Hi all,

I have been having a debate recently on another forum about database
market penetration and how to measure it.

My point of view would be that even for companies which *_only_*
produce databases (Oracle until recently, MySQL, say now), it is
impossible from their financial filings to determine how many

a) paid for installs there are

and

b) if an install is paid for, how much was paid (VAR, Corporate
resellers, OEM agreements, <insert contractual agreement of your
choice>).

The thrust of my argument has been that, in the absence of any other
reliable metrics, an analysis of job adverts in a particular market
(say country) is *_AS GOOD A WAY AS ANY_* to determine market
penetration (if no absolute revenue or number of (paid for)  installs
is available - normally the case - and even then, they are potentially
misleading).

Can anyone think of or does anyone know of a better metric for
measuring db penetration than job ads? They don't *_always_* mention
the db they are targetting, but *_usually_* there is some mention of
the particular implementation (be it Oracle, SQL Server, Sybase,
Informix, DB2.... whatever).

I would like to know

a) do people out there feel that this is a good (best) way of
measuring market penetration?

b) if not, how would *_you_* measure db market penetration given
currently available data from software companies and your particular
market?

I am not a bigot here, and am willing to listen to any/all ideas that
people might have about how to measure an entity which is very
difficult to get a grip on.

I mean, even if you have Oracle's revenue for (my country, Ireland),
that doesn't breakdown VAR/OEM/Corporate deals, nor will it detail
support (24 hour, 1 day, 5 day - on site, off site - whatever).

For example, I have read that Ryanair use Oracle on HP-UX boxes - more
or less the Cartier of db solutions - but could I obtain any idea of
how much Ryanair are paying for that - no chance. I would *_assume_*
that Ryanair have round-the-clock, engineer living on site type
support but

a) I have no way of proving that

and anyway

b) I have no idea of what Oracle is charging for that service.

So, guys, I'm basically asking - what is the best metric to use for
establishing (even a rough) an idea of db penetration in a given
market. Personally I can't see a better one than job-ads, but am more
than willing to listen to other ideas/concepts that people might have.

TIA.

Paul....