What’s wrong with Linux services?

I never expected to defend Red Hat’s business model in a column. But
the company has been getting pounded by the investment community for positioning
the company in the one segment where they could make some money. Go
figure.As you may have heard, Red Hat was “downgraded” last week by yet
another analyst, this one from ABN Amro. The result was another small
but sharp sell-off of the company’s stock. By most rational measures the
company has done well, according to its most recent quarterly report.
Red Hat reported smaller-than-expected losses last quarter, and grew its
revenues by another 15%. But ABN’s Keith Bachman said he was
disappointed. Bachman said, “It was in line with our projections
[but]… we hoping for more.” Other analysts didn’t share Batman’s’
disappointment. One, Katherine Egbert, of Thomas Weisel Partners,
slightly upgraded her predictions about the company’s future
performance. But the stock sank anyway.

The reason for these mixed reviews and weak prices has to do with the
shift at Red Hat, along with some other Linux firms, to a revenue model
based more on services and systems integration. “We’re a bit worried
about [the] revenue mix,” Bachman said.

What’s wrong here?

Most Open Source vendors have always said that their business models
were going to be built around services ranging from training to systems
development and support. Services have long been the fastest growing
segment of the IT marketplace. Many other IT firms, including more
traditional companies like IBM and Oracle, have greatly expanded the
service sides of their businesses. Firms like Unisys have become almost
completely dependent on services for their income. Many of these firms
generate respectable profits every year.

Growing marketshare

The Gartner Group predicts most major corporations will increase IT
spending substantially over the next five years, devoting 10% of their
revenues to IT spending, compared to around 7.5% today. Most of that new
spending will be on services. “There is an insatiable demand for
technology and services,” according to David Carlick, of Vantagepoint
Venture Partners, a San Bruno, Calif., investment firm.

Most studies indicate that the Linux-friendly customers will expand
more rapidly than other enterprises. Smaller firms and IT-intensive
e-businesses will generate a disproportionate amount of demand, and will
grow and increase their IT spending a lot faster than other
enterprises.

According to one study, the average solutions provider is generating
more than 80% of its service revenue from e-business services. And the size
of those deals has begun to grow dramatically. At MarchFirst Corp., the
average e-business services deal has jumped from $2 million to
better than $12 million in just over a year. Smaller specialty service
providers targeting e-businesses and dotcoms have seen revenue jumps
averaging better than 30% over the last year.

Linux, in particular, is a red-hot service market. Recently VAR
Business commissioned an end-to-end e-business study and found that the
number of companies looking for Linux-based solutions had
skyrocketed.

Growing competition

But had it grown as fast as the firms willing to provide Linux-based
services? According to the VAR business study, the number of solutions
providers had grown by nearly 40% in less than nine months.
More companies offer Linux-based services than all flavors of Unix
combined. That’s not necessarily a good thing. As I pointed out a
couple of weeks ago, Linux still trails operating systems like
Solaris as the O/S of choice among e-businesses, and traditional
companies have been reluctant to embrace Linux-centric service
solutions.

That worries investment pros. Several that I spoke to suggested that
the very nature of Open Source made it too easy for new firms to enter
the market. They worry about a glut of new Linux-based service vendors
depressing profit margins and stalling growth of individual companies.
They point out that international markets may already be closed to
U.S.-based Linux specialists because local vendors are rapidly
establishing themselves in local markets. A Korean vendor agrees. 3R
Soft’s Jeason Yeu recently told Business Week, “With the source of
technology open to all for Linux, Korean software designers are placed
on an equal footing with those in the U.S.” U.S. vendors don’t have
expertise in foreign languages or the peculiarities of foreign markets.

Many of these same pros think that they know something about the
service business, and the one thing they think they know is that
services are hard work. Revenue growth isn’t easy to generate. Ask
Unisys. Although the company continues to generate reasonable profits
each quarter, the company’s gross
revenues stalled and have been in decline over the last year, in
particular. Gross profit margins from services have been 25.1%, or
nearly half of the 46.8% margin the company makes in its “technology”
segment.

Linux firms are trying to play by different rules than old-line
service firms. But analysts worry that Linux service companies may be at
a severe disadvantage compared to larger firms capable of offering a mix
of services across multiple platforms. They suggest that the IBMs
and Oracles of the world are more likely to be able to put together the
kind of mixed platform, end-to-end array of solutions that most
customers are looking for. As the manager of one retail Web site said to
me, “If IBM can do Linux and Windows and manage my database and my
hardware, why would I use anyone else?” That kind of thinking helps
explain why companies such as IBM Global Services saw a $3.2 billion
increase in revenue their revenue last year. It also explains why
service-oriented Linux firms are in trouble with the investment
community.

But part of the reason Red Hat and other service-oriented Linux firms
have suffered in the marketplace is because the irrational greed that
rocketed Linux, dotcom and e-service valuations through the stratosphere
has turned into irrational panic. Even non-Linux services firms like iXL
and Cysive have been pummeled in the marketplace, as “analysts” look for
reasons to downgrade the same firms they hyped the year before.

Screaming growth rates? They’re a cause for worry that companies will
bleed red ink trying to manage their expansion. Moderating growth rates?
They’re a cause for worry that the company has crested and its upside
has been over-estimated. Deals growing in size? That’s a reason to worry
that the time to close deals and complete projects will grow even
faster. A customer base full of smaller, rapidly growing dotcoms and
e-business ventures? Instead of celebrating, it has become a reason for
panic the first time one of those companies defaults on its debt.

Is the current market as badly oversold today as it was over-hyped
last year? As I said a
couple of weeks back, the venture community thinks so. They wouldn’t
be pouring new money into Linux start-ups if they thought otherwise. And
while other segments of the Linux community (such as embedded systems)
may be generating the greatest buzz at the moment, services firms almost
never go bust. That’s something investors ought to be able to take to
the bank.