Re: We have Google Adwords Credit: How do we use it? - Mailing list pgsql-advocacy

Chander Ganesan wrote:
> I think we want to focus on areas where we don't have high organic
> rankings.  Furthermore, we don't want to target current PostgreSQL
> users - since they already know where to go for information...and have
> already made their decision.
>
> The big thing that comes to mind is targeting people who are looking
> for competitive alternatives (Oracle, DB2, Sybase, ESRI) where PG can
> present a compelling, cost effective case.  I'd list MySQL there too,
> but I don't think that the case is as compelling (from a "cost
> effective" standpoint ... despite what people say, IMHO MySQL is an
> alternative/competitor to PG, albeit not a good one).
>
> To make the $$ stretch we might want to look at less competitive
> keywords, since the more competitive ones will either leave us on page
> 2 (or later) or saddle us with high PPC costs ($5 per click doesn't go
> very far when you have a total of $200).
i'm currently running a small ($130/month) adwords campaign
for one of my websites. i'm still learning how the system works
(it's only been in adwords since may 29th), but i have observed
a bit about things and think i can contribute to this discussion in
a useful way.

don't worry too much about expensive keywords. leave the budget
optimizer on, and use relatively specific groups of keywords tied
to specific features of oracle, db2, informix, whatever, in order
to show up on those searches. alternatively, consider specific
search queries where you might want a postgresql ad to show up.
for example, i often google on the generic parts of text of error
messages to find online discussions of them. got any favorite
common/annoying/difficult oracle error messages? make them
keywords.

you can have multiple landing pages, you set up a campaign for
each landing page and run different keywords for different landings.
this gives latitude for customization of what part of the site people
get directed to.

even if you include expensive keywords, the budget optimizer
will only bid on them when the prices are low; they fluctuate
through the day so inactive keywords (turned off temporarily
by the optimizer) aren't that big a deal.

you can look in daily and see what keywords are yielding
click throughs and which are not.

the best thing you might get out of this is access to google
analytics, which is included for paying adwords customers
and might be thrown in with this $200 credit (ask the google
folks, maybe they'll let open source projects have analytics
for free in general.)

analytics will let you see the geographical locations of
visitors to the site, what their landing pages are, how many pages
per visit, what searches are getting them to their landing pages,
all kinds of good stuff. it will help you analyze which pages
on the site deserve the most attention; when someone hits a landing
page you want them to see something that causes them to
look around.

define some conversion goals in analytics -- downloads would
be good, registrations for mailing lists, etc.

that's what immediately comes to mind,
   richard