There are a handful of Windows vs. Linux TCO studies available. IBM and Opensource.org usually cites this one by the Robert Frances Group that was published in September 2002.
The rest of them, unfortunately, are all bankrolled by Microsoft. In the same way cancer studies funded by the tobacco industry always seem to put cigarettes in a favourable light, the majority of IT staff take these MS-funded TCO studies with a chunk of salt. So when Yankee Group and Sunbelt Software claimed to have released the first-ever independent Windows TCO vs. Linux TCO study, I was intrigued.
Even more intriguing was its conclusion: that Windows TCO was less for large corporations that have predominantly Windows-based, and more for smaller startup firms. These questions immediately come to mind:
Possible Conflict of Interest
- Sunbelt Software is a Microsoft Gold Certified Partner whose entire revenue is derived from selling Windows NT/2000/XP tools and utilities.
- Laura Didio, the primary Yankee analyst of this study, is infamously known as having championed SCO’s legal battles against IBM and the Linux community.
- The survey involved 1,000 IT administrators – however, the sample selection was not random. The survey was given to subscribers to W2KNews, a Windows NT/2000 newsletter published by Sunbelt.
- The study’s control case is of a company already running Windows systems. It would be logical to assume that sticking with Windows would be less expensive than migrating to another vendor, whoever that might be. After all, it is always cheaper to keep your existing car, even if it’s a lemon, than going out and buying another one.
Aside: In a way, this point inadvertently highlights the lack of interoperability within MS products; companies who are Windows shops are often “locked-in” with licensing contracts and sunk costs which leaves them helpless in negotiating pricing and support.
- No analysis on upgrade costs or marginal costs to scale, nor to the soft costs of running damage control for viruses and security exploits.