How Much License is Too Little License?
In the last few days, both PSFT and SAP did some explaining.
PeopleSoft missed badly on license revenue, but kept consulting
strong. SAP fell short on consulting, but not badly. What does this
mean? First PeopleSoft.
PSFT's License Falls
The Q4 earnings call for PSFT was impressive personally for
Craig Conway and Kevin Parker, but the key performance
metric for most analysts fell badly short. "License of $81 million
when they guided $125-135 in January," growled one analyst. "Is
anybody minding the store there?"
Conway's and Parker's explanation for the license revenue figure
was simple and, if accurate, exculpatory. It was a bad economic
climate. People were not willing to make capital investments. Several
deals fell away at the very last minute.
Most analysts I've talked to recognize that this was a bad time
for sales, butsome wonder why PeopleSoft was so very vulnerable,
when the license revenue of other software companies in the space
climate were not affected as much.
A few, I have to say, are going off half-cocked. There are those
who believe that the company has come to the end of the line, or
belive it is is completely out of control, or believe it has front-loaded
the bad news, so they could leap the bar the rest of the year.
Let me just say here that the latter two positions (for which there
is not a shred of evidence) don't make much sense to me. I can't
say I like Craig Conway a whole lot, but he's always shown a good
grasp of what's going on at his company day-to-day. So I don't think
the sales process is out of control. I also think that he has way
too much confidence in himself and his company to do something idiotic
But then how do you explain the shortfall?
Well, here's one man's attempt. I'll use rough numbers.
Let's say they missed guidance by 35 percentage points. Of that, let's say that 10
points are accounted
for by the harsh economic climate. An average miss of 10% for the sector (in license)
is roughly right, and that's all we need. I include in that figure the effect
of currency fluctuations, which may be giving PSFT the benefit of the doubt.
Beyond that, you have to recognize that PSFT is particularly dependent
on areas of the economy that were particularly hard hit: services
and government. First quarter has been a mess for
state governments; they've cut budgets and then cut them again
and again. As Conway pointed out, there was a collateral
cutback in university spending. PSFT has been getting 25% or more
of its revenue from government and education, and in January, when
the guidance was given, the extent of cutback in budgeted spending was
not as visible as it is today. So let's say they lost half
their projected license sales in the area. That's another
12 percentage points.
I'm going to assign another 5 percentage points to a factor that was
not mentioned by Parker or Conway, but one for which there is some evidence.
At the end of a good fourth quarter, like last year's, salespeople in any
to empty out their pipeline. Management knows this, so they try to take
some steps to fill it back up: it's called marketing. In Q1, PSFT
did not do any extraordinary. There was no Leadership Conference,
no new marketing campaign. In retrospect, that may have been
At the time, it was a reasonable mistake. It appeared then
that optimistic budgeting in late '02 and the release
of some new products was going to refill the pipeline
without too any extraordinary effort from PeopleSoft.
On this theory, by the time
the slowdown started to hit, there wasn't that much PSFT could do.
Admittedly, every other software company faced the same seasonality problem,
so I'm theorizing that PSFT made decisions about demand stimulation that were
worse than average. Yes indeed. Maybe Conway was overly confident. Maybe the
company wasn't agile enough. But so what. An isolated misguess about timing in
an extraordinary environment
does not indicate that there are serious flaws with a company. If they keep
on misspending on marketing, that's another issue.
That leaves 7 percentage points, plus or minus. If you want some
worrisome explanations for this last bit, here are two. First, PSFT may really not
have been able to get the upsells on upgrades that they expected. (One
can assume that the people who are quick to upgrade are more likely
to buy more than those who are slow, and PSFT may have mis-estimated
here.) Or, it could be that the sales cycle is changing fundamentally
and PSFT hasn't reacted as effectively as other companies. Conway gave
some support to this, saying in an aside that several deals were being
sent up to the CEO or the board and being stopped there.
But there are any number of explanations that are easier to take.
Seven percentage points isn't that many. It could be that PSFT
was put under a lot of pricing pressure in the last days of the
quarter and refused to discount excessively. (I don't have any anecdotal
evidence, sorry.) This would be a good thing as far as I'm concerned.
Don't Forget the Strengths
I'm not that worried about the last 7% or even any
major drop in license revenue, because I like the overall
The company has $2 billion in cash on a market cap of $4.8 billion.
They're on a pace to take in $800 million ($1 1/2/share) in maintenance revenue
this year. They had a service utilization of 57%, and they have seen
little or no price pressure on consulting rates.
PSFT is arguably best among the ERP vendors (I exclude Siebel
because I think their market is less diverse) in
all these areas: they have the most ability to invest as needed,
the most solid customer base, and the greatest demand for consulting.
These comparative strengths, moreover, are structural.
Take consulting. Because PSFT is
relatively US-centric and because the product footprint is less
industry specific (and arguably smaller), it is easier to staff
their consulting practice than it is for other companies in the sector,
so utilization should be higher. Also, PSFT has been more effective than
any other software company, from what I can tell, in making sure that most
PSFT projects have at least a few PSFT consultants on it, by far the
best way to deploy consulting.
You can make a similar case about PSFT's ability to invest effectively
and about its ability
to get customers upgraded and still paying maintenance.
are structural strengths. There are technical reasons for
these strengths (the architecture is very
strong), historical reasons (customers
have historically liked the company and been forgiving), and competitive
reasons (PSFT has relatively little competition in many important sectors).
These are strengths, therefore, that are almost impervious
to bad management. So even if the analyst
quoted in the first paragraph were right, PSFT would remain strong.
Does License Drive Revenue?
But isn't all this ultimately built on a steady stream of license revenue?
When license revenue falls off, the long-term potential for
service and maintenance drops, too. Right?
I think this has been an article of faith for way too long; it's time
to drop it.
You have to remember that what customers want is not software,
but solutions, solutions that provide clear benefit. When people
pay for license (that is, software),
they are actually putting a down payment on
a chancy capital investment. Before they see a return,
they'll have to spend at least 2-3 times more, use up
a lot of intangible internal resources, wait, in the
case of the initial implementation, 1-2 years for the implementation
and another 1-2 years before benefits start accruing.
In the late '90s, people looked at this fairly lousy
deal and said, "OK." They're smarter now. They want to pay
less up front and they want better guarantees. They don't
want a CD; they want a combination of license, service, training,
and maintenance that will insure that the investment is a good one.
This modern doubt about the value of a software license
isn't so bad for a software company, if it's well
run. The company can simply structure itself so that it
addresses the whole value proposition. In so doing,
it can also get more total cash out of the deal.
Don't you get better margins from a CD? Well,
on the maintenance side
and even on the service side, the margins can be pretty good, again depending
on how well run it is. And if you think of the maintenance and service
revenues as extra cash
that's returned on the same investment in R&D, isn't that a good
When you've deployed yourself to address the full value proposition,
you also gain a lot
of flexibility. If license drops down a little bit, you have
time to structure things so that you get a larger percentage of
the total value of the deal.
As it turns out, companies like PSFT that have moved toward
this model have also gained a lot of unexpected benefits. Customers of
all of these packages have had so much trouble actually realizing
the benefits that should have accrued, that a big market in
add-on consulting and add-on software has emerged that was
never part of the original plan. Companies that sell the full package
(like PSFT) are much better able to take advantage of this new
market--and they have.
If you're selling a complex package of license and services, then,
the fact that some component of the package assumes less importance
for a year doesn't matter all that much. A resilient and well-managed
company just deals with it.
A fact that no one seems to comment on brings this
into perspective. Last quarter, the sales and marketing expense for PSFT
last quarter exceeded its total license revenue by 50%. Are
they really spending a buck and half to bring in every buck of
license? I don't think so.
I have already put out two positive pieces about SAP this month.
I continue to believe that they are a money machine, but one that
isn't run at full capacity.
But I haul them before you again, because I think one thing they could do
to run at full capacity is to take
some lessons from PSFT's book.
This sounds weird, of course--wasn't it SAP that did relatively
well in the quarter? But hear me out.
In another very impressive personal performance last week, Henning Kagermann
took financial analysts in New York through some key financial metrics, then
thumped the podium for SAP's technology. No doubt, SAP reacted well to
the slowdown and relatively speaking, was barely touched.
But there were also some danger signals. Chief among these were the facts and figures
on upgrades. Bill McDermott: we're targeting 1016 customers in the US for upgrades. Henning:
30-35% of the contracts are upgraded (presumably, he means to mySAP.com, or does
he mean something else?). It appears to me that SAP customers are gradually doing
what runners do in a long race, separating into clumps. The fast ones are upgrading or
have upgraded and are ready to buy more goodies. But the majority have fallen back.
The Short Takes advisory board has been providing lots of color on this:
the high cost of upgrading to even to R/3 Enterprise, the unhappiness in
certain key industries with product functionality, occasional attempts to
ratchet maintenance revenue down. None of it is catastrophic; all can be dealt
with. But in the long run, it decreases margin and slows growth.
The second is the uncertainty surrounding SAP's new technology moves. On
the one hand, as I've said before, SAP is the only vendor among the
enterprise application vendors to understand that something will happen
in software support for enterprises after the 4GL/ERP/transaction model. NetWeaver
and the xApps are very good ideas, and they give SAP some chance of living
beyond the stage of vigorous near-maturity that PeopleSoft is in.
But they are slow. NetWeaver has yet to coalesce into a clear product
with a clear go-to-market strategy. SAP seems content so far to sell it
to clients who really need it, largely because they've asked for it, when in fact
they should be rolling up the space. And the xApps are still going through
an extended shakout period, one that is probably necessary, given the ambition
for the products, but still worrisome. It isn't as if the xApps idea has to win;
it could be that it will lose momentum and some other version will catch fire
To describe these products, a lot of tap-dancing seems required. Henning is good at
it--too bad he used to let Hasso do all of that--but it can still come off thin.
Thinnest of all was the entirely rational, careful, and thoughtful presentation
about web services, the best single thing I've seen on the subject, but what
it all seemed to amount to is that SAP won't be a web services company for
the foreseeable future.
The other worry has to be the Americas. Bill McDermott's performance was not up
to any of the others mentioned, either personally or in substance. The Americas did
OK this quarter, and it seems clear that SAP management is affording him some
breathing room. But it doesn't seem so far that he has successfully
dealt with any of the
essential problems with the region. And time marches on.
License and Service
So what does all this mean long term? First of all, it means some bumps in
the road at some point. Bumps getting the recalcitrant customers upgraded
or disposed of. Bumps getting NetWeaver to becme as seamless a revenue producer
as BI. Bumps selling the xApps. Bumps in the Americas (when haven't there been?).
In my view, many of these bumps could be smoothed out if
SAP chose a more flexible approach to managing the customer lifecycle.
As Henning said over and over, SAP is a software company; it focuses
on license revenue. This is appropriate, given its history and talents.
But it doesn't solve the customer's problem.
As noted above, SAP's customers need to have a reliable process
that insures that appropriate value is returned in an appropriate
amount of time from the software purchase. They also need a lot of
help with value realization.
The danger for SAP is that relying overly much
on partners to provide and guarantee this process
leaves the customers more at risk than they should be
or want to be.
This is not a new danger; it's been an issue since SAP started business.
But the complexion of the problem now is different from what it was. The product
is more complicated. Consulting isn't as good a business as it was. Frequent
failures have left customers impatient and risk-averse.
Under this view, the fact that SAP's shortfall came in the services area
is more disturbing than it probably is to Henning. (Especially since
currency was such an important factor in SAP's results this quarter.) I see
it as a sign that the business still hadn't figured out how to manage things
in an area that will become increasingly important. (Again, I don't think
Henning would agree.)
Oddly enough, one test of this hypothesis will be in sales of NetWeaver
and the xApps. In my view, if customers are insisting on solutions, then
they will be reluctant to take on a model where SAP provides the software
or co-develops the software, and they figure out what to do with it. They're
going to want more guidance from SAP.
I don't have enough visibility into this area to see whether I am right, but
I advance it as a way of judging my hypothesis.
In any case, we are talking about fine-tuning, not about whether SAP
can be a blue-chip company for the next 2-3 years. At 25, of course, SAP is
pretty fully valued, so fine-tuning does matter. A finely-tuned company
will justify the valuation, whereas a company that is focused only partly
on the right things, will have more fits and starts than those optimistic
shareholders have a right to expect.
To see other recent Short
Takes, click for a listing.