The company wants to turn great ideas into great products more quickly
What is it?
Have a guess. In a rare outbreak of jargon-free clarity from the company, Microsoft’s research division is called Microsoft Research (http://research.microsoft.com). Hallelujah!
Wait - Microsoft actually researches its products?
Now, now. Windows may not always work perfectly, but it’s not through a lack of testing. Since Bill Gates set it up in 1991 Microsoft Research has grown to employ more than 1,000 researchers and engineers around the world, in cities including Cambridge, New York, Beijing and Munich. Projects involve all aspects of computing, from artificial intelligence to speech recognition. But things are about to change.
In what way?
Microsoft wants the research to lead to more products. The company’s boss, Satya Nadella, thinks that in the past too much emphasis was placed on ‘pure’ research that wouldn’t necessarily lead to anything concrete. Engineers were guided by a ‘research for the sake of research’ ethos, not always focusing on the end result – such as products that would make the company even more money.
What will actually happen?
Researchers will work more directly with the teams that build products. Until now researchers worked in isolation, protected from the commercial pressures of having to build products that would prove popular in the ‘real’ world. That started changing in February 2014 when Nadella was shown a translation tool Microsoft Research had built. Impressed, he gave the team a three-month deadline to make it work with Skype. It’s now known as Skype Translator, and comes with the Desktop version of the program (www.skype.com/translator).
So we’ll see inventions turned into products more quickly?
That’s what Microsoft hopes. In the past too many good ideas were overlooked, giving rival companies the opportunity to launch products first. For example, back in 1998 the research team showed Bill Gates a program for making digital maps. Nothing came of it until, seven years later, Gates saw Google Maps and ordered his engineers to create their own version within 100 days.
Is Microsoft copying Google then?
Yes. Google’s recipe for successful innovation is to make sure its researchers work closely with the teams building products. Google’s management says that this makes it easier for them to spot an idea’s potential, and then suggest ways it can be turned into a product people will want to use.
What are the dangers of this new approach?
Critics argue that research has to be given the freedom to develop at its own pace, allowing for happy accidents along the way. They fear that a more disciplined approach may stifle creativity, leading to a short-term vision that will actually result in less innovation. However, Microsoft Research had earned a reputation for developing technology that was fascinating, often groundbreaking, but ultimately pointless. It looked great in a laboratory, but was never likely to appear in a person’s home. It’s widely accepted that a change was needed.
How will I benefit?
Hopefully by getting your hands on better products sooner. In fact, that’s already happening. As well as Skype Translator, better collaboration has led to products such as Cortana and the holographic headset HoloLens (www.microsoft.com/hololens). But we’ll know that a new era has dawned only when the teams work together to release a tool that helps users block Microsoft’s incessant Windows 10 upgrade messages. At least Google can’t beat Microsoft to that one.