The Gathering Storm. Microsoft's recent analyst day provides
us with a good excuse to initiate coverage of Microsoft as a serious
enterprise applications vendor.
We can imagine two, diametrically opposed reactions. One. "What took
you so long? Microsoft has announced its intentions. It's committing
huge amounts of capital. It is already a major player". And two. "Yawn.
Microsoft is just Great Plains + Solomon + Navision + Damgaard + wind.
None of those companies have been major players. And Microsoft's saying
it doesn't make it so."
The first position is held by many financial analysts. And the second
is the prevailing view (we believe) among ERP company executives.
We've been sympathetic to the second view, so far. But we're now
beginning to see how Microsoft can exploit some fissures in the market that
the ERP players don't want to admit exist.
We think the subject deserves some attention, in any case, so we're devoting
three Short Takes to it. In this, the first one, we explain why
the ERP executives aren't worried yetand why they should be.
What, Me Worry? To many financial analysts, the mid-market,
where Microsoft plans to play and where all the major ERP players
already play (more or less) is a single, small territory, and they
are already assigning market share to Microsoft in their models.
But to the ERP executives, the mid-market is a vast continent.
If they're all playing in the far Northeast, Microsoft has landed
somewhere down in Mexico. Before we need to worry about them, the
executives think, Microsoft needs to make its own colony viable,
then cross deserts and the broad Mississippi.
Let's say they're right. (They don't have to be; remember,
the now-defunct minicomputer manufacturers weren't too worried about
the PC, said it would be decades before they collided.) And let's
look at the territory they say Microsoft has colonized.
Microsoft Territory. Mid-market ERP software can be segmented along
five dimensions: the spectrum of functionality provided, the complexity
of company supported, the geographies covered, the industries where
special functionality has been provided, and the channel.
Along all of these dimensions, the Microsoft products have taken segments
that have very little overlap with the ERP players.
Functionality/geography/company structure. Great Plains
and Solomon are American accounting packages at heart, designed
primarily for companies with a simple company structure. They have
good sales and purchasing and a little bit of inventory control,
but not much in manufacturing.
With that functionality profile, there are lots of places where
they don't play. Go to Europe, Great Plains or Solomon don't work.
Make or distribute stuff in serious amounts, you need good manufacturing,
planning, and inventory control; again, neither Great Plains nor
Solomon are good candidates.
Navision and Damgaard do work in Europe, and they have better
inventory control and good interfaces. But they don't
necessarily support American financials or sales processes.
All good products, no question, but in America, GP/S are best suited
to small, non-manufacturing businesses, branch sales offices, or
some service businesses, and severely limited when a user makes
things, distributes things, or has a complex business structure.
In Europe, with N/D, there is a similar level of limitation.
Industry. Given the functionality, it has historically made
sense for Solomon and Great Plains to target business areas that
the typical ERP company rarely touches. Think small consulting companies
(we do), NBA teams, plumbing supply distributors, or commodity trading
companies. Think doctors' offices, general contractors, or franchisers.
Think quarries, stock jobbers, or marinas. Think small software
companies and independent insurance agents. In Europe, add in repair
depots, branch distribution centers, and custom manufacturing.
Many of these businesses have certain quite special needs. A powdered
milk trading company we knew that bought Navision needed to work
with total landed cost. The NBA teams use Great Plains partly because
the state tax for the players is very complicated. Navision and
Great Plains don't necessarily start out meeting these needs, but
they can be easily and cheaply customized by the channel partners,
who will then support the customization.
By contrast, the more highly integrated and more complex ERP packages
are harder to turn to these specialized purposes, and because the
channel partners are the Big 5 (3?), special-purpose applications
are more expensive to build and maintain. This difference, mind
you, is as much in the channel as it is in the software.
Channel. The Great Plains/Solomon channel is
uniquely suited to the functionality and the target audience,
but may turn out to be a serious barrier if Microsoft
wants to move up the food chain.
The channel is enormous--30,000 partners worldwide.
And it is the only
way of buying the software; Microsoft doesn't sell direct.
It is, of course, the same channel at heart that sells and installs the
other Microsoft products. This makes sense. The doctor's office,
the rare book
dealer, or the producers of syndicated radio shows want to buy
from the same guy who sells and maintains their NT network,
somebody who sets it up, makes it run, and comes by once a week to
troubleshoot. I always call this guy Micky, in memory of one of
them, who was a September 11 casualty.
It's important to understand that a lot of Microsoft's projected
margin is gained at the expense of Micky. The Micky I knew always
ran his little company on a shoestring, working long hours himself
and paying his friends too little as well. He would count the hours
till his next check came from the client. (Exactly like the companies
he was serving.) Micky isn't exactly Accenture.
But as long as Micky doesn't have the skills that Accenture has,
the Microsoft suite is never likely to break out from the doctor's
office. Micky can put the software in all right, and he can get
the system to create GL transactions, but he isn't necessarily going
to have much insight into how the company ought to be reporting.
The more complex the requirements, the more judgement required
in setting up the software, the more the internals of the software
need to be exploited to get benefit, the harder it is for the Mickys of
The past is prologue. None of this is new. Great Plains
et al. have never had either the functionality or the channel to
compete effectively in the areas that are core for the ERP vendors.
In the past, they could and did nibble away at JD Edwards (and
Lawson and QAD), picking off a distributor here or a small engineering/construction
company there. But over the long run, this is a self-destructive
strategy. To win deals where the fit is poor, you have such a high
cost of sales, and later, to keep the customer happy, you have such
high maintenance costs, that you simply can't mount a sustained
campaign. This is what happened to JD Edwards when it tried to compete
with SAP. And historically, Great Plains has known this and has
given into temptation only tactically.
Even now, with Microsoft's marketing prowess, there is no reason
in the short run why anything should change. Great Plains (or Damgaard
or Navision or Solomon) can simply roll up the stone quarries and
the union offices, where the only competition is the spreadsheet
and Quicken and Peachtree. Why should they move into a more difficult
market while there's still low-hanging fruit?
To move into the ERP core areas, they'd have to improve the fit.
And every software executive knows that it's really hard to develop
functionality in the areas where the ERP software is strong. Even
if Microsoft does spend $2 billion on development, they'll have
to spend the $2 billion extremely well, or it will be a waste. And
Microsoft has never shown that it's good at spending money efficiently
on software development, especially when it underestimates the difficulty
of the space.
Remember, they can't just develop ERP-level software that is comparable
to the competitor's. The competitor's software is already somewhere
inside many, if not most mid-market companies where they would be
competing. Once you have an ERP system in anywhere at a company,
you need a crowbar to get it out. So once Microsoft goes up-market,
they have to build the crowbar; they can't just build a comparable
system. And that's quite a slog.
Is Microsoft Rational?
Unfortunately, Microsoft thinks
it has the crowbar. It's .NET.
Microsoft hopes to see a new .NET product rise
like the Phoenix from the ashes of Damgaard,
Navision, Great Plains, and Solomon. This product will magically connect to any
other enterprise application through .NET. With that
product, Microsoft thinks
it will have a technical edge, a reason to replace SAP.
I don't agree, but what I think doesn't matter. The point is this. While
Microsoft is trying to bring the perfect .NET product to the market, it
will compete aggressively on price.
And, unfortunately for the ERP vendors, price is an area where they
are vulnerable. For better or worse, years and years of gasp-inducing
software price tags,
long implementation times, high integration costs, and 20-watt
benefits have made every executive wary.
The hottest part of the ERP market these days is the extension
market, where a vendor sells more software into an existing customer.
What if Microsoft starts offering comparable extension functionality,
the promise of easy integration in the future, at 1/5 the cost?
If customers really believed in ERP software, they wouldn't go for
it. But with today's disillusion, at the very least, it will cause
what we might delicately call price erosion.
Microsoft is big enough that this kind of strategy could
eventually change everybody's underlying
expectations about how much these packages cost. Our tiny company
expects to pay a few grand-plus and DuPont expects to pay a few
tens of millions. All of us could see that shaved in half by the
time Microsoft is through. They can afford it (maybe), but their
But if this does happen, don't jump to any simple conclusions about
software company revenue. This is pressure on total price. An enterprise
application involves costs in three areas: hardware, software, and
consulting. It's entirely possible that this pressure may set off
a cat fight of Biblical proportions between the Accentures, Suns,
and SAPs (or their equivalents among the et al.) In such a fight,
you have to bet on the software company. They can take measures
to reduce the required investment in hardware and consulting. So
they won't take the brunt of the blow.
But lowballing is only the first of two hammers that Microsoft
can let fall. We'll talk about the second in the next Short Take, which
we call, "All that Cash."
If you missed a recent Short
Take, click here for a listing.