[long] Picking A Target Market (was Re: A cohesive sales message)

Rob Napier wrote:
> But it soaked up a
> few hours this week that could have been more usefully devoted to working on
> the task that I had undertaken to do.

Welcome to the world of unmoderated mailing lists!

A number of options to help reduce the time we have to spend reading the
list:

1) We could start a new list (eg: pgsql-marketing) which is moderated.
However, you have to find someone to moderate the list.

2) People could learn to change subject lines and include stuff like
"[OT]" or "[offtopic]" to indicate that a message is offtopic for the
thread or the list.

3) People could learn proper list ettiquette. ie: put answers right
after a question, don't quote portions of messages that aren't relevant,
don't quote the whole thread of a discussion, only the message you're
replying to, etc. It's really annoying me that about 30% of the posts
to this list involve someone posting a "me too!" response to a two-page
message, where they quote the entire two-page message and stick their
one-line contribution at the bottom!

Then life would be easier for people who are trying to follow one
particular thread. That's not just you, but at least me, and I suspect
80% of the list readers. Once a thread becomes "old" or "stale" (in my
judgement), I don't bother reading followups. Saves a lot of time, but
then some donkey goes and replies on-topic in an off-topic branch... D'Oh!

> ... you have devoted so
> much time to debating various topics this week while there has been
> virtually no response to my specific questions. Does this mean that no one
> has the answers? I expected to be inundated with information and ideas.

Lessons to learn from Usenet: "Usenet is like a herd of performing
elephants with diarrhea -- massive, difficult to redirect,
awe-inspiring, entertaining, and a source of mind- boggling amounts of
excrement when you least expect it." Gene Spafford, 1992.

The trick is learning that the crap will come, it will be in high
volume, but somewhere in there will be something of value.

> I see no point in trying to second guess the target market, product profile,
> features, advantages, benefits, etc.

At present, the target market for PostgreSQL could probably be best
described as:
1) People who have chosen to use PostgreSQL for various reasons (these
people are the targets for "further education" such as books or courses
or certification)

2) People who are using that other database. These people would need to
be targetted based on performance and ease-of-use. They aren't going to
switch unless we can prove that switching to PostgreSQL is less
inconvenient than sticking with that other database (I believe the hook
term is "Cognitive Dissonance")

3) People who are looking to set up a dynamic website and don't even
know what a database is, much less that they need one in order to avoid
reinventing the wheel (and making all the old mistakes)

4) Conference attendees - but there are so many different conferences,
you'd really have to tailor the message to the conference. A standard
PostgreSQL CD would need to have documentation and sample databases
(along with some CGIs to interact with the database) so the potential
user could see how PostgreSQL could work for them.

(4) is where we get into all that marketspeek such as "branding" and
"awareness" - getting the Label out there everywhere so that by the time
the word "database" has flitted across some poor victim's brain, the
association has already been made with "PostgreSQL" or at least "Elephant".

Josh Berkus pointed out that we'd need to have two distinct sets of
material:
A) the geek-oriented buzzword-free version that addresses technical
issues (eg: foreign keys, outer joins) with examples that illustrate how
the technology works and how it saves you heaps of time (and keeps your
data clean).
B) the suit-oriented buzzword-compliant motherhood statements
accompanied with check-box items comparing PostgreSQL to other DBMSes
such as MySQL and Oracle.

Robert Bernier's CD may be of interest as part of a CD-kiosk for a
conference. Even if some suit just gets a copy of "one of each" from
whatever kiosk is set up, PostgreSQL will end up in some techies hands
eventually ("hey guys, I just got back from my all-expenses paid holi...
business trip, and here's a grab bag of stuff you might be interested in").

> Then I could spend time writing marketing
> material and submit it for consideration and receive hundreds of emails
> (perhaps thousands if this last week is anything to go by) telling me where
> I missed the mark.

What you could do is ask the list if someone is willing to collect and
moderate the responses on your behalf, and post summaries. Perhaps
someone out there is willing to help the marketing effort to this extent.

> That is not how things are done in marketing: The agency gets a brief.

But *that* is not how things are done on Internet mailing lists - in
most cases, people subscribe to mailing lists to share angst, not to get
things done ;)

If we're going to run a "Switch" campaign, we need to have testimonials
(and not manufactured ones like the Ellen Feiss one, either ;), feature
comparisons and a whole lot of effort into "Migration HOWTOs".

If we're going to run a "further education" campaign, we need to have
references such as:
- Mr. Momjian's PostgreSQL book,
- decent database design books (one of which I'm aiming to
obtain for myself and review by the end of this month),
- some books on user-interface design.
- courses in database design
- courses in PostgreSQL management/maintenance
- certification (for user, DBA, sysadmin)

To support a "you need a DB" campaign, there would have to be some way
of triggering the association of "manipulating data" -> "DBMS" ->
"PostgreSQL".

A "branding" campaign would raise market awareness if nothing else.

regards
Alex Satrapa