"...Someone Had Blundered"
Half a league, half a league,
Half a League onward,
All in the valley of Death
Rode the six hundred.
The JD Edwards User Conference gives
us a better view of one of the two likely scenarios:
PeopleSoft acquires JD Edwards.
Last week's JD Edwards user conference gave us a lot more detail
about the proposed acquisition. Here are the basics:
- PeopleSoft is Acquiring JD Edwards. The JD Edwards senior
management will stay (for up to 2 years), and one member of the
JD Edwards board will sit on the PeopleSoft board, but Craig Conway
and his team will run things.
- One Company. The basic strategy is to combine the companies.
Services, support, and infrastructure will be combined as soon
as possible and as much as practicable. JD Edwards will not be
left as an independent subsidiary.
- A Friendly Agreement. Discussions have been going on
since November, and the merger teams on both sides have learned
a lot about each other. It was a surprisingly well-kept secret,
it turns out, partly because the teams on both sides were small.
Only one of the five PSFT general managers was in the know, according
to reliable sources.
- Three Product Lines. The PeopleSoft product will be the
basis of an "enterprise" product line. JD Edwards' One World will
be the basis of a mid-market product line. And JD Edwards' World
product will be the basis of an AS/400 (i-series) product line.
For all three product lines, the strategy is to create a product
that combines the best of both (PeopleSoft 8.8 and JD Edwards
5). It has not been decided whether that will be achieved through
integration or by rewriting JDE in PeopleTools. It may be decided
on a case-by-case basis.
- Much Planning Left. As I write, teams from PeopleSoft
and JD Edwards are meeting to decide how the merger will go. According
to all accounts, this is a merger done by committee, teams from
each side in each area getting together to decide what to do.
- Cost Savings Unknown. At the conference, $80 million
of cost savings was promised; this would be savings on infrastructure
for the combined companies, e.g., having one data center and one
office in countries like Switzerland, where there are two offices,
"a few blocks away from each other." The new $150 million to $200
million numbers will presumably come from combining workforces.
Everybody at the conference, including me, thought that the managements'
stated goal, having three product lines tuned to the needs of the customers,
each containing the best of all three products, was a great thing.
Most JD Edwards customers were hopeful, but skeptical--and reserving
judgement when it came to committing dollars. I was even more doubtful.
The Three Product Line Strategy
Let's look at the idea that the company will maintain three
product lines: an enterprise product line, a mid-market product line, and an
i-series product line.
On the surface, the lines of demarcation
make very little sense. PeopleSoft's product
already serves both the mid-market and the enterprise market. So do JD
Edwards' One World and World. JD Edwards decision to call itself
a mid-market company was largely a marketing decision; with the same product,
before Bob Dutkowsky came in, Edwards used to go after the large enterprise
market with moderate success. Even today, this "mid-market" company includes
Wal*Mart, Shell, and Fidelity among its customers, and both Wal*Mart and Fidelity
are new customers, announced since the mid-market strategy was put in place.
The lines of demarcation that do make more sense are along the boundaries
of industries.
PeopleSoft serves far more services companies than JD Edwards,
and JD Edwards serves far more industrial products companies than PeopleSoft.
Even industry lines are blurry in places, and there
is overlap in some sectors (e.g., government and CPG). I've thought
about this for some time, but I simply don't see a simple, but
accurate description of where each product line's "natural" market lies.
Alas, the teams now working on the merger
need such a description. "Do we give you this; do you give
us that?" are questions that need to be answered right away.
Unfortunately, I did not see any sign that the guidance
was any more clear or detailed than what was published.
(And I did talk to several people who were involved.)
I also fear that
even a better demarcation wouldn't help.
I'm not sure there is a "right" way of maintaining three living, growing ERP product
lines, each based on a different technology. However
proud, brave, intense, determined, and indignant the management, and however
willing the troops, the plain fact is that each code base has to be sold,
serviced, maintained, and developed separately, which means separate organizations and
separate skill sets. People think this merger might be like HP and Compaq, but
for my money, it's a lot more like Mercedes and Chrysler.
Do you remember that merger? The leadership
promised to exploit available synergies, keep the brands, but
provide the best of both to the customers, and end up with
a single company that was stronger than both. They were utterly
defeated (relative to their promises) by the fact that the
infrastructures just wouldn't combine. It wasn't only that you
couldn't make Mercedes cars on Chrysler lines. The engineering,
purchasing, and sales departments proved equally difficult to join.
I think it's just about as hard with
software companies. Not quite, perhaps.
With two software companies, you can, of course, combine processes, metrics, and
(to a small extent) management. You can combine offices and data centers.
You can combine some back office operaitons. You can combine first line of support and low-level sales (the people who take and route
the phone calls.
But all this doesn't really add up to that much.
In most well-run organizations, the first line of support is
already pretty busy most of the day. The utilization holes--where the
bench-sitters really take money from the bottom line--is in the second line of
support, where consolidation is much harder.
You need to maintain a certain amount of specialization, even when
there aren't that many calls for it.
Is a merger that eliminates a few areas of duplication
good enough? Only if the merger doesn't also involve extra costs.
A History Lesson
Is there an example in the software industry of
a company that tried to have a combined organization, but run
two different code bases? One leaps to mind: it's
called JD Edwards.
Seven years ago, JD Edwards developed an open systems
platform called One World as the successor to its AS/400 World
software. The original plan was to have everyone migrate to OneWorld,
since OW ran on any hardware platform, including the AS/400. In the
meantime, World customers would be allowed to coexist with One World (which
means that they could run World and OneWorld simultaneously). In
addition,
enhancements in OneWorld would also be put into World.
The idea was thus to have a single set of intellectual property
implemented in two code bases, which were interoperable.
We all know that it didn't work; in fact, it nearly drove the company
to ruin. The first problem: One World didn't work and didn't work
and didn't work. The second problem; coexistence was impractical; the product
lines soon diverged. The third problem; AS/400 (now i-series) users
didn't want to move.
I should add, of course, that this was among the most successful of the
various attempts to move code bases from the AS/400. With JBA or SSA,
the companies did disappear entirely.
Why did this happen? It is arguable that it was a failure of people,
not a failure in principle. According to ex-JD Edwards executives, the
original design for One World was excessively complex. Then it was complicated
by having to live up to reckless promises made by the CEO. There were some
failures caused by inexperience with the system. There were failures
due to bad judgement about when to go to market. It all added up to
a failure, one might say, of the system. Producing a competent, workable One World
was simply beyond the capacity of the organization that undertook the project.
Of course there were reasons why it was so much harder than
anyone thought it would be. Myriad reasons. It turned out,
for instance, that you couldn't just use the RPG code for One World. You had
to build new code tuned to the capabilities of the new platform. And this was harder
than it looked. Do you want all the horrible nitty-gritty about why? It turns out, to take
one example,
that a "commit" (the act of storing a segment of data) is pretty automatic on the
AS/400. But in an open systems environment, you have to designate the segments and make
decisions about how they'll be handled. A very irritated JD Edwards executive told
me some years ago that there were 37 major flaws like this.
And Three Code Bases?
If PeopleSoft tries to move the best of JD Edwards to its code base and
the best of PeopleSoft even to One World, I think it will run into precisely the same
kinds of
problems.
Clearly, they won't be able to just write code that
works in all three code bases. They will have to to take an idea
and then write three sets of code, each of which works
works on its own code base. But it isn't just writing code three times.
Each product is different, and each customer base is different, so
a product idea needs to be adapted to its product and customer base.
So it's not just writing code three times. It's creating three slightly
different designs.
There are smart people at PeopleSoft; there are smart people at JD Edwards.
Perhaps these problems can be overcome. But how long will it take? And is using
the R&D budget to work through what ideas ought
to go into each product really the best thing for the PeopleSoft, One World, or World
customer? Sometimes yes. But a lot of times no.
The Justification
Handed out at the conference were 12,000
3 x 5 glossy cards with the "PeopleSoft-
JD Edwards Merger Rationale." Here are
the headlines.
- Global Scale -- $2.8B Company with 11,000
Customers
- Best of Large Enterprise & Mid-Market Solutions
- Solutions and Expertise for YOUR Industry
- Extended Products (sic) Lines for YOUR Business
- "1+1 = 3" -- Two Customer-Focused Cultures
The card concludes with "Have questions? Send them to
Welisten@jdedwards.com".
All this makes a lot of sense if you believe
that the two companies will make one company. But if you
believe that it will really be three companies, each
focusing on a code base more or less moribund, it simply doesn't.
This new, larger company will simply consist of the old,
smaller companies, each of which is distracted by the
fact that there is one management that is trying to
take advantage of synergies.
If I were a JD Edwards customer, I would react to
this card in exactly the way the customers I talked to did react.
Hope it works. Don't commit any money to it. And plan
on an exit strategy if it doesn't.
At the conference, I myself was looking for a little bit
more than a 3x5 card. I thought that after 8 months, there
would be a pretty clear notion about how the not unobvious problems
I'm talking about would be addressed. But executives stayed consistently
on message, which means consistently refusing to address the question
of why their merger would be different. Is this just good public
relations? Or do they really not have the grasp of the actual that
this merger will require?
The Conference
What happens to the new PeopleSoft if PeopleSoft
bangs its head against the wall for a few years trying
to make itself into one company.
In a sense, this conference gave us a view.
At this writing, roughly 60% of the JD Edwards customer
base consists of World (AS/400) customers. These people
pay maintenance; if they grow, they also pay license fees.
(JD Edwards instituted an aggressive audit program a few
years ago.) But they don't get a lot for this money.
To get the latest functionality, World customers have
been told, they will have to migrate to One World. Some
are experimenting with this, gingerly. Others are
planning on living with what they have.
There is some development in World, but it is
limited by a rule that JD Edwards will not change the
data model. The developments that are provided are handled in a
fairly shabby way. According to one person who spoke
at the executive Q&A, the documentation for the enhancements
that are provided is shockingly inadequate.
I talked to a lot of World customers, and this
wasn't the only instance of shabby behavior. I don't want
to repeat them; you can
always hear such stories at any user conference, if you
look for them. But overall, I was surprised at
some of the things that happened.
The One World customers I spoke to were a little more
cheerful. For them, the conference provided
quite a bit of good news, relatively speaking. JD
Edwards has made a major push to get more industry focus in
its product. It has rebuilt its marketing organization, identifying
37 sectors that it wants to focus on. (I can't tell you what they
are, because JD Edwards will not provide me with the information, and
yes, I find this as odd as you do.)
With the industry sectors where they are making the biggest push,
there are Special Interest Groups, groups of customers who tell JD Edwards what is really
needed in the area. In some cases, JD Edwards has also provides funding
to colleges with programs in the sectors where they want to make an impact.
I went to several meetings where recently released enhancements and
the road map in what is now called ERP 9 were discussed. On the one hand,
the sessions were quite encouraging. Many shortcomings of the old product are at
long last being fixed. Many of the enhancements are exactly what is needed
in these sectors. An organization built from scratch in the last two years
is indeed making its mark.
On the other hand, what was announced and what was promised showed
how very far JD Edwards still has to go. The road maps that I saw
were a hodge-podge of really good things and other things not so good.
A lot of the reason for this seemed to be the relative inexperience of
some of the marketing managers. To be effective,
these people really need to know their industry and the competition
inside and out. Unfortunately, with some of the marketing
managers, who are hired "as much for their ability to articulate
as anything else," the depth didn't seem to be there.
In at least one case, for instance, they were announcing
enhancements that had been introduced by integrated suite ERP vendors as much
as ten years ago. I asked the head of industry/product marketing whether
it made a lot of sense to put this stuff in his product, and he said that what
he offered
customers were these same enhancements in an integrated suite.
When you've been around for a while, you can learn from
what happened to the enhancements that have been in other vendors' products
for a long time. You learn what sounds like a good idea, but isn't something
people will pay for. You learn what is harder to do than it seems. You learn
what will take several tries before you can get it right. (One thing announced,
for instance, was "enhancements" to "advanced pricing.") Some of
the people I met were old hands. But others were not.
I'm not faulting these people. They're working hard, and they're
accomplishing something. What I am saying is that JD Edwards' experiences
over the last five years have left it struggling for resources and
struggling to catch up.
It isn't just in marketing. I talked at length to the head of Americas
consulting, who has been there for a long time and has a really
stupendous overall utilization. Very good indeed. But he is only
now beginning to introduce programs to increase participation by
JD Edwards personnel in all new product sales, programs that were
introduced by the competition as much as 5 years ago.
Perhaps I'm going into too much detail. But I want to show in a
fair and accurate way how much damage has been done to this company
and how far it has to go. Throughout the conference, there was much
that was encouraging, as I've said. But I myself experienced some
of the ineptness that I heard about in the stories that World customers
told me. Bob Dutkowsky and Les Wyatt have stanched the wounds and
done as much as they could to put the company on its feet. But it
remains a property that may not be able to achieve the really, really
difficult task that is in front of it.
And If It Doesn't Work?
So what happens if it proves to be too expensive or too difficult
to maintain the two JD Edwards code bases and add in the PeopleSoft
functionality, the PeopleSoft marketing, and the PeopleSoft panache?
The answer is obvious: the JD Edwards code bases will be retired and what
there is of the intellectual property will be put into PeopleSoft. A combination
of support price rises and promises of de-support will force JD Edwards customers
off their beloved i-series, and PeopleSoft will do what it can to provide
migration programs and services.
Essentially, JD Edwards customers will be given exactly what PeopleSoft
customers are now being offered by Oracle: a chance to upgrade to a completely
foreign system when the
time comes.
When that time comes, it will be hard for these customers to believe
that this wasn't the plan all along. Certainly, there is a contingent
of intelligent people who believe that now. They say that Bob Dutkowsky
was hired to sell the company, that he will get an enormous pot
of money if the merger goes through (I have no evidence to support
this and do not believe it), and that the reason for X, Y, and Z
problems (what I've been calling ineptness) is something more sinister:
that he never wanted to invest in making the company grow.
If PeopleSoft has to retire the JD Edwards products, it won't just
be a disgruntled minority that believes this, no matter how hard
the PeopleSoft public relations team works. And that, too, will
be a problem for the new PeopleSoft.
If this merger doesn't work the way it's planned, is the core PeopleSoft
corporation vulnerable? Well, it will have invested years and hundreds
of millions in making this work. It will have paid net $1.2 billion
for a company whose valuation is largely based on people's belief
that it can still grow, but found that the company can't grow after
all. (By contrast, Baan, a company with one-third JD Edwards' revenues,
but a company that people believe can't grow, went for $135 million.)
It will be damaged.
No matter how good its product is (I think it's very good) and
no matter how well run it is (again, quite good), it will have lost
ground to its competitors (SAP's innovation is already outdistancing
PeopleSoft's), and it will have lost a lot of money. If it looks
very much as JD Edwards looks today, it will not be a pretty sight.
Final Notes
A merger is an exciting thing to do, and a merger
in the face of a challenge from Larry Ellison is twice
as exciting. On the JD Edwards side, the people who
are charged with making this merger work are energized,
even exhilarated. "I'm thrilled," were the words of
arguably the most insightful person at JD Edwards.
The fact that many people at both organizations want
to make it work is a real plus.
If courage, energy, and hard work are enough, then
maybe they will be able to achieve the outcome that I
am convinced would be the best possible for the customers
and shareholders of both JD Edwards and PeopleSoft.
But I fear that these courageous and energetic people are being
sent off to face the cannons. It will be a heroic struggle for them.
But in the end, do they want to be memorialized as Tennyson memorialized
the Six Hundred:
Someone had blunder'd:
Theirs not to make reply,
Theirs not to reason why,
Theirs but to do and die:
Into the valley of Death,
Rode the six hundred.
To see other recent Short
Takes, click for a listing.
|
