Subject: Re: Larry Ellison on FSBs
From: "Ben Tilly" <btilly@gmail.com>
Date: Thu, 27 Apr 2006 00:02:47 -0700

 Thu, 27 Apr 2006 00:02:47 -0700
On 4/25/06, Stephen J. Turnbull <turnbull@sk.tsukuba.ac.jp> wrote:
> >>>>> "Ben" == Ben Tilly <btilly@gmail.com> writes:
>
>     Ben> Furthermore MySQL doesn't put together its improvements at
>     Ben> any expense to itself.
>
> Neither does Oracle or any business that consistently posts ROI > 0.
> Why is it an advantage that the revenues that cover the expenses are
> earmarked for those expenses?  Isn't it rather a disadvantage, that
> MySQL is locked in to those expenses rather than activities that might
> be more strategic and long-term?

You were positing that it is a disadvantage for MySQL that what they
put together at cost to themselves can be  borrowed for free by
Oracle.  I'm pointing out that maintaining their software is not a
cost center for MySQL, so that isn't a disadvantage.  Furthermore
MySQL gets feedback on what exactly it is that people want.  If Oracle
tries to get involved, Oracle has to spend money doing what nobody
asked it to do, and Oracle doesn't get customer feedback to help it
understand this target market.  (What people want changed about MySQL
and what they want changed about Oracle have little relation, so
Oracle's existing experience doesn't translate so well.)

>     Ben> Given their dual licensing model, as long as MySQL can
>     Ben> maintain control of the software base, they are the only
>     Ben> people you can get those improvements from.
>
> In other words, MySQL *owns* *something*.  What does this have to do
> with the open source threat to Oracle?

What MySQL owns is a competitive advantage in maintaining the
maintainership of an open source project.  That open source project is
a future threat to Oracle's installed base.  It is both owned and open
source.

>     Ben> Please explain your acronyms.
>
> XP = extreme programming, AKA refactor until you're down to all
> primes, and then run some more primality tests just for fun.  XB
> (which acronym I just made up) is the same idea: "extreme business."
> No sacred cows, rethink everything every day.

OK, I knew that version of XP.  But I thought that you were thinking
of some version of XP that was related to an acronym XB, and I didn't
draw the right conclusions.

> SAP and IBM are not selling software.  They are selling business
> process consulting, business processes which are supported by their
> software.  Ellison is saying Oracle is doing the same thing.

Fine.  I can tell you from direct experience that Oracle's software is
orders of magnitude more difficult to install and maintain than MySQL.
 Furthermore Christensen's thesis says that Oracle's internal cost
structure should be a lot fatter than MySQL's.  Not with unnecessary
overhead - to go after the big clients you need infrastructure that
costs a lot.  However with overhead that smaller clients don't need
and don't want, which gives MySQL a competitive advantage over Oracle
in a market that is more than large enough to interest MySQL.  The
result is that Oracle can't go head to head against MySQL and
outcompete it unless Oracle is willing to throw a very large amount at
the problem.  (And if they did, the result would be an opening for
another company like MySQL.)

>     Ben> If we're going to talk about Christensen, then we need to use
>     Ben> the words in the way that he does.
>
> OK, I'll have to go look at it again more carefully.
>
>     Ben> Where things get very difficult is when it is obvious that
>     Ben> your long-term interests are best served by doing A, while
>     Ben> your short-term interests are best served by doing B.  In
>     Ben> that situation it is hard to do A, and harder still for
>     Ben> organizations than for individuals.
>
> Precisely why I emphasize that Microsoft and Oracle are still more
> personality cult than organization.

Have you read http://minimsft.blogspot.com/?  Microsoft looks an awful
lot like an organization to me, and one which needs to watch its
cholesterol levels.  Oracle may be in better shape.

> You know, there's a reason why some executives get paid numbers of
> dollars that look like interplanetary distances, to manage companies
> whose accounts look intergalactic.  That reason is that boards can
> convince themselves that these are the guys who have done A before and
> will do A again.  The jury is still out on whether they really do, but
> what if LE has *already* done A with Oracle.  That's his claim.

You know, last I heard there was a NEGATIVE correlation between CEO
compensation and company performance across the Fortune 500.  Which
suggests that CEO compensation is less about the official reasons
given and more about management working to subvert stockholder
interests and boards turning a blind eye.  Many observers, including
Warren Buffet, would agree.  See
http://www.berkshirehathaway.com/letters/2002pdf.pdf and go to page 15
in the PDF (labelled 16 on the page) to see his comments on the
subject.

As for Larry's claim, I'm sure that he sees the need to diversify and
I'm sure that he has done a lot of diversification.  But I strongly
suspect that as Oracle the database falters in the market, Oracle the
company will show the hurt.

>     Ben> By his terms that is absolutely right.  You can't charge a
>     Ben> lot in open source, and you certainly can't charge as much as
>     Ben> Oracle is structured to need.  Unless Oracle can restructure
>     Ben> itself to not need that (organizationally very difficult), it
>     Ben> needs to seek business opportunities elsewhere.  (Of course,
>     Ben> like IBM, those opportunities might be found in complementary
>     Ben> businesses to open source ones. Indeed I suspect that Oracle
>     Ben> is going that way.)
>
> Why "suspect"?  LE said each of those sentences in very similar ways
> in that interview.  The main difference is that you, for reasons I
> don't grok yet, seem to believe that open source is going to drive
> Oracle, whereas LE clearly believes that open source may get to where
> his business is sometime in his career, and then again, maybe it
> won't.  And he certainly isn't going to bet G$6.5 that Red Hat will be
> the company to do it.

I don't believe that open source businesses will get to where his
business did - it is the nature of the game that open source
businesses have thinner profit margins.  However I don't think that
he'll find that his business forever has the kind of generous margins
that it has become accustomed to.  I could be wrong.

>     Ben> Actually the technology is NOT a horse race.  Take 2 machines
>     Ben> that are installed and ask two competent people to install a
>     Ben> database on it. MySQL will be installed faster, and requires
>     Ben> lot less competence to install.
>
>     Ben> By contrast Oracle installations that I've seen are using
>     Ben> features that simply aren't found in MySQL.  (Or weren't
>     Ben> until recently. Though plenty are using features that still
>     Ben> aren't in MySQL.)
>
> You're missing my point.  Oracle can afford to provide *both* a free
> software EZ-DB for the masses---competing against MySQL on that almost
> as good EZ-DB plus whatever snake oil LE claims is their *real* core
> competence (this is the "horse race"), and also their full-featured DB
> where MySQL *can't* compete (yet) for those who need it.

When incumbents try to offer a low-cost alternative they generally
fail.  That's because they don't know that market well enough to know
what needs to be left out and what needs to be left in.  Certainly if
you take "Oracle Lite" (eg their personal edition), it is a miserable
failure in the terms that MySQL users judge MySQL on.  (It is harder
to set up, it is harder to manage, you can't find cheap web hosting
providers that will support it, etc.)

> Sure, this implies one of your central themes AIUI: because that will
> cannibalize Oracle's full-featured DB market, Oracle has to stop
> selling databases and Do Something Else[tm].  My question is "What is
> MySQL gonna do if for Oracle Something Else[tm] is a done deal?"

Christensen would predict that MySQL will make plenty (for it) of
money in market niches that Oracle can't be interested in because
Oracle can't make enough money there to make a profit.

Cheers,
Ben