The real breakthroughs in technology don’t come from asking customers what they want.
“You go and you find out their problems,” says Steven Drucker, a researcher in Microsoft Research, the software giant’s 15-year-old research organization. “But you typically cannot rely on them for a solution.”
Drucker, who met with a small group of Canadian technology journalists at Microsoft’s corporate campus in Redmond, Wash., this week, doesn’t mean customers’ needs and opinions don’t matter. Once Microsoft Research comes up with a promising technology idea and one of the company’s product groups decides to run with it, he says, there will be lots of consultation with customers before the idea turns into a commercial product.
It’s just that when it comes to truly innovative ideas, most people don’t really know what they’ll find useful until they see it. “I don’t think customers are very good at coming up with any leaps,” Drucker says.
The more than 700 researchers who work on upwards of 50 research areas at Microsoft Research are looking for ideas that go beyond refining current products. Not all their ideas ever find their way into commercial products, and that’s fine with Drucker, who suggests that if more than half a research organization’s work is being commercialized, it’s probably not doing the right kinds of things.
He estimates that between 10 and 30 per cent of what Microsoft Research comes up with ends up in the company’s products. This “product transfer” is one of the factors on which Microsoft evaluates research staff, but it’s not the only one. Patents are also a factor – if Microsoft never makes anything based on a patent, it may collect licence fees or trade the technology to another company – and just as in academia, published research and presentations at conferences are also important.
“The policy of research is to try and hire really bright people and to try to not manage them so much as to say, ‘Here’s what your success factors are,’” Drucker explains.
Drucker – who spent time at the Massachusetts Institute of Technology’s famous Media Lab before coming to Microsoft – works in a small research group that is interested in solving problems related to online media. As an example, he cites the growing volume of digital photographs many consumers have on their computers. Drucker’s group has been experimenting with ways of making those photos easier to organize.
Some of the group’s ideas have already found their way into Microsoft products. Some are still in the idea stage – such as a prototype designed to make it easier to find photos without requiring the user to enter any information about them manually.
What the trial version does is to divide photos into groups based on the date and time information that digital cameras store automatically, then represent each group with one picture it judges to be representative – because it resembles several others in the group, for instance. In a broad view, each group is represented by one picture, the size of which is proportional to the number of pictures in the group. Zoom in, and the individual pictures in the group appear.
To show visitors technology ideas its researchers believe are close to becoming reality, Microsoft has set up the Microsoft Home at its Redmond campus. It’s a replica of a fairly upscale home, but with technology going a bit beyond what we have today.
Jonathan Cluts, Microsoft’s director of consumer prototyping and strategy, says Microsoft revamps the home regularly, and never includes anything its researchers don’t believe will reach a commercial price point – defined as no more than 20 per cent above the cost of the most comparable commercial technology today – within six years.
Today’s Microsoft Home features several large video screens that can display television, video on demand, home movies, games, and menus that provide access to other services. There are also other displays, such as a monochrome screen set into a wall and painted over so that it disappears into the wall when not in use. It can show weather reports and menus to control home appliances. Similarly, a kitchen bulletin board incorporates a low-cost, monochrome display to show notes and other information.
Radio Frequency Identification (RFID) tags figure heavily in the Microsoft Home this year. Place a bag of flour and a food processor on the sensor-equipped kitchen counter, ask for recipes, and a list of recipes involving flour and a food processor is projected onto the counter. Place an article of clothing in the washing machine, and the machine reads its RFID tag and provides washing instructions. A child’s tablet PC lists toys that are not in their proper places and lets the child earn “fun credits” by putting them away.
Though the idea is that these technologies will be accessible to average consumers in half a dozen years, Cluts says people will buy them selectively, often as older devices wear out. He is under no illusion, he says, that in a few years most people’s houses will look like Microsoft’s present showcase.