Subject: Re: Larry Ellison on FSBs
From: "Ben Tilly" <>
Date: Tue, 25 Apr 2006 07:49:01 -0700

 Tue, 25 Apr 2006 07:49:01 -0700
On 4/25/06, Keith Bostic <> wrote:
> On Apr 24, 2006, at 7:10 PM, Ben Tilly wrote:
> > Between PostGres and MySql, Oracle's current business model is toast
> > within a decade or so, and there isn't much they can do about that
> > reality.  Given the level of lock-in that people have with databases,
> > they'll take a good while dying.  They may find new businesses to be
> > in.  But their current business is doomed.
> This isn't correct, Oracle's revenues aren't license based, and
> haven't been for some time.
> The last time I heard, 60% of Oracle's revenue was unrelated to
> database licensing (support, tuning, custom software, middleware).
> That was a few years ago, and I'm sure licensing makes up a smaller
> percentage now, not larger.

Accepted.  But how many of these auxiliary business opportunities do
they get by virtue of owning the database in use at the organization? 
When companies use another database, how likely are they to want
Oracle's support, custom software and middleware?  Their core business
is a bigger piece of the pie than licensing alone would suggest.

> Sure, MySQL is going to drive Oracle's licensing revenue down and
> Oracle knows that.  The trick for Oracle is to manage the revenue
> curve as licensing revenues decline over the next decade.  Maximizing
> licensing revenue will cede more of the marketplace to MySQL; keeping
> more of the marketplace away from MySQL will tend to minimize license
> revenue.  (As an example of managing the curve, buying Innobase
> probably shoved the curve out 12 to 18 months, all by itself, a
> pretty good use of a few million dollars.)

Yes, Oracle can delay the transition.  But they cannot stop it. 
Incidentally people with existing Oracle installations are not
noticing any generousity on Oracle's part - Oracle understands
customer lock-in quite well and takes full advantage of it.  Only
people who are considering a database get generous prices, and that
behavior is exactly what  Information Rules  says that Oracle should
do.  (It is also what Oracle has done as far back as I remember.)

> Regardless, Oracle's "current business" is largely the same as
> MySQL's (except for being a couple of orders of magnitude larger),
> and has been for years.

Their businesses are similar, but are both tied to their databases. 
As you replace Oracle installations with MySQL installations, that's a
growth opportunity for MySQL and a problem for Oracle.