Now Oracle wants to buy PSFT. Does this make any more sense
than the Net PC? If it does--and every so often poker players do draw the
straight card they need--it will be a more boring world.
Oracle Bids for PeopleSoft
Why on earth is a company whose core
product is of no earthly use to Oracle
worth two times revenues? Here are
(The Conway Theory). The bid is purely
destructive. It is meant to slow down
PSFT's sales (who would buy a product
about to be de-supported?) and derail
the PSFT JDEC merger.
(The Silly Theory). Oracle feels that
it needs to be bigger in the applications
market in order to compete effectively with
SAP, so it is buying market share and customers.
(The Eviscerate PeopleSoft Theory). Oracle wants to buy PeopleSoft's
$800 million in maintenance revenue, the
$2.2 billion in cash, the $300 million in
real estate, the intellectual property (? million)
and the smart programmers and salespeople (if any would
go)--and simultaneously reduce the size of the
(The Visionary Theory). Oracle is willing to take
a very large short-term hit (in cash, management time,
and lost revenues) in order to position itself as an
enterprise platform player. In this theory, there
will be three players serving the enterprise in 5 years:
Oracle, IBM, and Microsoft. (No, I didn't forget SAP.)
I don't know which it is; maybe
it's all of them. But let's look at each in turn. Maybe
one makes sense.
The Conway Theory
Craig Conway has already denounced the offer
as a way of derailing the threatened merger with JD
Edwards. On the surface, this theory makes a lot of sense.
Consider the following:
- The offering price is low; already, the street price
of PSFT has gone well above the offering price.
- There is a poison pill and no real reason for it to go
- If Larry (and I) are right and the JD Edwards merger makes
no sense, why actually acquire PSFT? Let them fry in their own oil.
- Oracle, as noted above, has no real use for a code
base that overlaps theirs closely.
This theory may be right; the low price is certainly not an indication
that they actually want the deal to work. But there
are three considerations that militate against it.
- The price may be right.
If Larry (and Chuck) are half as smart as I think they are,
they are counting on this nuking PeopleSoft's second quarter revenues.
pre-announces," as one analyst said, "and the stock drops to
12, then the 16 price doesn't look so bad."
- A purely destructive bid helps people on the sidelines even more than it
helps Oracle. Why should Larry do something as risky and potentially
self-destructive as pretending to bid for PeopleSoft when the
primary beneficiary is SAP.
- This offer has been in preparation for some time. You don't
get financing from CSFB in a day or two. The timing may be
partially dictated by the merger announcement, but the notion
has to precede it.
The Silly Theory
If Larry is serious, the silly theory has it that he is
buying market share (or perhaps, as one very
silly analyst said, the satisfaction of
still being number two.) The plan under this theory is to keep
PeopleSoft as a subsidiary of Oracle, get some leverage from...
somewhere, and sell the applications "side by side." Lots of people
subscribe to this theory.
As august an organ as the WSJ, for instance,
says the acquisition "could help Oracle expand
and improve its applications business."
Larry's already said he's not going to do this. His plan,
he says, is to migrate people off PeopleSoft 7 and PeopleSoft 8
to Oracle applications--"write some scripts" he says.
Why doesn't this theory make any sense? It's the same reason JDE and
PeopleSoft doesn't make much sense. Whatever is seen as
the secondary code base is going to see a catastrophic fall-off,
because customers simply won't believe that it will continue to have
a useful life. We all know that software mergers just don't work,
even if you want them to. Why should this one work, when there
is no strong reason to keep PeopleSoft alive?
Well, people say, surely there are synergies somewhere
in the organizations. Support can support either code base. (Larry
said he would "combine support organizations.") Sales
can sell either code base. Developers can develop for either
code base. Alas, all these ideas are also silly.
The cost structure at PSFT is there for a reason; most of the
obvious "synergies" will simply result in less customer service,
fewer sales, and less revenue.
The Eviscerate PeopleSoft Theory
Larry isn't quite saying this is what he wants to do, but it's the
only thing that makes sense in the short term.
The economics don't look too bad on this. PeopleSoft has
$800 million a year in maintenance revenue. It has $290
million in real estate. (Thanks to one of the advisory board
members for this.) If you can buy the company for
net $3.9 billion and can net, say, $400 million in
maintenance, it's not such a bad deal.
It isn't quite that simple, of course. There are the
costs of unwinding PSFT--$500 million, maybe--and the
plain fact that customers will start going off PSFT's
very expensive maintenance contracts as soon as they
see that PSFT is going away. So the value of this acquisition
goes down every year.
Set against this are a number of positives. You might, as
Larry says, offer people a migration from PSFT to Oracle.
Might is the operative word, though. It would be very, very
expensive to do plausibly; it isn't just a matter
of writing scripts. The underlying
structure of the PeopleSoft software is quite different from
Oracle's; even though they do the same thing, they do it
in different ways. And don't forget that those PeopleSoft
customers are not likely to embrace an Oracle that took
their beloved PeopleSoft away from them.
Still, there are other benefits. You will still be able
to get some new license revenue as PeopleSoft customers
continue with plans to buy additional software. You will
also get a chance to upgrade Oracle software using PeopleSoft
know-how. It would be expensive and require an almost
super-human tolerance from the programmers, but you could do it.
You can also give a boost to a couple of your other businesses.
The Oracle database business won't be hurt if PeopleSoft stops
selling its DB2 version. The Oracle hosting business won't be
hurt by having some new customers. And there might be
a few synergies. Both Oracle and PSFT
understand how to make some money from consulting; combining
these organizations won't hurt. Maybe some of the empty office
space across the Bay can be filled by PeopleSoft people.
Highly risky, I admit. Not silly. But scarcely worthwhile unless there's
a long-range plan.
The Visionary Theory
I think Larry is one of the smartest people in the
software business. And one of the reasons he's smart
is that he can see the 5-year trend.
So how could this problematic, expensive, destructive
offer make sense in five years?
Imagine, if you will, that a new business emerges: providing
the enterprise stack to companies. The enterprise stack: operating
system, desktop, e-mail, database, business applications, you name it.
Who plays in that business? IBM, Oracle, and Microsoft.
What about SAP? Well, IBM is the biggest enterprise stack
player around, except in the area of business applications. And it
has a huge consulting business that will be shut out of the
Oracle and Microsoft consulting markets. Do SAP and IBM
want to sit there arguing over who gets the consulting business
and the middleware/app server business? I've always said that
IBM would never acquire a big business applications company. But
this might be the exception.
If that's the picture in 5 years, it makes all kinds of sense
for Oracle to pick up PeopleSoft (and JDE for that matter, though
it's a little more problematic).
Here's the problem with this idea, though. This isn't a five-year picture; it's
a ten-year picture or more. It assumes, too, that there will
be no revolutionary innovation in the technology. It even assumes
that there isn't much innovation left at all. (The natural path
for innovative software companies would be to be purchased by
one of the Big 3; indeed, it might make a lot of sense for all
three to open up their application stacks so that developers could
innovate for them.)
In the meantime, you've spent $3.9 billion and a lot of management
time on what will prove to be a drag on you and your attention.
Conclusion: Winners and Losers
So which theory is right? Well, notice that there's nothing
inconsistent about any of them. Larry (and Chuck) can be destructive
in the short term, pick apart PeopleSoft at their leisure, keeping
the bones with meat on them, and gradually build up the enterprise
stack. All it takes is money, near-perfect execution, luck, and
better vision than I have. They may have all four.
In the meantime, who wins? Well, it's not bad for Oracle, though
it will take a lot to make it work. But SAP wins big, big, big.
In the short term, you know who else wins big? Lawson ... QAD ...
Intentia ... MAPICS ... SSA Global Technologies ... Navision ...
Damgaard ... Great Plains.
Suddenly, doubt is cast on their biggest
competition. Who knows, it might even help out i2.
The losers? PeopleSoft. JD Edwards. IBM, for sure.
And perhaps all the rest of us, who love the applications
market. Whatever else happens, this is almost sure to stifle
innovation. Ultimately, an enterprise stack limits what you
can do with the software to whatever is allowed by the design.
A lot of good ideas will be inconsistent with what the big
boys are doing, so they'll never even get off the ground. And
a lot of other good ideas will never get the funding they need.
The current highly competitive, arrogant, freewheeling market
has produced an unbelievable amount of bad behavior that has
been nothing but bad for customers. But like a lot of adolescence,
it can be way more fun than maturity.
To see other recent Short
Takes, click for a listing.