "Why wouldn't you want people going crazy with your products?"
Business Week does a good job summarizing the situation and Matt and Karim why such hacks should be made an asset rather than a distraction or a liability.

The fundamental problem is Microsoft’s business model: they brought the hardware to the market cheap assuming they’d make all their profits off of games (which generate either direct or royalty revenue). They thought they were making a peripheral that made the Xbox 360 platform more attractive — but instead they were creating a entirely new platform.

It reinforces a point I emphasize with my technology strategy students: cross-subsidies are always dangerous, because someone will find a way to arbitrage the subsidy. If you give away razors and make it up by gouging people for razor blades, then someone will sell cheap blades that don’t need a subsidy.

One might argue that Microsoft couldn’t such hacking coming, but that wouldn’t compute. Almost nine years ago, hackers figured out how to take (the heavily subsidized) original Xbox and turn it into a cheap Linux machine.

Microsoft is missing a significant market opportunity by not being open to third-party enhancement of the Kinect hardware. Although the volume will not be as big as for a hit game — as predicted by Karim and his research — third-parties will identify markets and solutions that Microsoft never anticipated.

Over the past few years, Nokia faced a similar conundrum as university scientists hacked N-series phones (such as the N95) to use as a mobile sensing platform. At least a few people in Nokia Research (back when there was a Nokia Research) understood this and were working on even better unlocked hardware for this purpose.

It’s not too late for Microsoft to turn things around. The Kinect COGS will continue to fall. Instead of more aggressively pricing the hardware, it can use the savings to increase the gross margin, reduce the subsidy and make the Kinect a self-supporting stand-alone device.

In the short run, Microsoft can also (quite inexpensively) support a university or hacker conference on other applications of Kinect. MIT, Stanford, Carnegie Mellon or even its hometown UW would be glad to take the lead.

No comments: