What exactly is a market share anyway?

My co-author made a post about the extremes of Linux market share, citing statistics from Net Applications. The stats in that particular post gave Linux a meager 1% and Windows a whooping 92% market share!
What boggles my mind is what people mean when they talk of market share. In that report for instance, the stats actually refer to desktop market shares. Now that then brings to mind the basis of those stats. How are they really measured such that Linux has 1% and Windows 92%? What sites were used in the aggregation to arrive at such a conclusion?
I am naturally a person who always takes stats of whatever nature with reservation. A look at the definition of Market share by Wikipedia states
“…according to Carlton O’Neal, the percentage or proportion of the total available market or market segment that is being serviced by a company. It can be expressed as a company’s sales revenue (from that market) divided by the total sales revenue available in that market. It can also be expressed as a company’s unit sales volume (in a market) divided by the total volume of units sold in that market. It is generally necessary to commission market research (generally desk/secondary research, although sometimes primary research) to estimate the total market size and a company’s market share.”

This definition is really interesting. It is worthy to note that if for instance, desktop market share is actually going to be determined by the number of units sold, then Linux is likely to have less than 1%. Then too if we take the part that states that the proportion of the total available market that is being served, Linux would likely make about 10% of that.
I believe making an accurate measure of the market share of Linux on the desktop is very difficult in the best of circumstances. For instance, at home, I use a HSDPA which runs very slowly on Linux for whatever reason. Thus I am compelled to boot into Windows XP anytime I need to do some work online. This probably means my unit may not be counted as part of the total statistic. 
Also the kind of sites that are used in the aggregation is of much interest.f you take this site for instance, Linux will have a whooping 80% and Windows 7%, leaving the rest for Mac and the rest to share. However, if you take a site like NYtimes, then Windows would certainly win there. To estimate the market share of an OS on the desktop is no small task. It is also one riddled with inaccuracies and errors, not to talk about subjectivity.
To say Linux has X% of market share to me, is just for convenience. I believe the true number of Linux users on the desktop cannot be accurately measured. All that can be done would be to do some subjective maths. Windows no doubt, has the lions share of the market, but Linux has moved form the time when it was a pushover to a force to be reckoned with on the desktop market.
However, on the server side, Linux is the king. Pure and simple. No stats is needed to prove that. Anyone for example, of the 200million hits Google receives everyday is actually processed on a Linux system.

In short, market share is mostly a subject measure of how a particular product or service is fairing relative to others on the same market. This measure will vary greatly even when the factors on which they are based are varied in the slightest way. Linux on the desktop is small, but no longer a pushover. What’s your take?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.