(FORTUNE Magazine) – Like most things Microsoftian, the software giant's recruiting department has taken on mythic proportions. There's the story of the recent college graduate who was asked in a job interview how he would weigh an airplane without a scale. And then there's the tale of the software writer who had to prove his mettle by writing code on demand--with a programmer peering over his shoulder. Not your typical job interviews, to be sure, but market domination doesn't result from a run-of-the-mill human resources operation either.

To get the best people, the Redmond, Washington, company puts applicants through a rigorous hiring routine--after computers sort out resumes by looking for keywords. The man who runs it all is David Pritchard, the director of recruiting. No human resources lifer he, Pritchard worked at Microsoft as a public relations man, a software writer, and executive vice president Mike Maples's chief of staff and business manager before assuming his current position. He recently spoke with FORTUNE's Ron Lieber about how Microsoft woos its stars.

So what's Microsoft's secret of hiring?

What keeps the recruiting operation--and ultimately all of Microsoft--successful is three things: employee involvement in hiring, the process that we use to find candidates to come talk to us, and what we think are innovative tech-niques to interview them.

How do you get busy Microsoft employees to take hiring seriously?

To this day, Bill Gates might call a few people right out of college and say, "Hey, we're interested in having you come to work here, and I just want you to know that we're all aware of your skills and we think you'll have some positive impact on our company." It doesn't seem like something a CEO should be spending his time on, but if Bill and the vice presidents weren't involved, employees would probably think they didn't care, and if they don't, why should we?

Besides, the best thing we can do for our competitors is hire poorly. If I hire a bunch of bozos, it will hurt us, because it takes time to get rid of them. They start infiltrating the organization and then they themselves start hiring people of lower quality. At Microsoft, we are always looking to hire people who are better than we are.

How can human resources help the company avoid hiring the bozos?

Recruiters need to have a very clear idea about the needs of the division they are recruiting for. Though they might report to me, they are truly partners in those organizations. Recruiters attend those business meetings and get feedback from current employees, so they understand the short- and long-term business goals of that particular division. This helps them find the right people and meet hiring needs before they become critical.

How do you go about finding and attracting the best people?

We get about 12,000 resumes a month, and every one gets logged into a computer with information on when it arrived, what people thought of it, and when we sent a response. We've got computer programs that search the resumes for keywords that, for instance, would indicate the candidate knows C++ computer language. That helps, but the technology isn't flawless. So we've been taking a lot more time recently to have recruiters read resumes and even call the people up to see what they are like rather than simply deciding to fly them out.

In interviews you ask a lot of brainteasers. Do you like watching people sweat?

The view you're describing is the "I went through hell, so everyone else should" approach. It's that fraternity-hazing thing, and I just don't agree with it. The interviewing process is going to be stressful for the candidate no matter what we do. I think it's important for us to give people a process that lets them be successful, that lets us see how innovative or technical they really are. Whether they end up getting a job or not, we need to give them every opportunity to succeed.

So what kinds of brainteasers do you ask?

Oh, there's a list, but it certainly isn't something we publish. (We don't ask senior people those questions, only college recruits.) I always see reports in the paper about how we ask why manhole covers are round [so they don't fall in the manhole] or how to figure out how much water flows through the Mississippi daily.

We're not looking for the "right" answer. We're looking for the method, and we want them to talk about it. We want to see what information they ask for, like the length of the river or the flow rate at certain measuring points. It's a learning process.

One of the things we look for is smarts and experience, but we also want to know what they will bring here over the long term. Are they flexible? Can they learn new concepts? In this industry, things are changing on a daily basis, and if you're not capable of learning new things, you won't be successful. We might even teach them an algorithm in a morning interview and then ask them about it in the afternoon to see how much they have learned.

So interviewer No. 2 may call No. 5 and No. 6 and tell them what to ask?

Exactly. We try to keep it very well orchestrated so that throughout the day, people are aware of what technical questions have been asked.

What wouldn't impress you?

Well, I might be concerned if someone said to me, "That's a really stupid question." That's not a wrong answer, but I'd certainly want to have a conversation about why that is. In my mind, though, the worst way to answer a question is "I don't know, and I don't know how to figure it out."

What do you remember from your own interview 15 years ago?

I met with many people, but what stands out was [executive vice president] Steve Ballmer asking me point blank, "What excites you?" That's a question I often use myself. If I can get you to start talking about your business, I can ask a whole pile of questions that go along with that and make the interview as much an educational experience for me as it is an interview for the candidate. That will show me how knowledgeable you are about that business, how you got to be that knowledgeable, and what your thinking is about that business's future.

Is there a preferred answer to that question? What if someone talks about sea kayaking instead of software writing?

No, I just want to hear about something that excites them. If you're a marketing person, you might not know anything about software. It might be cars, it might be Jell-O. Steve Ballmer worked on a dessert called Coldsnap at Procter & Gamble. To this day, he is passionate about it--he's still got a package of it in his office actually--and he can tell you why it failed and tie it back to whatever he's doing now.

I hear Gates has called some people he's interviewed stupid.

Well, he might jokingly say, "That seems kind of stupid to me," or he could say, "I don't understand how that would work." It's not a putdown necessarily, but a learning process for Bill. He's trying to understand how that person is thinking. Or if Bill is interviewing a senior person, he might want to see how they handle a tough question by saying, "Why do you want to work here, after all the nasty things you've said about us in the press?"

What about your approach to testing? Lots of companies conduct psychological tests. Does that interest you?

It doesn't really interest me much. In the end, you end up with a bunch of people who answer the questions correctly, and that's not always what you want. How can a multiple-choice test tell whether someone is creative or not?

What do you do to snag the best candidates?

You have to sell them on the idea that this is an innovative place to work. Even though we're a much bigger company than we were even five years ago, people still have this perception that we're just producing the next version of the same product over and over. They don't always realize all the innovative things we've got going on here because we don't often talk about them in the press. The best way people can understand that is by spending some time here and seeing that we're still organized in such a way that you as an individual can see your mark on a product people will use all over the world. That's what turns people on.

When has the Microsoft approach to recruiting failed?

I always look at this when people leave. If anyone leaves within the first 12 months, I like to know why. I try to see what went wrong. A lot of times it's nothing. We didn't do anything wrong, and neither did they. They can be god-like in an interview, but you just never know. If hiring people was so easy that we could say turn up the volume here or down there, that would be great. But we're dealing with individuals in a company where the best asset is our intellectual horsepower. You just can't measure it correctly all the time.

Research Associate Erin M. Davies