I was going to avoid posting anything about this week's departure of Bill Gates from his full-time role at Microsoft; everyone else is making a big deal over it, and I suppose if sentiment is a factor then it's time to pay its due, but in fact Bill has been considerably removed from the operational side of the business for some time now and inertia is such that there isn't likely to be any sharp changes in the business any time soon after his departure.

Then I read Joel Spolsky's and Mary Jo Foley's respective articles on What It Means that the world's richest programmer is leaving the world's most successful software company, and their general agreement on the fact that a programmer, not an MBA, is required to run a software company well, got me thinking. What exactly is their criteria for "well-run" that Microsoft fits the bill?

Their points are that, being a brilliant programmer, Bill could push down into a remarkable level of detail when evaluating the company's products, detail that a more conventionally bred executive could never approach. And although the company has long been too large for Bill to have his hand in every detail (although he managed to drill down far more than you might imagine, even later in his career), he was able to take that and set a culture which resulted in the same effect among top lieutenants and managers.

But did this ever result in really great products? Or are they paying homage to a rather geeky quirk that is certainly worthy of respect from other coders and nerds, but which had little ultimate effect on the success or failure of the end products?

Although the company has had great success, it hasn't been one resting upon the technical merits of its programs, at least not for many years now. It's been the result of marketing and a certain element of foresight which placed the business in control of the dominant platform in the industry early on, a fact which was leveraged both legally and illegally to achieve its current position bestriding the planet.

Spolsky talks of watching software companies being run by non-programmers in pejorative terms such as "...like watching someone who doesn't know how to surf trying to surf." He apparently doesn't spend much time watching companies like Apple (Woz was the coder; Jobs just does vision) or Sun (Scott McNealy has-gasp-an MBA). You might argue those are more properly hardware companies; but their code has been a significant component in their success. And I can't help but wonder if Spolsky and Foley have it backwards; Gates certainly paid tremendous attention to details, but in many cases they were the wrong details. Is Windows a towering technical feat of intellect? Perhaps. But my Mom can't figure out how to answer half the pop-ups it throws at her.

I respect Gates' intellectual prowess, but it strikes me that a non-programmer more attuned to the consumer's sensibilities and capabilities might in fact have churned out much better software over the years, or at least better in the sense of being more useful, and useable.

It strikes me that the role of the CIO is similar. Although I consider myself reasonably technical, I think my strength is that I don't have to be; it's not the only way I can relate to things, unlike many colleagues. I think a good CIO is one that can grasp the realities of the business and the challenges it faces. A great CIO can do that, and master the technical intricacies of his domain as well. But I wouldn't want a detail oriented CIO who couldn't relate to systems on the average user level.