Studies related to Microsoft

There have been a number of studies related to Microsoft. They are always a source of great controversies, since the studies are often funded by people or companies having a stake in one of the sides, and there are a lot of advocates on both sides.

IDC study: Linux TCO vs Windows 2000 TCO
In October 2002, Microsoft commissioned International Data Corporation to determine the total cost of ownership (TCO) for enterprise applications on Windows 2000 versus the TCO of Linux on the same enterprise applications. IDC looked at security and other infrastructure tasks, and Web Serving. According to the report, Windows 2000 had a lower TCO for four infrastructure items and Linux had a lower TCO for web serving. IDC's report was based on telephone interviews of IT executives and managers of 104 North American companies in which they determined what they were using for a specific workload for file, print, security and networking services.

IDC determined that the four areas where Windows 2000 had a better TCO than Linux &mdash; over a period of five years for an average organisation of 100 employees &mdash; were in the use of file, print, network infrastructure and security infrastructure. They determined, however, that Linux had a better TCO than Windows 2000 when it came to web serving. The report also found that the greatest cost was not in the procurement of software and hardware, but in staffing costs and downtime. The report did not take into consideration the impact of downtime to the profitability of the business (although they did apply a 40% productivity factor, in order to recognise that employees are not entirely unproductive during periods of IT infrastructure downtime) though it did find that Linux servers had less unplanned downtime than Windows 2000 Servers. They found that most Linux servers ran less workload per server than Windows 2000 servers and also found that none of the businesses they interviewed used 4-way SMP Linux computers. IDC also did not take into account specific application servers &mdash; servers that need low maintenance and are provided by a specific vendor &mdash; when they performed their study. The report did emphasise that TCO was only one factor in considering whether to use a particular IT platform, and also noted that as management and server software improved and became better packaged the overall picture that was being shown in their report could change.

Criticism
This study was one of the most criticised studies, since Microsoft funded it. Many editorials were written about the study, including one by The Register.

Jon "maddog" Hall has stated that "the TCO is useless, because that's only about what the software costs- it doesn't talk about what the software is worth." He explains that from a consumer perspective, cost is only important when considered in relation to the utility gained from the software, pointing out the frequency of downtime experienced by users as some variables which should be considered in a making a rational decision.

Cybersource TCO study: Linux versus Windows
Melbourne-based Cybersource compared in 2004 the Total Cost of Ownership (TCO) of running Linux versus Windows in the enterprise. Its studies found that Linux was 36% cheaper than Windows, when taking into account the software cost as well as service, support and upgrades. The study done in 2004 was an update on their previous studies in 2002, and found the same results.

"Get the Facts"
Get the Facts was an advertising campaign launched by Microsoft in 2004 to convince users to switch from Linux to Windows servers. It was originally focused on comparing the total cost of ownership (TCO) of Linux to Windows, but later compared reliability, security and interoperability. Microsoft claims that its products have a lower overall TCO than open source programs, proposing that the Windows server platforms offered improved ease of use leading to these savings.

As part of the "Get the Facts" campaign Microsoft highlighted the .NET trading platform that it had developed in partnership with Accenture for the London Stock Exchange, claiming that it provided "five nines" reliability. After suffering extended downtime and unreliability the LSE announced in 2009 that it was planning to dump Microsoft and switch to Linux in 2010. .

Response
Microsoft's figures are disputed by a variety of organisations, notably Novell and The Register. Some websites suggest that some common inaccuracies in Microsoft's figures stem from including figures for Unix and Solaris with figures for Linux. Individual Linux and Unix administrators may have higher salaries than Windows administrators, but they tend to be more efficient and thus able to handle more servers.

In 2004, the Advertising Standards Authority (ASA) of the UK warned Microsoft that an ad from the campaign which claimed that "Linux was ... 10 times more expensive than Windows Server 2003", was "misleading", as the hardware chosen for the Linux server was needlessly expensive. The ASA's complaint was that "the measurements for Linux were performed on an IBM zSeries [mainframe], which was more expensive and did not perform as well as other IBM series." The comparison was to Windows Server 2003 running on two 900 MHz Intel Xeon CPUs.

In 2009 Microsoft launched another "Get the facts" campaign which compared Internet Explorer 8 to Mozilla Firefox 3.0 and Google Chrome. It was criticized for misleading comparisions. Microsoft says on the site that IE ties with Firefox and Chrome on web standards, however, at the time the latest Firefox beta scored 94 on the ACID3 test, Google Chrome scored 100, and IE only scored 20 on the test. Another misleading statement was that IE is the only browser with tab isolation and crash recovery, however Chrome also have them.

Windows XP Home Edition vs. Windows ME Public Report
American Institutes for Research conducted a comparative usability study of Microsoft’s Windows XP Home Edition and Windows Millennium Edition (ME) in August and September 2001. The goal of the study was to assess the differences in usability between the two operating systems’ user interfaces.

Thirty-six individuals participated in the comparative evaluation. AIR recruited 12 people in each of three user experience groups—Novice, Beginner, and Intermediate—as categorized by Microsoft’s Windows Knowledge Screener. Each participant performed the same tasks using both operating systems. Further, participants had no experience with either Windows ME or Windows XP before participating in the study. In order to eliminate learning effects, half of the participants performed the tasks using Windows ME first, and half used Windows XP first.

The study was conducted as a task-based evaluation in which participants attempted to complete a series of timed tasks. The set of tasks was designed to reflect ordinary usage of both operating systems. Intermediates attempted all 30 tasks, while Novices and Beginners attempted 28 tasks. Test administrators recorded whether or not participants successfully completed each task as well as the amount of time it took them to complete the task. In addition, they administered the System Usability Scale (SUS - © Digital Equipment Corporation, 1986), a ten-question survey that yields a score describing a system's overall effectiveness, efficiency, and user satisfaction.. At the end of each participants’ second test session, the test administer also asked which operating system they preferred and why. The results of the study suggest that, overall, the Windows XP interface is an improvement over the Windows ME interface. Primary findings of the comparison are as follows:
 * On average, participants were able to complete 17.20% more tasks using Windows XP than using Windows ME.
 * On average, participants completed the same tasks faster with Windows XP than with Windows ME.
 * Prior Windows experience appeared to correlate with task performance more so with Windows XP than with Windows ME. With Windows XP, Intermediates successfully completed the largest percentage of tasks, followed by Beginners, then Novices. With Windows ME, Beginners completed the largest percentage of tasks, followed by Intermediates, and then Novices.
 * Statistical analysis of the System Usability Scale (SUS) indicates that participants subjectively rated Windows XP as being more usable than Windows ME.
 * After using both operating systems, 78% of participants preferred Windows XP over Windows ME.