Tuesday, January 27, 2004
Studies on Linux help their patron: Microsoft
Critics question how independent the analyses are
Microsoft Corp. is touting the results of "independent analyses" in its latest effort to show corporate decision-makers the merits of its Windows operating system vs. Linux, its biggest open-source competitor.
The studies were, in fact, performed by well-known, independent research firms such as IDC, Giga Research and Meta Group. But the reports themselves tell the rest of the story: They were conducted "at the request of Microsoft," "commissioned by Microsoft" or "prepared under contract from Microsoft."
Microsoft, in other words, paid for the studies to be done.
Some analysts say that doesn't meet the definition of "independent." Microsoft says it does. But no matter the outcome of the semantic debate, the situation is drawing new attention to the company's controversial practice of funding and publicizing research to support its competitive claims.
The reports at issue are available through a new Microsoft Web site called "Get the Facts on Windows and Linux." Although the company has sponsored and publicized individual studies in the past, the site collects them in one place, with an overarching message: "Leading companies and third-party analysts confirm it: Windows has a lower total cost of ownership and outperforms Linux."
Cost and performance are two central issues in the debate over open-source software, which is produced by communities of developers who make the underlying code available for free use and modification. Microsoft says the cost of using Linux, particularly on corporate computer servers, is more than the cost of using competing Windows software, considering such factors as installation, development and service.
More than defending the company's position, however, the "Get the Facts" site and a related ad campaign show how seriously Microsoft is taking Linux as a competitive threat to one of its biggest products.
"As Linux became more mainstream, as Linux became more commercial, as IT professionals wanted to make more pragmatic decisions on it, they needed a set of data to help them sort through their decisions," said Martin Taylor, Microsoft's general manager of platform strategies and its point man on Linux issues.
In that way, the research fills a critical gap, he said. Companies "couldn't say, 'Maybe I should just choose Linux because there were 52 Slashdot postings saying that Linux is better,' " he said, referring to a popular Web site for technology news and commentary. "That's really what drove a lot of the emphasis to say, 'Hey, fine, let's really give a set of facts.' "
Taylor said he considers it correct to call the studies "independent" on the site because the firms themselves are independent and the studies in question, though funded by Microsoft, are based on the firms' own methodologies. Findings are also presented in such a way that they can be duplicated by others, he said.
The Web site itself makes no reference to funding for the eight third-party studies it cites. But several of the studies, available for free download from the site, indicate in their text that Microsoft funded them.
"It's very clear what we commission," Taylor said. The results would be the same regardless of the funding, he added, saying the company doesn't influence the outcome.
But others aren't convinced that Microsoft is justified in its use of the phrase "independent analyses" to characterize the studies on the site. More than a question of wording, it's a debate over how much weight a reader should give any given report.
"I think the firm is independent, but it is a stretch to say the report is truly independent," said Frank Catalano, a technology consultant and analyst based in the Seattle area.
"If Microsoft or any company pays for a report to be written about itself, you have to take what the report says with a grain of salt. Some will be truly independent, but others will either consciously or subconsciously slant the report."
One of the most outspoken on the issue has been Steven Vaughan-Nichols, editor of eweek.com's Linux & Open-Source Center.
"This is not to say that all research is crooked. It's not," Vaughan-Nichols wrote in an October column, after a controversy over one Microsoft-funded study. "Most magazine and other independent research is done with a desire to uncover the facts, nothing less and nothing more.
"I'm also not saying that if Microsoft went in and demanded a study showing that Windows for Pizza 1.0 delivered 12 percent more pepperoni per pie that it would get a report saying it had 12 percent more pepperoni," he continued. "I am saying that when vendors sponsor a report, they always get to put a thumb on the scales. It may be subtle; it may not. But the thumb is usually there."
Michael Cherry, lead analyst for operating systems at Kirkland-based research firm Directions on Microsoft, said he had heard about Microsoft's "Get the Facts" site but hadn't looked at it because he had "assumed it was propaganda."
"I think it's great that Microsoft is providing some information about it, but it's just like if I went and read the Red Hat site -- you have to consider the source," he said, referring to Linux vendor Red Hat Inc.
One problem, he said, is that companies will tend to keep under wraps the results of commissioned studies that turn out unfavorably. That means the public may get only part of the story when it reads a report sponsored by one of its subjects. "We're only seeing the ones they want us to see," Cherry said.
Microsoft isn't the only company that publicizes competitive research for which it pays. But the company's prominence in the technology world -- and the divisiveness of the Windows vs. Linux debate -- has helped bring the issue to the forefront.
It reached a boiling point last year, when Microsoft publicly disclosed the results of a study by Forrester Research subsidiary Giga Research that showed cost advantages of Windows over Linux. Microsoft commissioned the study, and Forrester Research Chief Executive Officer George Colony said in a public statement afterward that the firm erred in letting Microsoft talk publicly about the results.
Forrester had policies in place prohibiting clients from publicizing commissioned research comparing their products to those of competitors, said Dan Mahoney, Forrester's senior vice president of research for North America. (Commissioned research not making competitive comparisons can be publicized under the firm's policy.) In the case of Microsoft and another client, however, Forrester mistakenly approved the external use of the reports.
The problem with publicizing such reports is the public perception that the client somehow "bought" or influenced the results, Mahoney said. It's a mistaken impression, he said, noting that the firm stands by the integrity of the research. But it's a common impression nonetheless, and it's something the firm wants to avoid.
The funding "never affects the research -- at least at Forrester it doesn't," Mahoney said. "There are some vendors, not the one in question here, but some vendors who say, 'I'd like you to write this thing, and here's what I'd like it to say.' The great majority of research firms at that point will say, 'Thank you very much,' and hang up, or point them toward one of the guys who does that."
The very report that caused the stir for Forrester is included on Microsoft's "Get the Facts" site, which was launched this month. Had there not been the one-time arrangement letting Microsoft publicize the results, "I would not have wanted that to happen," Mahoney said. He noted that the initial approval was Forrester's error and that Microsoft is within its rights to keep publicizing the study.
Meta Group allows its clients to publicize research they fund, so long as the research isn't cited out of context to make a competitive comparison, said Samantha Finnegan, a firm spokeswoman. The "Get the Facts" site includes a report in which Meta audits the methodology used by Microsoft in a study about Linux.
Microsoft's Taylor acknowledged that corporate information-technology managers may, after reading the Microsoft-funded studies, consult alternate research from a Linux vendor or similar source. In fact, he said, "That's kind of the behavior that we want to drive from this" -- customers looking closely at the facts.
The type of research the company has been funding isn't something the outside research firms have been interested in conducting on their own, Taylor said. It's "just not exciting research for them to go do," he said. "I told them all, 'Hey, if you guys do these on your own, I don't want to have to commission these things.' "
Linux isn't the only subject on which Microsoft has published commissioned studies. When the company launched its Microsoft Office System 2003 last year, it cited a study it paid Navigant Consulting to conduct. The study showed the ways the software streamlines business processes and saves money.
Dan Leach, group product manager for the Microsoft Office System, was asked at the time whether Microsoft would have publicized the results if they hadn't been positive. He answered that he had been so confident in the software's benefits that it "was never going to be a question."
Advertising Sections:Local Businesses: Business Directory
Send comments to firstname.lastname@example.org
©1996-2010 Hearst Seattle Media, LLC