At the end of July I spent two hours with Gary Flake for a recent Wired story about Ray Ozzie. Flake is one of 10 technical fellows at Microsoft, but more importantly he is one of the people on the front lines of the innovation rethink Ozzie et al are pushing at Microsoft. He runs something called Live Labs, and his job is to make sure that the company gets more bang for the buck out of its research arm – something it has come under increasing criticism for. Microsoft hired Flake about a year ago from Yahoo, where he ran R&D. I found him to be thoughtful and candid about the issues Microsoft is facing, but as happens often in journalism a lot of what we talked about didn’t make it into the story. Below are edited excerpts of our conversation.
FRED VOGELSTEIN: Why does innovation in Microsoft Research take so long to get incorporated into Microsoft products? How is Live Labs going to help fix it? Some say the problem is that MSR can’t get the product teams to pay attention to its work.
GARY FLAKE: I think there's always room for improvement. We actually have had a lot of really great successes. Folks at MSR are having tremendous impact on the search team, for example. But there is some truth to what you are asking.
Imagine, if you will, I'm an engineering manager. There are certain problems I have to solve that I understand, and there are certain things that are a bit of a mystery to me. And let's also say that I have finite resources. So if I'm an engineering manager, it's rational for me to make the investments in the things that I understand, not the things I don’t understand.
Let me draw you a diagram. I might say there's Windows, there's online stuff over here, there's Office, there's Xbox, there's mobile. If I’m an engineering manager this might be my schema for how I think the world is organized, and how I try to categorize things.
Now, if I'm a scientist at MSR, I think about programming languages, algorithms and theory, machine learning, databases, distributive systems, operating systems -- a whole list of other things.
So the languages that these two groups of people use are actually wildly different. They have a different set of priorities. They have different beliefs in terms of what's important and what's not important. They're thinking about things on different timelines. So it shouldn't be surprising that there's a bit of an impedance mismatch.
And so what we (at Live Labs) do is come up with a way of characterizing all this -- it's really something of a Rosetta Stone that helps communication between these two worlds.
FRED VOGELSTEIN: How are you taking the Live Labs philosophy and extending it out into the rest of the company?
GARY FLAKE: What we've done is we've put a little bit more structure in terms of how business leaders should think about managing their own activities so that people not only have permission to do innovation, but the requirement to do innovation. So one of the major things that we're doing is actually making a number of the performance metrics a function of how you're executing on innovation.
So let me explain that. We haven't locked in on what innovation means, because it's one of those horribly overused words that means different things to different people. But one that seems to resonate with the most people is the following: Innovation in the context of our company is anything that intuitively you believe is the right thing for the company to do, but goes against your existing performance metrics.
FRED VOGELSTEIN: Anything that would get you in trouble?
GARY FLAKE: Sort of, but it's still the right thing for the company.
Let me give you an example: My job could be building a Web search engine, but as a requirement for building that engine I’d also have to build infrastructure software to allow one person by themselves to administer either a thousand or 10,000 machines. Well, hell, that piece of infrastructure software is something that's clearly good for the company. But this person has a set of performance metrics that are about making search, not infrastructure software, number one.
FRED VOGELSTEIN: And also there's someone who's in charge of managing all those servers who's going to get their back up about somebody trying to show them up.
GARY FLAKE: Yeah, yeah. And so what we are trying to do is carve out a little bit of a safety net, a career safety net, where not only are you allowed to do things like this, but there's the expectation that you will do things like this. So that's part one.
Part two is that we've actually made a number of investments in … for lack of a better way of putting it, let me just call them innovation collaboration tools. So we have a system that we use in house called Idea Factory. It's about a year old. Basically it's a tool that allows people across the company to submit ideas and give feedback on those ideas to one another.
FRED VOGELSTEIN: Are you saying that for the first time there's actually a system where everybody can throw ideas in and submit them for peer review -- that everybody sees up and down the company.
GARY FLAKE: That's right. And the thing that's really exciting is that we've fundamentally changed things. Now, if you don't participate in an innovation project, your numbers aren't going to be as good. But we are also making it so that one of the ways you can fulfill your innovation goal as a business leader is to cherry pick something out of Idea Factory.
FRED VOGELSTEIN: So in other words if you're Jeff Raikes you get a bigger bonus if --
GARY FLAKE: If you take some thoughtful risks. Yep, yep. So this is, like I said, I want to be really clear, this is a work in progress, but this is to a great extent inspired by some of the changes and processes that we've made within Live Labs.
FRED VOGELSTEIN: The history of Microsoft has been that every innovation has to in some way help promote the sale of Windows, Office, Internet Explorer usage versus Firefox, and so on. How do you square your mission with that?
GARY FLAKE: Look, look, let me hit the pause button. I don't mean to be rude on this, but the more you know about me, the more you would realize that that couldn't possibly be counter to who I am. I've written over a hundred thousand lines of open source code. I've been a user, I've been a contributor, and I'm still a maintainer of open source projects in the outside world.
Let me tell you about a conversation that I had with Ray Ozzie, Kevin Johnson, Steve Ballmer, and Bill Gates before I'd accepted the position here.
I walked into this company and said, look, here's my DNA. I think the Internet is more important than software, period, full stop. Those are my biases, my opinion, and I think that thoughtful people can have different opinions on this. I'm not religious about this, but when I kind of take a look at the evidence, I think that that's the thing that motivates me to do what I do - the potential of the Internet. Now, to the extent that Microsoft’s assets are aimed towards pursuing that mission, I'm all on board, I'm a team player. To the extent that there's a little bit of conflict, I'm all about pursuing my original goal.
And so I said let's all do an expectation level set with one another -- that if I'm coming here, these are the goals that I'm pursuing. It was within their right to embrace that sort of attitude or suppress that attitude, and I was actually warmly embraced with that.
FRED VOGELSTEIN: Why come to Microsoft instead of Google, or staying at Yahoo?
GARY FLAKE : Three reasons. First, I thought “Who has the means to do interesting things?” Yes, all three companies have tremendous assets, but when I went out and considered Microsoft's connection to both the desktop, small business, mobile, living room, the whole kit and caboodle, that was pretty exciting. What really pushed it over the top for me was that they asked me to evolve the relationship between MSN and MSR. I happen to think that MSR is the greatest computer science research institute in the world. The people there are rock stars, I’ve been following their work or they've been my friends and colleagues for over a decade. And so I gave Microsoft the edge in terms of the raw resources they had at their disposal.
The second thing is basically I saw Yahoo and Google as companies that had a fairly well defined mission in terms of what they wanted to do. Microsoft is clearly a company at a transition. We're very public and open about that, that we're transitioning into a world and an industry that's gone from software that's shrink-wrapped to software and services. And for me being part of the transition was a very, very exciting thing.
The third thing was that because of my experience working in the Valley at various search engines, I felt disappointed in the industry for the advances it has made in search technology in the past few years. Look at some of the major changes in the user experience over the past few years: We've got spell correction and tabs. Those are really the two big ones. And, yeah, we've made some modest improvements in relevance in the aggregate. But when you look at it over five years it’s a bit disappointing. Google and Yahoo can’t introduce innovations and changes to the search engine industry that put their business model at risk. They've painted themselves in a corner in that sense.
We'd like to do better on search, but at the end of the day we're not painted in a corner in terms of what we can do or what we can't do.