Commerce One: Everything but Profits
Commerce One began life in 1994 as DistriVision, originally focused on
developing CD-ROM sales catalogs. The company took its first steps toward
fame and fortune at age two-and-a-half when it grabbed Mark Hoffman, who
had been CEO of Sybase, to be its leader. As new CEO, Hoffman quickly
raised more than $7 million in venture funding for the infant company,
and brought several former Sybase employees along with him. Hoffman said
at the time that he supposed the company could reach $100 million within
few months later the company had a new name and a new product release
to go with it. In April 1997 Commerce One released products for building
online catalogs, for shopping and for transaction processing. Following
upon this release the company entered into its first significant relationship,
a partnership with MCI. Under the terms of the partnership MCI provided
network infrastructure and 24/7 operations for Commerce One to build its
marketplaces. MCI Systemshouse and Ernst&Young Technologies were signed
to provide consulting services to suppliers. MCI also became one of Commerce
One's flagship customers.
1998 Commerce One teamed with Microsoft, PriceWaterhouseCoopers and SAP
to offer purchasing services in Europe. Microsoft's presence was due to
Commerce One's strategy of basing its products on standard Microsoft software.
SAP was to provide ERP software. Also in 1998 the company made a fundamental
commitment to XML by purchasing Veo, Inc. Veo specialized in developing
XML based products for trading networks. Veo had been spun off from CommerceNet,
where it was instrumental in developing CommerceNet's proposed standards.
By March of 1999 Commerce One had released a fully functional online product
suite based on XML as a connector that would enable other companies to
link with its software. With this software Commerce One's customers could
go through Commerce One's marketplace, called MarketSite, to reach such
vendors as Office Depot and Grainger. An XML tool, called the XML Commerce
Connector, allowed buying companies to link from their own internally
maintained catalogs into MarketSite, and also linked MarketSite to supplier
catalogs. Overall, Commerce One's position regarding business languages
in general and XML in particular has been to support open connectivity.
same year, 1999, was Commerce One's IPO, which brought the company $69
million. However, soon afterward Commerce One suffered a setback when,
as a result of its merger with WorldCom, an Ariba customer, MCI switched
to the Ariba Network. But the loss was not fatal (although. Ariba's PR
people made the most of it). By the end of the year Commerce One's stock
price had shot up twenty fold, giving it a market capitalization of $20
billion, and it had partnerships with companies like British Telcom, Nippon
Telegraph and Telephone, PeopleSoft and such financial institutions as
Toronto-Dominion Bank of Canada and Banacci of Mexico. It had signed to
develop General Motors' GM Marketsite, and had purchased CommerceBid to
gain the technology to add auction capabilities to its product line.
into the year 2000 Commerce One has been moving strongly, especially in
the international part of the world. It has developed regional trading
hubs in China and South Africa to add to its others in Japan, Australia
and Southeast Asia, Europe and North America.
Figure 1 shows the company's revenue history. Through Q3 of 1999 license
revenues were growing faster than exponentially and service revenues were
growing slower than linearly. This was an excellent position to be in.
(Growth in revenues has been so amazing that were it to continue at those
rates it would exceed $5 trillion within two years. So, of course a slowdown
was to be expected, and must be seen as healthy. Growth in Q4 seems to
have slowed in geometric terms, but this should not be of much concern,
especially given the dramatic upturn in Q1 of 2000.)
licensing is not where the company expects to make its money. Much of
its revenue is expected to be in the form of transaction fees for the
purchases made on the various networks it manages or has revenue partnerships
with. The company also expects to garner revenues from commissions on
auctions and from subscriptions to hosted services. There are also revenue
sharing agreements with at least 75 portal partners and the company is
developing other services.
Overall, to date the only significant revenues have been from license
fees. This is in line with all reasonable expectations, but we'd look
to see those revenues begin to flow by Q3 of 2000, and Commerce One agrees
that the next year or two will be critical in proving their financial
Strategy and Trajectory:
Commerce One's initial strategy can be contrasted with that of rival Ariba,
which went IPO a week earlier. Both companies discussed desktop (buyer-side)
software and marketplaces. Commerce One claims that their emphasis from
the beginning was always on the marketplace and that Ariba's was focused
more on the desktop. It is possible to read the initial documents this
way, although both companies left plenty of room to reset their directions
after launch. However it is clear that Commerce One has been moving aggressively
to build a large global network, and that the development and growth of
the marketplace is a keystone of its strategy.
Commerce One identifies three types of marketplaces as being central to
Mid-market: These are marketplaces dedicated to a single midsize
company's supply chain or to a vertical niche. The Promus Hotel chain
runs a marketplace of the former kind and OmniCell has one of the latter,
in the healthcare industry.
Exchange Markets: These are larger markets organized around supply
chain optimization. Examples include Commerce One's partnerships with
General Motors, Shell and CitiCorp.
Markets: Commerce One has been busy building regional marketplaces
though partnerships with such companies as BT in the United Kingdom, Cable
and Wireless Optus in Australia and NTT in Japan, to name about a quarter
of these partners.
Commerce One ties these three market types together loosely in a Global
Trading Web. For example, a pub equipment supplier in Australia might
create a marketplace for its supply chain. If the marketplace operator
allows, members of the marketplace can be given access to the regional
market or even to the Global Trading Web, which would in principle allow
them to sell their beer mugs and skittles to companies across the world.
There are about twenty core regional participants in the Global Trading
Web, with about twenty core verticals in each region.
Besides its marketplaces Commerce One maintains its desktop purchasing
application, Commerce One BuySite, which currently supports about fifty
enterprise customers. Commerce One's target customer is the large multi-national
member of the Global 1000 or Fortune 200 with $500 million in indirect
spending and 500 or more employees. The company is also looking for ASP
opportunities, which Corio is the first to develop.
Commerce One has the philosophy that partners should share risks and rewards;
as one executive puts it, "both sides should put their skin in the game."
Thus, the company will take equity positions in small market makers, trading
revenue sharing for lowered license costs. On the other hand, some of
Commerce One's larger partners have taken equity positions in it.
One tends to believe that it is on track with both the BuySite application
and the marketplaces, and is looking to extend its capabilities and offerings.
A prime area here is developing supply chain solutions, which are being
pursued through partnerships with Aspect Development (but see below) and
with Adexa (see TEC note: What's
in a Name for Supply Chain Vendors?)
is in the area of auctions; its Commerce One Auction Services product,
based on technology acquired with its merger with Commerce Bid, can be
sold as a standalone product or integrated with a marketplace solution.
The overall direction of the company is to be a provider of enterprise
portal solutions that are leveraged thought the Global Trading Web.
Commerce One's hierarchical network structure is an attractive environment
for buyers and suppliers.
Commerce One has also built a very strong position in XML and catalog
management, the latter through its acquisition of Mergent Systems. It
has displayed a willingness to be an open system, much more than has Ariba
to date, and this is helpful because it enables partnerships of many kinds.
Not the least of these is its willingness to make network membership available
through second tier products. Peregrine Systems and RightWorks are two
of at least fourteen partner companies that can compete for the smaller
companies within Commerce One's range, but which proudly offer access
is important to recognize that Commerce One's commitment to XML is far-reaching
and represents a real technical strength. In particular, where XML itself
is a standard that is technically neutral to the values passed in fields,
Commerce One has incorporated data field validation ("strong typing")
into its schema language.
For any company in Commerce One's position - high market capitalization,
done almost everything right, never shown a profit - the biggest challenge
is going to be to keep the music playing. Although losses have been growing
steadily, Figure 2 shows that over the last year the loss per dollar of
revenue earned decreased significantly, which suggests that the company
may be moving in the right direction.
that Commerce One's business plan still calls for it to make the majority
of its revenues from transaction fees, it is unfortunate that much of
the control of whether that happens is passing into the hands of other
large companies. We have no reason not to expect that the various regional
marketplaces will develop as long as the global economy stays strong,
although it is not clear at what rate that will happen. However with the
large exchange markets Commerce One is somewhat at the mercy of the relationships
between the major partners (such as GM and Shell) and their supply chains.
Those are complicated relationships that might not bend easily to the
use of an exchange market.
One will likely be under pressure to match the incredible partnership
between Ariba, i2 and IBM (see TEC note: B2Big
Deal for IBM, Ariba, and i2)
While it has partnerships with Aspect (which the company says will continue
despite Aspect's absorption by i2 (see TEC note: IBM
is not Enough: i2 Snatches Aspect and SupplyBase), with Adexa, and
with various system integrators, they do not match the magnitude of Ariba's
sheer numbers a case can be made that the difference isn't huge. For example,
PriceWaterhouseCoopers has 500 consultants dedicated to Commerce One and
there are similar arrangements with Arthur Andersen. IBM has not stated
how many of the 138,000 people in its Global Services Division will be
dedicated to the Ariba/i2 partnership, although it has promised to have
"thousands" of salespeople working on it.
Regardless of numbers though, Ariba's deal is breathtaking because of
the involvement of IBM. Of course Commerce One will continue to develop
solution and partnerships that reach through the supply chain, but it
is not clear that Commerce One should try to match the enormity of this
partnership. The integration aspects of getting the three companies to
play together at the technical level have barely been examined. (And Ariba
has other integrations on its mind as well). If they prove difficult Commerce
One may look stronger continuing as it is going. Whereas if the integration
is smooth and easy the worst that will happen to Commerce One is that
it will end up being a successful and rich "number 2."
completeness, we note that Commerce One has been keeping a large cash
reserve for more than a quarter, much larger than the management textbooks
like. However, we suspect that Commerce One doesn't need our suggestion
that those resources can be better used than for stuffing a mattress.
Commerce One has enough momentum that even a major stumble will keep it
moving forward for a considerable time. But the precise direction it takes
will depend on small forces that are difficult to discern.
We think that Commerce One is quite committed to building a huge revenue
base from fees (transaction, commission, and subscription) in its markets,
and that unless everyone is guessing wrong about the digital economy those
revenues should be looking healthy within the next 18 months to 2 years.
Whether that means profitability for the company as a whole depends on
whether Commerce One gets to slow down its expenditures. While absolute
expenditures have been increasing, as Figure 3 shows for the key areas
of Sales and Marketing and Product Development, Figure 4 shows these areas
shrinking as a percentage of revenues.
In short, it seems that if Commerce One keeps doing what it is doing,
its future will be rosy. We think that only a major change in its business
model would interfere with this happy story. But such a change is likely
to be imposed on them by the growth of new models for doing business on
continue its move toward profitability without giving up its technological
growth Commerce One may choose to give its stockholders both the satisfaction
of success that everyone desires and the thrill of remaining on the roller-coaster
car named Internet. It could do this by spinning off a profitable market
management business as a separately traded subsidiary, or spinning off
a number of operating divisions under a holding company shell, leaving
something looking like a B2B version of CMGI.
While the next product directions have not been announced, we expect to
see strong use made of the company's depth in the area of XML; an enterprise
level content management system of some kind could be a likely candidate.
Commerce One has already acquired Mergent Systems for its catalog aggregation
and management technology; we're postulating that a more general content
management product (such as a corporate portal) might be a logical way
to gain value from the XML investment (probability 30%).
When Commerce One and Ariba started on their historic journeys, their
focus was logically on larger enterprises. However, the mid-market is
huge and extremely interested in cost reduction. Some of these companies
will find their way to verticals, but many are primarily buyers of MRO
goods. Commerce One can capture a lot of this market's transactions by
continuing to build partnerships with smaller vendors of purchasing software
and purveyors of corporate portals.
can imagine good results from a tasteful marketing campaign about how
Commerce One is very interested in this segment. Such a move, we think,
would require Commerce One to provide marketing and technical support
for its partners. Commerce One could win big in transaction revenues and,
incidentally, spread its flavor of XML very widely.
Companies in Commerce One's target market should have no concerns about
selecting Commerce One's BuySite for MRO purchasing. We think that Commerce
One will have to work a bit harder to make sales to companies with aspirations
of bringing e-commerce to their entire supply chain. We don't doubt that
they will have an adequate solution, but the potential of the i2/Ariba
partnership and the uncertainly about where Commerce One's cooperation
with Aspect will actually lead requires evaluators to ask particularly
hard and probing questions.
large market makers will also be comfortable putting Commerce One on their
short list, and should give it extra points for its strong international
companies, those at the low end of Commerce One's target range and below,
can also consider Commerce One through license or, as the capability develops,
ASP. But they should shop more carefully because the additional capabilities
available from portal and self-service vendors may meet their other needs,
in addition to purchasing, in the same package.