Steve Ballmer's Big Moves

I was thinking more about the BlackBerry email devices. You make Microsoft Outlook, which is the dominant email application, and which keeps adding features all the time. Along comes BlackBerry with a much smaller device, where the screen is tiny, the commands are limited, and the keys are much too close together. Yet it's incredibly mobile -- and everyone wants to have one. What's going on?

Are they really a threat? Are they redefining the core scheme of the way email gets done? Do you really know very many people who would get rid of their personal computer to have a BlackBerry? No. The company that makes it totally uses our Exchange software as its back end. It's primarily an add-on. It would be trivial to do one. But even so, you have to ask, Will appliances like the BlackBerry disrupt PCs? The answer might be yes. So if we don't fight that battle, we'll lose.

About a year ago, you acknowledged that every new version of a software package gets bigger in terms of the number of lines of code, and you portrayed that as a good thing. Do you still think it's a good thing? Or at some point, does big become bad?

Big is great. Big per se is not a problem. If you look at how fast hard disks and machine-processing capabilities are growing, software is relatively smaller today than it's been at any time in the past 20 years. If you use size for more complexity, that's bad. But if you use size to bring more simplicity, that's fine. In our next release of Office, we put in more bytes to make it easier to use.

Tell us a bit about your current thinking on outside partnerships. It seems as if every company is more attuned than ever to working with a whole constellation of allies. What's the right way to make that happen?

That's the $64,000 question for us. We have five or six businesses, and we have different needs for partners in every one. There's no way we can succeed in isolation. In every business, the partner ecosystem is very important. But when people come to see us about partnerships, it isn't even obvious to me where to send them.

I had a company in here recently that builds -- I'd better not say exactly what, but let's call it middleware that they've integrated into Windows. It would be good for them and for us to build products that could be embedded in other things and that would therefore be tied to our embedded-systems group. That could produce solutions that have something to do with what we're doing in the consumer area. And those solutions would need development tools to go with them.

You're right. If you aren't careful, you'll end up giving them a scooter and sending them to every single building on your campus.

Exactly. We do too much of that. So the question is, How do we get a crisp point of contact? How do we get somebody within Microsoft who can speak powerfully for multiple businesses?

The Internet makes all of this more complicated. If it weren't for the Internet, we wouldn't be looking at MSN. We wouldn't have dot-net middleware. We had our partnering model pretty well taken care of, really, pre-Internet. It only involved Windows, Office, and a couple of servers. And then, darn -- along came that Internet. It brought so many more people that we need to touch -- and with whom we need to form partnerships in different ways.

How about within Microsoft? How do you get all 40,000 of your people to think about the Internet with a new sense of urgency?

It's not a problem to get this company Internet-oriented. It's really pretty easy. People act as if it's a hard thing, and it's not. The harder thing is to get people to agree on what you want to do. We need enough consistency in our platform and in our vision that we take customers in one direction instead of six. When you need to do that, you can't make everyone happy. Sometimes you try, but you almost never can.

In the past few years, with the dotcom mania, some people left because they thought they could make a lot of money at a startup. That has changed. We're back to doing better at recruiting and keeping the people we want. And we're spending much more time focusing on the quality of the job. We're thinking hard about how to keep jobs big and full of impact. That's the key: doing more than just fixating on compensation.

We have a culture in which people really want to move quickly. So senior management has to keep all of this commotion organized, and sometimes that means slowing things down. Our people complain that we move more slowly than we used to. But some of that more measured approach is appropriate, and it's much better than being stuck in a situation where the CEO has to invigorate a place and keep shouting, "Come on guys, let's move!''

George Anders (ganders@fastcompany.com) is a Fast Company senior editor based in Silicon Valley. to learn more about Steve Ballmer and Microsoft, visit Microsoft on the Web (www.microsoft.com).