By Brian Proffitt
Managing Editor
Analyst’s report reveals that Linux is more expensive than
CP/M!
News at 11!
And so it goes. Analyst reports on the so-called expense of
using Linux are a dime a dozen these days, with one analyst group
praising Windows and another anointing Linux with the same praises.
It is hard to know who is right and who is wrong. It becomes even
more frustrating when it is revealed that some reports are
sponsored by Microsoft and other proprietary software companies,
for the seemingly express purpose of detracting Linux and open
source software.
But are analysts’ opinions really bought and paid for, as so
many Linux advocates maintain?
Here’s a shocker: they are, but not in the way you think.
Looking at the typical analyst firm’s business plan, there are
few ways to generate revenue for the company. One is to do some
research, create a report, and then hope enough people are
interested in the subject of the report to buy copies of it. Or,
better yet, hire the analyst to do more in-depth research for them
on a longer-term basis.
Another way, a more stable revenue-maker, is to supply research
based on a client’s request. The client sponsors the costs of the
research and the creation of the report, and is given a copy of the
report at the end of the proverbial day.
So, if a client sponsors a report, than the analyst will always
have the report conclude the end results the clients wants,
right?
Not if the analyst is honest or wants to stay in business very
long.
Microsoft commissioned Giga to do that study on Windows vs.
Linux development costs a while back, the results of which were
released this month. Clearly, Microsoft liked the results, but did
the very fact that Microsoft sponsored the report automatically
ensure that the results would be favorable to Microsoft?
To reach that conclusion, either one of two things would have
has to happen: Microsoft requested a specific “study” with specific
end results or someone at Giga decided to please their client and
give them the results they thought Microsoft would want to
hear.
Let’s tackle the first scenario. If Microsoft did do such a
thing, and Giga submitted to the request, then both companies are
foolish. By not commissioning an honest study, Microsoft would be
blinding itself to the realities of the marketplace and, in this
case, of honest development practices. By accepting such a request,
Giga–if discovered–would be forever branded as a data cooker and
would never be treated with any amount of respect (or revenue)
again.
If Giga voluntarily provided soothing conclusions to Microsoft,
then Giga is more the villain by sending fraudulent data to its
client, and Microsoft is more the fool for listening to such
information.
With billions of dollars of sales at stake, it would be
surprising to me if Microsoft were deliberately cooking the data or
surrounding itself with sycophant “yes-men.” Yes-men may soothe the
ego, but ultimately they will be wrong and lead you down the garden
path to destruction.
So, would Giga’s report immediately be dishonest because of
Microsoft sponsorship? With so many negatives for such a scenario,
I think this answer is no. I think most honest analysts will
deliver what they believe to be honest information to a client. If,
for instance, Giga’s report had gone the other way and they found
Linux to be cheaper to develop than Windows, that information would
still be useful to Microsoft, and would give them a goal to shoot
for.
The only difference would be that had that specific report gone
the other way, no one outside of Microsoft would ever see it.
I know honest reports are delivered to Microsoft (and other
companies) all the time. Linux has too many advantages for
proprietary software companies not to receive bad news about one
thing or another. After all, if all the analysts’ reports were rosy
and cheery for Microsoft, why would the company ever take action
against open source?
No, I think company-sponsored analyst reports are more honest
than we would believe. But the release of reports is dictated not
by the analyst firm, but by the company. Which is why the media and
the rest of the community very rarely see reports negative about
the report’s sponsor released to the general public.
In many cases, when a company hires an analyst to create a
report, the final dispensation of whatever information revealed in
that report belongs solely to the hiring client. Which means that
report and its conclusions belongs to the report sponsor and no one
will ever see it… unless the sponsor decides to release the
information.
In other words, Giga’s report was bought and paid for by
Microsoft, but not the conclusion of that report.
Which is clearly what happened in the case of this report.
Microsoft liked what it saw, and they authorized Giga to release
the information. Without that authorization, Giga would never have
revealed the outcome of that report. To do so would be legal
suicide. If a report is created that Microsoft does not feel is
favorable to them, then they do not release the information.
We know such reports exist. Specifically, some of Eric Raymond’s
Halloween documents recently revealed that analysts were reporting
that Microsoft’s continued attacks on Linux and open source
software were backfiring. Sure enough, the company changed its
tactics. But we were never supposed to know about those reports. If
it were a perfect world for the kids at Redmond, their tactics
would have just changed and we never would have known why. But the
information was leaked (intentionally or otherwise), and now there
is a kinder and gentler Microsoft.
This is my reasoning for maintaining that corporate-sponsored
analyst reports are not inherently dishonest. I am concerned that
the media and the community is far too eager to dismiss such
reports because of the implied fakery that is believed to be in
such reports. There may indeed be some grains of truth in any
analyst’s reports, no matter who sponsors them. Would we be as
dismissive of a report sponsored by Red Hat stipulating that Open
Source Rulez? The media would, but the community might not.
My point here is to try to focus arguments on the content of
such reports, and not so much their origins. Crying fake just
because we don’t like a report’s sponsor may be a tactical mistake.
The origins of an analyst’s report should be noted but it should
not be the crux of any argument against that report. The
methodology of the report, the cited examples, the
data-gathering… these are the areas that should be focused upon,
if we disagree with the conclusions.
I am not, nor do I intend to be, an apologist for analysts. But
knee-jerk cries of foul will not help win any case against their
conclusions. The focus should not be on analysts’
honesty–something which would be hard to prove, regardless–but on
the accuracy of their claims.
Because even if analysts are being honest with what they
conclude is the right answer, that in no way means they are
right.