"I think I have the best job in the company," says Stevie Bathiche, Distinguished Scientist of Windows Hardware at Microsoft, where he’s worked since 1999. Bathiche is something of a mad scientist in residence, a guy whose job it is to tinker for a living—and to hire and manage a team of similarly curious tinkerers.
That team, about 20 people strong, is called the Applied Sciences Group, and is made up of a diverse collection of scientists and engineers. The Applied Sciences Group’s fingerprints are everywhere, in algorithms for Kinect, in Microsoft’s multi-touch mouse, in some of the underlying technology of the Surface tablet and its accessories.
Fast Company caught up with Bathiche to extract lessons from his experience running an experimental wing of a large corporation.
Thinking Long, and Short
The Applied Sciences Group occupies a funny niche between research and product development; they’re encouraged to think big and not be slavish to the marketplace, but at the same time, there’s a need to make concrete contributions to the business. Bathiche walks this tightrope by having his team pursue a portfolio of projects, some with long time horizons, others with short ones.
"Some projects are lower-hanging fruit, and might have impact in the next six months to a year," he explains. "Other projects are more mid-term, two to three years, and others are longer term, three to five years. Those are all important for us to have." If the team only pursued short-term projects, nothing would distinguish it from Microsoft’s product groups. And yet if it exclusively looked at the long term, "we wouldn’t be all that relevant," he says.
There’s an art to striking that balance between blue-sky dreaming and on-the-ground realism. "It’s a fine balance," Bathiche says. "You have to tweak things in order to stay relevant but also surprise the business. We do our jobs when we give the business something they want but didn’t know they wanted. It’s that surprise that we’re here for."
Drill Dry Holes
Entrepreneurs routinely say that their failure contained the seeds of their success; the lessons learned in a doomed first venture inoculated them against making the same mistakes in the future.
Bathiche also is an evangelist for failure, but in a different sense: Sometimes you simply need to get lost in the jungle in order to stumble on Shangri-La.
Bathiche’s team members routinely come up to him with a twinkle in their eye, saying they’d like to pursue a certain project. And often, Bathiche thinks the project is misguided, doomed to failure. He lets them do it anyway.
"I will always allow ideas to happen that I think are stupid," he says. "I know that the work you’re on today will lead you somewhere else, where there will be a really good idea. Several times in the past I’ve started a project that people thought was really dumb. But I’d take it down somewhere, and it leads me to something else." He was sure the gold was over here, upstream, when in fact it turned out to be over there, round the bend. At moments like that, he recalls thinking, "Oh my God, that’s why we did this project."
It’s crucial to hire people you trust immensely, if you’re going to let them be wrong. Bathiche hires carefully, and his team members often have a proven track record of innovative thinking. "There’s a little bit of intuition as well" that goes into the decision to give someone a long leash, he adds.
"You have to give them enough rope to ask those questions, and to be wrong. They have to feel like they can be wrong," Bathiche says. "If you don’t feel safe to be wrong, you will never surprise anyone. You have to be wrong in order to hit the big ideas—if you’re always right, you’re probably not stretching yourself enough."
Timing Really is Everything
The Applied Sciences Group's members often generates plastic bins filled with files and prototypes of old projects that went nowhere. But these bins are hardly graves. They’re seeds waiting to sprout.
A prime example is a pressure-sensitive keyboard Bathiche designed back in 2002. "It was a cool idea," he recalls. There was a notion that gamers might like it—you might press a key down harder, and it would cause your character to run faster, perhaps. Or maybe you could use the tech for dynamic font input; hit a key lightly and a small letter might appear, then press it hard and the letter might register in larger type. "But the application didn’t really pop," Bathiche recalls. It was cool, sure—but hardly necessary.
It went away in a plastic bin.
Fast-forward about a decade, and Microsoft finds itself wanting to design a marquee tablet to rival the iPad. One of Microsoft’s chief concerns? Designing an ultrathin keyboard that can be embedded in a cover accessory. "That’s when the killer application for pressure popped. It was like, ‘My God, this is why we have pressure.’" The pressure-sensitive keyboard has since found its way into the TouchCover for Microsoft’s Surface tablet, which enables users to have an ultrathin physical keyboard.
"The timing of an idea is just as important" as the idea itself, says Bathiche. "You can have the right idea, but at the wrong time. It has to be the right idea at the right time. For us, the key is learning the intuition of when the right time is."
You have to feel out when to reopen the plastic bin.
Let ’em Laugh. For Now...
Ultimately, all these lessons boil down to a more essential one: Be willing to be laughed at. Earlier in his tenure at the company, Bathiche might've been taken for something like the Redmond Court Jester. "So many times I’d have an idea, and I would get laughed at," he recalls.
Once, he pitched the idea of a teddy bear robot. The room erupted in snickers.
He built the teddy bear anyway. At its heart, he says, was a serious technology problem. "If you remove the fur," he explains, "what it is is a computer that understands who’s talking, and listens." Bathiche was intent on solving an AI problem: the ability to figure out where to attend to in a room full of noise. "It was about a computer that understood whether you were trying to talk to it or not," he says—a smart Teddy Ruxpin.
The slow rise of digital personal assistants indicates a nascent demand for such a feature. "We’re still not there yet," says Bathiche, "but we will be." He reflects for a moment. "And then no one will laugh at this project anymore."