Earlier this week I was part of a small group of journalists and bloggers who spent the day questioning and listening to Microsoft’s top executives from Xbox, Office, Bing and other parts of the company. Just about every senior leader (except for those from Windows) took part in on-the-record briefings.

It was an annual event called TechForum, led by Microsoft’s Craig Mundie, the company’s longtime research and strategy chief who recently announced plans to retire next year. He’s serving as senior adviser to Microsoft CEO Steve Ballmer in the meantime.

There was a lot to absorb during the day, and I’ve been posting this week about some of the projects and trends that struck me during the day. But before the event fades from memory, there was one exchange toward the end of the day that’s worth sharing.

It came when Mundie was asked why Microsoft has struggled to keep up with Apple and other competitors over the past decade. It’s a common question, but Mundie gave a particularly introspective response, starting by cautioning that it wasn’t as simple as simply getting faster. Here’s an extended excerpt from his comments.

In many of the areas where people said, oh, you’re behind — Apple’s products are a great example, they did music players, and then they did touch devices and then they did phones and then they did tablets — well, it turns out that we had all four categories of those devices in the marketplace more than one year or two years before Apple did their first one. But for a whole variety of reasons — they were just business choices we made at the time — we didn’t end up capitalizing on them. And once we didn’t capitalize on the lead we had to fight our way back. So in each of those categories you go from leading to lagging not because we didn’t have the technology, and not because we didn’t do it fast enough. Some people said we did it too soon. It is a delicate balance.

On the other hand, we also shipped what have been our major products, particularly Office and Windows, at a scale that nobody else ever does. Our beta tests our bigger than the lifetime deliveries of most people’s software products. In every category. The only ones that even come close now are people doing phones, because phones as a device category have gotten big enough in absolute volume that you kind of get into the same zone. …

So those things had a long lead time, when we were packaging them up and testing them. And then shipping them in this, I’ll call it, increasingly uncontrolled hardware ecosystem. And so one of the things that has changed is we’re trying to provide more controls on the hardware ecosystem so it’s not so unruly, you could say, because that variety our testing process more complicated. And then the other thing is, because we’re able to move these things more into the services environment, as opposed to just shipping them as an upgrade, we bear the cost of managing the upgrade.

If you listen to an enterprise CIO over the last decade or 15 years, most of them would tell us, every two or three years, that’s absolutely the most frequently we would want to change.  In fact if you look at it, most of them who have paid for our software are skipping whole generations. They’re only upgrading every five or six years. In that environment, there was more pressure to be deliberate.

When we run these things in our own infrastructure, and you just buy a service, like you’re buying electricity, we say, hey, you know what, every 90 days we add a new feature. They say, great, we get a new feature. Didn’t have to do anything. Didn’t buy any new hardware, didn’t have to upgrade all the machines. So there is an element of the services component that allows the cycle time to be reduced. (Microsoft Office president Kurt DelBene) acknowledged it today when he talked about the Office evolution, and the Bing product evolution has been a whole rapid cycle. They introduce a new generation every 90 days.

So I think the company has the capability to engineer at whatever speed we need to. And it’s clear that there are certain categories now where getting it done more frequently is beneficial, and I think you’ll see us do that.

A lot to chew on there. It feels like Microsoft has gotten caught in the middle of the different markets it is trying to serve, from businesses to consumers. As Mundie makes clear, it’s not good enough to have the idea. The company needs to execute quickly and make the right business decisions.

Yes, consumers are driving tech trends in businesses these days, but the needs of businesses remain very different from the needs of consumers. The big question in my mind is whether Microsoft can become more nimble while trying to focus on both.

At the end of the day (literally) I still needed to be convinced that the company has made the fundamental changes to ensure it doesn’t end up trying to come from behind, again, when the next big thing arrives.