Good points, Ed.
It makes sense that vendors would go with the path of least effort,
especially when it seems to their advantage to measure in that way.
Thus, the much more difficult task of proving real marketshare falls
to this project! With IBM on our side, as well as Sun, we might be
able to figure this out, or at least promote our interpretation a
little more.
Firefox also benefits in that websites keep stats of which browsers
are visiting them. In a slightly similar way, we can search for the
number of ODF vs. MSOOXML documents available on the web (and some
have done that and written about it), though this is still an
inaccurate proxy measurement.
At the very least, we can strongly promote our meme, which may cause
users to demand a reinterpretation of the landscape. Breaking down
the 95% assumption will benefit us in many ways.
-Ben
On Oct 12, 2007, at 2:44 PM, Edward Buck wrote:
Kaj Kandler wrote:
Benjamin,
good observation. What are the other marketing departments of OSS
saying
about this? Mozilla, Apache, Ubuntu, ...? I think this is a wise
spread
issue if true and we should focus on making all in the OSS community
aware of it. Can we try to collaborate with some economics
professors,
in order to study the general phenomena and then use their
research in
order to blog, give presentations at conferences, unconferences,
post in
mailing lists, etc.
Having worked with a lot of market research firms in the past (an
industry filled with conflicts of interest and poor methodologies),
there's a very simple reason why they like to use revenue numbers:
it's
much easier to compile. They simply call up a company and say, 'we're
doing a market share study and would like to know your revenue numbers
for x product.' Then, they tally up the numbers and issue a
report. In
other words, there's very little primary research involved. A REAL
market share study would have to look at actual usage but that means
those doing research would have to do some real work and solve real
problems.
The revenue == market share phenomenon exists in many markets,
especially software. For example, mysql hardly registers in some
market
share studies of database servers because a significant number of
users
get it for free. It also doesn't help that those sponsoring these
reports are often the big commercial vendors themselves, i.e. Oracle,
IBM, Microsoft, etc.
The only reason why Firefox doesn't suffer from this problem is
that the
alternative is free too, e.g. IE.
Ed
I think we should pose the question to the researchers, how they
measure
the market share.
Kaj
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
