The Judgement of Paris
We're using the end of 2002 and beginning of 2003 to take a look at some out-of-favor areas that we
like. In this issue, we return to sourcing, looking at a perennial power (FreeMarkets), a company with endless
bad news (Ariba), and an up-and-comer (Emptoris). If you could shake off the history, my choice among the three
would be Ariba, but this might not be everyone's take on it.
The Sourcing Shakeout
In the summer of 2002, we released a report on e-sourcing that
covered 43 vendors. We found a crowded space and a confused one.
The 43 solution offerings came from every direction. There was
software; there were services; there were outsourced services. There
was specialty software by spend category. There was specialty software
by function and by industry. And (surprise, surprise) there was
plenty of Power Point software from the ERP vendors.
In another time and place, the existence of so many applications
might be thought of as a response to customer demand. But when
we talked to customers, they told us a different story.
Customers didn't want choice. They didn't want
highly specialized applications. They wanted a single solution:
a single relationship, a
single piece of software, or a single approach, one that would save
them money this year, but also improve the overall maturity of their
sourcing processes.
Alas, the best the vendors could do at the time was respond with
messaging. Frictionless said it had an end-to-end solution. B2eMarkets
said it had a strategy solution. Ebreviate said it had a
full-service solution. Perfect said it had a Perfect solution.
And, of course, PeopleSoft said that it had everything.
The hope, obviously, was that what happens elsewhere would happen
here. Early adopters would take a flier on inadequate solutions
and provide the funds to build out what was actually needed.
This just didn't happen. Why? I think the gaps were so large that
even software buyers (not a notably perspicacious group) couldn't
see any way around them. Since no total solution was in the offing,
they decided to go to something with an immediate return, namely FreeMarkets.
The Elusive Total Solution
To understand this market at all, you need to understand
this reaction. In my view, it is not as if people preferred
FreeMarkets or thought it was a good alternative. It is that
they saw too many gaps in more technology-oriented proposals.
It is important to understand why those gaps were so big and so
obvious.
The essential
fact about purchasing departments is that they buy
everything a company buys. Everything. Legal services.
Toilet paper. Electricity. Floor tile. Airline tickets. Fork lifts.
If you're a purchasing department and are going to go to the effort
of buying and installing a software solution, you're going to want
something that helps you buy all this stuff. But most of the software
out there is pretty clearly oriented toward a particular buying
process, one that just plain isn't useful for a lot of these other
things. If you're buying legal services or real estate, for instance,
you're probably not going to go to the corporate catalog and punch
out to legal.com or realestate.com. If you're buying airline tickets
or implementation services or cafeteria management, you're not going
to think that auction software is just the ticket.
The leaders last summer were essentially unable to convince
their customers that they could help them buy everything. Today,
this has left the software vendors in a bit of a pickle.
Frictionless is losing staff, B2e (in the words of one observer
close to the company) is in M&A mode, eBreviate has been offered
around (with no obvious takers), and PeopleSoft has moved to Plan
B, which is redefine the market.
The software companies might still have gotten some traction if
there hadn't been another problem. Even when their solution had
some real promise—had the right buying models and some other
needed bells and whistles—it was still going to be hard to
bring it into a company. To make such a solution work, you also
need expertise, expertise not just in using the solution, but also
expertise in the spend categories where the solution is put to work.
But most of these companies had said, "We're a software company.
You don't get expertise from us."
For the expertise, they said, you can go to our consulting partners
or get it from your own organization. For the very few buyers
who already have lots of expertise, that was great. But for most
buyers, the buy was getting too complicated. It was simpler to turn
back toward FreeMarkets.
It's also important to understand what FreeMarkets was offering
that the others weren't.
Let me repeat. FreeMarkets was not providing the total solution that
people wanted. They're fundamentally
in the business of selling hosted auction events, and that is by
definition piecemeal. What differentiates them is that FreeMarkets
is a one-stop shop for these events. They provide
all the services and (more important) all the expertise that you need.
When you use them for an event—as opposed to running an auction
event yourself using auction software—they help you determine whether
an auction will work, how it should be structured, and what the
goals should be. They even recruit and train suppliers.
So, from a customer point of view, even though FreeMarkets doesn't
cover every spend category or process, it was and is preferable because
when you do go to them, the event is held right. Not that they were
exulting over this; FreeMarkets
is expensive. An event might well cost roughly 50% of the savings,
and FreeMarkets isn't exactly shy about estimating savings.
So Who Has Traction Now?
In the last six months, the allure of a total solution
has not dimmed. But there are still very few alternatives
to FreeMarkets.
Of the 43 companies, many have just lost the race. They couldn't
persuade enough people, customers or VCs, to fund development of
the products they claimed to have already. Six or so of the 43 are
out of business or acquired. As noted above, most of the rest are
laying low. But there seem to be two exceptions: Emptoris and Ariba.
Emptoris has a new and much improved version out and some strong
commitments from brand-name customers. And Ariba seems to be
in every deal. In the words of one rueful competitor, "Ariba has deep pockets,
and while their product isn't all the way there, it's come far enough."
The competitor is particularly struck by the new software
that Ariba has been developing
for the last 9 months under the rather inviting rubric, Project Venus.
(Inviting, that is, to those of us who have a weakness for cheap witticisms.)
So now, the hapless consumer seems to have three, quite
different choices about where to go. Since one of those
choices is Project Venus, I guess I'll have to call
the problem for the consumer the Judgement of Paris.
Paris, you may remember, was the son of the King of Troy. He was
asked to decide which Greek goddess was the most fair: Hera the
powerful, Athena the wise, or Aphrodite (Venus) the beautiful.
By analogy, customers are now asked to choose between the powerful
(Emptoris), the wise (FreeMarkets and the consulting companies like
Kearney), and the beautiful (Ariba's Project Venus).
The Powerful
To see what the first choice looked like, we visited Emptoris and
saw a demo. Originally a vendor of a pure auction/evaluation tool,
Emptoris is now going to market as a full suite vendor, with a particular
focus on long, complex sourcing projects. Their name clients are
Glaxo and Motorola. Several other well-known companies (like Avon)
use them through ICG Commerce. (Emptoris is the engine that ICG
Commerce uses for hosted events.)
The product clearly would appeal most to companies that
want to manage a long sourcing process for a
complex set of (often direct) goods. The old RFP/auction engine
is now embedded in what you might describe as
a multi-participant, multi-stakeholder project tool. This allows
managers to create a process specific to a buying opportunity
and get a degree of involvement from many people without
impeding the process. A group can be asked, for instance, to evaluate
vendors, and the evaluation is simply fed into the total scoring. Or,
a vendor can be shown where it is falling down and be asked to
modify a bid.
This kind of capability is pretty good, relative to, say, an ERP
vendor with an auction tool. The ERP vendor simply doesn't provide
the process management or project management you need in order to
get full-scale auction management. In an ERP system, you can create an RFP/bid,
distribute it, and score it, but you can't
manage the process with any delicacy.
With Emptoris, on the other hand, there are workflow, project,
and RFP toolsets, all in the same system. With these, you
can model the processes that companies like Motorola or Glaxo actually
use when they buy. (Clearly, this is why ICG Commerce finds them valuable; ICG
runs outsourced procurement projects, and they can use their internal
experts to tailor Emptoris to their customers' needs.)
Unfortunately, for companies like Motorola or Glaxo, the ultimate value
proposition for this is still questionable. You can use this toolset
to create a full sourcing platform, but brother, you're going to
work at it. Somebody at your shop is going to have to convert the
workflow tools into processes for each separate sourcing process.
I get worried when I think about doing that for the forklift
buy, then for the airline ticket buy, then for the legal services
buy, across countries and continents and divisions. It's a lot of
investment, just in infrastructure.
Concerns like this make me think that the only companies that will
like Emptoris are the ones who have decided that they can afford
to commit substantial resources to the project. In my view, there
aren't too many of them.
The Emptoris plan might be to get a limited number of committed
customers, then sell less capable, less committed customers based on
the rave reviews of the initial buyers. But will those reviews
be forthcoming from these customers when much of the value add came
from them? Will they think it's worth what they paid for it?
The Wise
For the wise, the choice is simplicity and a clear
return on investment. Go to FreeMarkets (or ICG Commerce
or Logistics.com), pay their rather high fees, but get
a better deal than you'd be able to get on your own.
Go, in other words, to these companies for the
expertise and the process management that you can't
get yourself.
From a customer point of view, there's nothing wrong with this,
but there's a strong desire not to do it forever. For one thing,
you still can't use it for everything. For another thing, it sells
your own efforts short. Why can't we do this ourselves, at least some day? It
isn't, as we say, the total solution, one that lifts our organization's
effectiveness permanently.
Of course, this complaint about high-cost expertise is not new.
People have been going to Kearney or the boutique consulting
firms for years for even more expensive expertise.
They keep on saying they would like to bring expertise
in house, but they don't. And relative to Kearney's cost,
FreeMarkets is providing expertise at quite a low price.
They can do this at a lower cost than Kearney because of two potential
economies: one, they are focused on events, not strategy consulting. And two,
their experts are supported by software.
Mind you, it's still expensive, so expensive that you have to
wonder about the business model. FreeMarkets supports
literally hundreds of spend categories, has "trained" thousands of
suppliers across the globe and hires or has access to
thousands of category experts.
Is this enough investment in expertise for
FreeMarkets to be a reliable source of expertise in any area a
company might want? It's unclear. Even with this
investment in expertise, can the company be
self-sustaining and highly profitable? It's unclear.
The cost basis is really
high. You have all these experts sitting around, and they have to be used.
FreeMarkets has tried to solve this problem by keeping their
rates high. According to many customers, too high.
The Beautiful
The problem with beauty, people tell me,
is that it is only skin deep. This has indeed been a problem
with Ariba applications. The
products do, but they don't do too much.
The new set of products is still a little surface-y, but it is not
too bad. Crucially, it is end-to-end (with workflow,
project management, etc.) Project Venus
allows you to manage activity by spend category, and it integrates
analysis, strategy, buying, and contracting in a relatively straightforward
way. There is a rather thoughtful workflow package and more than a little
attention paid to knowledge management—all pluses. It may not have the
toolset that Emptoris has (I think they're roughly comparable, but can
see why Emptoris is attractive). But it has at least crossed over
the hurdle that the ERP companies are balking at.
The end-to-end management may not be perfect, it is done
with some care and some effectiveness. Unlike Emptoris, which emphasized
its toolset, Ariba emphasizes good looks and (hence?) usability. Demos
are full of nice-looking dashboards, practical-seeming work lists, and
good communication features.
Underneath, there are also some interesting
wrinkles, such as some serious support for direct (BOM) spend and a somewhat
mature optimization engine attached to the RFP evaluation tool.
Still, it isn't so good that by itself
it would inspire rueful comments from the competition.
What I think really gives it some potential is
Ariba also trying to provide the service and the expertise that this
space evidently requires.
To begin with, they've decided to build up their service capabilities.
They've hired a director of consulting services from the consulting industry
and made it clear to buyers that significant amounts of services and some
expertise are
part of the package.
Leading with a lot of service is unusual and somewhat
risky, of course. Software companies are always told not to get into the
service business. And clearly there are some worries about
Ariba. The fact that Eileen Basho, a really good services
person who came from Andersen left
Ariba just as they were beginning to ramp up the service business has
to be worrisome. But
to me, the logic of thing is inescapable. For sourcing,
services and expertise are part and parcel of
delivering a total solution. To set up the right processes, templates,
stakeholders, and strategy for any spend category, you can't just
throw in some off-the-shelf software. Either you don't manage
the category, or you do some implementation, which means services.
The other element in the mix is that Ariba is
trying to develop knowledge capital. I wouldn't call this a bet-the-company
initiative, but the ideas are interesting. One idea is a "category
sourcing kit," a set of processes, templates, strategies, and data that
Ariba develops, which can be used by a company as a starting point for
its own spend strategies. Another idea that already seems to have legs is
to allow companies access to cleansed total spend information gathered
through the Ariba Supplier Network. Yet another is fairly robust capabilities
in Knowledge Management, so that experience with one spend event or supplier
can be fed back into the choices.
Of these, only the kits are a revenue source.
The kits will be created by development (I've talked to one developer,
but haven't reviewed them), but added to and maintained
by the services group. Interestingly, if they are done right,
they become a subscription service and one with significant potential
for Ariba. There are lots of spend categories and lots of industries; if
users simply download new versions once a year, like TaxWare, the
revenue opportunity is not trivial.
The Competitive Outlook
So which goddess do you choose, Mr. Paris?
According to the legend, all was not pure and aboveboard when Mr.
Paris was picking his goddess. There were bribes. Hera promised
power, Athena wisdom, and Venus promised a woman named Helen,
who then lived in Sparta.
In this market, too, success may well depend on the quality (and
plausibility) of the promises.
A lot of these promises will turn on a notion that is common in
the procurement world: maturity. The more mature an organization,
the better organized, the more strategic, and the more effective.
The promise of a company like Emptoris is that they will bring a
reasonably mature organization to the highest level of maturity.
A FreeMarkets, on the other hand, promises a less mature organization
that it can be more effective today, but without much long-term
gain in maturity. Ariba promises a somewhat wider range of more
mature organizations that with time, their maturity will improve. Different
promises really and in one sense, therefore, a market with room for
all three.
But I think the Ariba promise has the
greatest likelihood of hitting home with the average prospect (just
as Venus's did). Unfortunately, with that prospect, there are some other
issues.
Beauties in vain their pretty eyes do roll...(A. Pope)
You see, however attractive the Ariba offering, it just hasn't
grabbed the market the way Helen of Troy did. We are sending this
out in advance of Ariba's earnings announcement, but frankly, we
don't expect that the announcement will reflect any big jump
in demand.
At Ariba Live! in April,
I saw or talked to several teams of big-name retail and CPG
customers who were pretty serious prospects. But in November,
when Ariba announced Project Venus, not one of them was on
the podium.
I've
talked to a number of people not at Ariba, but in a position to know about
Ariba deals, and they confirm my worry. Deals
for this new software have been tough to get. Better sellers
have been new modules (like Invoice) that are extensions to
Ariba Buyer.
Not that sales are zero. Ariba has been pretty good at maintaining
a steady state of $50-$55 million a quarter, and I expect nothing
different for the still-delayed Q4 '02 or Q1 '03. These
numbers are big enough to sustain the company while the new
product gained acceptance, even if acceptance was slow.
But we still haven't seen any jump. Is this due to the general
slowdown in IT spending? Yes and
no. Buyers for this kind of software are purchasing people or the CFO, not the CIO.
These people
are more likely to be slowed by skepticism than by lack of IT budget. At
the same time, being purchasing people, they hate spending money, and they're
always looking for a good deal.
So what's slowing things? Well, there's always the fact that Ariba is Ariba.
I myself think
that the company has moved well beyond what it was in the Keith Krach-Larry Mueller
days. But you'd scarcely know it from the announcements. The earnings announcement
is still delayed as Ariba figures out what to do with one of the Krach-Mueller
shenanigans. (And, inevitably, there have been shareholder suits.)
The balance sheet/income statement is still so encumbered
with losses and faux good will from those days that it is impossible to understand
the actual operations. Analysts don't like this, and at some point, neither do buyers.
There is one ray of light in all this. The idea
that a best-in-breed is right has clearly established itself in the space.
For a while, it appeared that the ERP vendors would rapidly catch up to the
niche vendors. All announced
SRM initiatives and e-auction software last year. But all have
been a day late and a dollar short. SAP has been particularly disappointing,
unable to exploit several natural advantages. But neither PeopleSoft
nor Oracle (probably the best product of the lot) have done any
better. In the process, a "wait for our ERP vendor" strategy
has begun to look pretty bad, and these days CIOs have as little tolerance for looking
bad as they do for spending money.
Ultimately, I'd like to agree with Pope, who continues the line
above with "Charms strike the sight, but merit wins the soul."
At Ariba, I've seen the software,
gone over the issues with developers, and talked to the customers, and I think
the merit is there. So I'd give the golden apple to Venus. But in doing
so I might just prove once again what everyone knows already, that
I'm no son of kings, no slayer of Achilles.
To see other recent Short
Takes, click for a listing.
|
