This odd headline in PCWorld, "Windows 7's Market Share Gains Start to Slow," reminds me again that the enterprise software and IT market-share silly season is upon us. On May 7 I posted about the Gartner absurdity that BEA was one of the leading middleware providers of 2008,except for that little problem that its growth rate plummeted vs. 2007.
Of course, Windows 7 market share can't change trajectory because it is not yet in the market place. And there really isn't a Window 7 market separate from Vista, XP, ME, etc. anyway. And a measurement based on how one company's -- NetAppllications -- clients access certain Web sites is totally unscientific. These are just a few of the problems with PCWorld's highly parsed headline and less than thoroughly reported article.
These two examples are extremes; other misleading market share information is more subtle. But what does enterprise software market share mean and does it matter to anyone but the vendors involved? If the answers matter to IT managers and staffs in some sense other than bragging rights, then it is worth digging deeper. Some of the subtle sidebars of the market-share discussion include where is the product used, what is the metric (money, units shipped, renewal rates, etc.), what does middleware (plug in your favorite IT buzzword) even mean?
The easy answer to the "does it matter to IT folks?" question is: if it matters to your suppliers, it matters to you. Suppliers you have invested in heavily go out of business based on market share statistics (you might be having the same issue if you are thinking of buying a new GM or Chrysler product). Or they are less able to maintain or upgrade your software.
And ultimately, based on market share, Eurocrats and apparently the Obamanistas may fine them out of existence (or at least raise your costs) if some market share statistics are twisted from "a large majority of us like this product" to "if a large majority of IT folks like a product, its developer must be breaking the anti-trust laws." The latter is a particularly bad use of market share data because it means governments pick and choose which enterprise and consumer software and other IT products you can use.
As the market share silly season continues, I'll keep an eye out for other examples. Consider: Is Quickbooks still the leading ERP software? How rapidly is Google-funded Chrome overtaking Google-funded Firefox? Is IBM's Linux business overtaking its mainframe-software business? If you have any questions like that, do not hesitate to e-mail me or leave a comment.