Is "market share" thinking "old school" thinking?

The tech industry spends a lot of time talking about market share. Whether it's hardware (iPod, iPhone, Macs ...) or software (Windows, Linux, Mac OS, browsers ...), we seems to want to read a lot into this particular metric. But is "market share" thinking "old school" thinking?

I think that the mistake that many fall into it to look at market share and equate this to importance, the idea being that ultimately there's "one XYZ to rule them all" (XYZ can be an OS, a computer, a cell phone or whatever). This is old school thinking, because over time all markets fragment. Look at markets such as cars, home appliances and consumer electronics and watch how over time these changed over time and how an increase in the number of companies competing for users cash, along with changing user tastes, has meant eroded the idea of one single product dominating a particular market. Even a killer product such as the iPod is vastly outnumbered by cheap junk music players. The big difference here is that Apple has a massive PR and ad budget to tell us how important the product is, while the companies making the cheap junk players don't.

Market share for software is even more dubious. What does it mean, for example, when Mac OS X gains an extra market share percentage point? Does it mean more Mac sales, or is it that people are using their Macs more? Or does it just mean that Steve Jobs is sitting at home in front of a sea of Macs, furiously browsing away? Does it even matter?

Equally, what does it mean that Linux has hit a market share of 1%? It only begins to mean anything if you take "Linux" to be a monolithic entity, which it isn't. Without information on particular distros, the 1% figure (which is itself debatable) is meaningless beyond being tokenistic.

When I consider market share I'm usually only interested in one thing, trends. While specific numbers are open to debate and interpretation, trends are a bit more solid and allows you to get a feel for how things are changing over time. Trends are interesting because you can see a cause/effect correlation between an event and a trend shift. These are far more interesting that the numbers themselves.

The bottom line is that ultimately the idea of dominance is dying, and that people buy and use what works for them. Sure, there are going to be people who follow the herd and buy what's popular, but even the notion of what's popular is transient, with new products replacing older ones faster than ever. Even dominance is fleeting.

Is the notion of market share meaningless nowadays?