RE: Marketing list action: Market Research for GNOME and GNU/Linux

["John Williams" <JWilliams business otago ac nz>]

> I share your point of view. On the other hand, GNOME's 
> potential market
> (or audience) is worldwide (or at least very large), isn't it?

The size of the market is not relevant to
(1) The correct way(s) to do market research
or
(2) The correct way(s) to construct a marketing strategy

> Thus, some questions arise. For example: What would be a 
> representative
> sample size? How do we define the sample? How do we make sure 
> the sample
> is indeed random?
These are easy questions to a market researcher or statistician.  The
answers are:

(1) The question is ill-posed ("representative" and "size" are
unrelated)
(2) The sampling frame is determined by the target market.  Thus,
research cannot proceed unless the marketing strategy has first been
defined.
(3) There are many ways. All are easy.


> But the most important question would be: What do we do with the
> results?
Great!  The number one question that needs to be answered before any
research is commissioned.  And again this points to the need to have a
strategy defined ASAP.
 
> For example, we all know that network economics protects Microsofts
> position. 

> I also like to add that I've seen no concrete marketing goal yet.
Me neither.

> Thus, I like to suggest we start small: Let's skip the "real" market
> research for a few years.
There is no need.  The sort of research that I envisage would cost no
more than USD$1,000 as long as we could get volunteers to lick stamps
and punch data.

> A web-based survey would be 
> sufficient for the
> start. 
Probably far to biased; but depends on target market.

> Nevertheless, even a low-quality survey would be great 
> indeed. I'm sorry
> if my position is a little bit too pragmatical.
Is the wrong information better than no information?