- by Jack Bryar -
- Open Source Business -
On a day when VA Linux announced
major restructuring of its core business, I interviewed a
very chatty Red Hat CFO Kevin
Thompson, who thinks his company is on a fairly stable track. Thompson said that
his company has proved that you can make money in Open Source, and that
the only uncertainty is "how much and how fast."
In the short term, Thompson suggests it could still be a quite a roller coaster ride for Linux companies and their stockholders, because "we're still in a very early stage
... no new technology gets adapted overnight." Still, he says Red Hat,
under his watch, has been "very conservative" and he thinks that strategy,
plus a suite of new products and services, will help the company ride out a
techno-recession that seems to be lasting longer than he thought it would. In fact,
Thompson suggests there's a big upside for Linux developers if they understand
the implications of this shakeout.
Q: How does it feel to have positive cash flow?
A: It feels better than having a negative one. We've worked
very hard to get here. It's always good to make money. Frankly, it could
have been bigger ... six or seven million [dollars]. We made a point to take care of a
lot of a whole lot of payables this quarter rather than wait.
Q: Does this mean you're a real business, not
a "charity case"?
A: I think we've proved that Open Source does not equal "free."
We've never wanted to be a charity.
Q: Where's the money coming from these days? There's not much
detail in your financials.
A: Our revenue is coming from four to five main areas. We are still selling
Linux retail, and globally. It's still a significant contributor to
revenue. It's not growing as rapidly [as other parts of the business] but it's
consistent. As time goes on it is a smaller percentage of our income but it is
still a good cash source. We're still generating income from support
services ... to a range of companies, many of them small- to medium-sized companies
all the way up to companies like Amazon. We have a healthy custom
engineering and consulting business, using tools we've developed to service
firms ... especially semiconductor and device companies, using a fee and
subscription model as we provide support. We service a lot of embedded customers: Toshiba, Fujitsu, etc. We did in the neighborhood of four and a half million
last quarter. We're also providing engineering services planning and
technology assessments of infrastructure. That's five million. Open Source Services
and training is growing rapidly for us. The Red Hat Certification program
is very rigorous and has a lot of credibility. We're doing some operating
system consulting, that's new for us this quarter. We're looking to
deploy development consulting ... building a solid consulting practice.
We had only one revenue stream when we started, today we have over 22
different revenue streams.
Q: I see you personally contributed to positive cash flow. You made
some real money on investments over the last quarter, is that
A: We've been very solid, and conservative -- we've focused on
investments rated AAA or above. It's worked out well. Recently we've been
impaired a little bit along with everyone else in this market. Not great but
Q: You got any tips for the rest of us? My 401K's getting
A: I'm right there with you.
Q: You took one heck of a big write-down this quarter. I'm
sure most readers wouldn't understand why -- do you want to
A: A lot of this is the normal amortization that takes place after
large purchases. For us, that's running at roughly $15 million a
quarter, and that will be continuing until Q1, 2002. In addition, we
have been writing off a lot of our earlier investments in Red Hat Ventures.
We seeded a lot of startups who were developing product elements we
needed. A number of them have had some trouble in the marketplace or have had
difficulties raising additional capital and we've needed to address that by taking
Q: Now despite your operating profit, Red Hat got beaten up by a
couple of industry analysts -- you wouldn't give Prakesh Patel or any of
the other guys at Hambrecht or Goldman Sachs any "guidance," any vision
of where future profits were going to come from -- is that right?
I saw that Patel in particular, slapped you around pretty good.
A: Yeah, he seemed to take it personally.
Q: Was Patel fair?
A: Most everyone else understood that the device market,
especially, was just too large and too uncertain. That's a $4 1/2 million market
for us. We've had no cancellations, but we have had delays.
Q: He implied that good management ought to have an idea of
where their money was coming from, and you didn't seem to know. How do
you respond to comments like that?
A: We think we have good management. We have tracked revenue
carefully. We think we have a very mature team.
Q: Since he was picking on you, you want to say anything
A: No. [laughs] Most analysts we've spoken to, and I've spoken to,
understand our position.
Q: What do you expect will be the consequences of your
A: There's been a little, not a lot, so far.
Q: Do you expect you'll need to go to the financial markets anytime
A: That's one thing. We are not under a lot of pressure to go to the
financial markets. Right now we have $295 million and are cash flow
positive. If we were cash strapped there would have been more pressure to provide
guidance [to the analyst community] anyway. Since we don't need to
raise money right now we can do the right thing.
Q: What do you say to institutional investors and others who
rely on first-call reports and analyst evaluations?
A: Well, obviously we're not satisfied with the stock price. We
will do everything we can to drive our market value. Right now we are cash
flow positive. We are trading at a multiple of earnings that is relatively
high compared to some others . But we will continue to do everything we can
to enhance our value.
Q: Why did you decline to play the expectations game?
A: We felt we had built up a lot of credibility. We have met or
beat expectations for quite some time. We didn't want to put out guidance to
the analysts in the face of a lot of uncertainty.
Q: What kind of range of earnings are you looking at?
A: As I said, we've declined to give guidance at this
Q: Surely you have some kind of internal goals for revenue. What are
A: Of course. You have to have internal goals to shoot at. We have
a number of internal goals. We track our progress daily. It's just that
we are not confident enough to put these out on the street as
Q: You've put together a pretty good consulting team -- how
important are they to future revenue?
A: They're very important. Network consulting, system consulting
planning assessment ... [are] helping determine what's portable to a Linux environment.
We see a strong trend towards Linux as a unified Unix standard. We're
seeing a lot of migration from a proprietary Unix to a Linux environment
especially. There's a real trend away from proprietary Unix platforms.
Q: Isn't this a tough environment? How do you compete or
partner with an IBM?
A: That's an everyday challenge. IBM has been a wonderful partner,
especially on the hardware side. Now on Global services ... we're
trying to put together an engagement model to develop a way to work together.
We are in several deals together. We know that we will be competitors
sometimes, and partners other times. We are working on rules of engagement that
work for both of us.
Q: I understand Red Hat is increasingly targeting large
enterprises as customers.
A: We've reached an inflection point where we have credibility
among CIOs. We've made a strategic decision [to go after this business]. We
recognize that these kind of accounts will require a lot of work. We've had
successes with companies like Cisco, Nortel ...
Q: Roughly what is the size of your customer base among
these kind of enterprise accounts?
A: At one level, it's still really small. A half dozen. At the same
time we are probably in nearly all Global 2000 enterprises. We got lots
of customers -- just not a lot writing checks. Well, maybe 30 bucks for
a license. But they have seeded the market. And we do have accounts
with Ford, Toyota. Engineering services are driving lots of business.
Q: I noticed that you've estimated sales cycles in the range of a
hundred days. Given the amount of time it takes to run capital purchases
through an average Global 2000-type company, isn't that actually kinda
A: You're right. That 100 days is based on averaging small and
larger customers. But yes, the sales cycle for enterprise customers can be a
lot longer. Cisco took longer. Nortel was, I think only 95 days. But in
most cases we expect the [enterprise] sales cycle to run four to six months. And as
we get into markets like banking, where they are not generally early
adopters, we know those guys will be slow. But you have to go back six
months -- we had no visibility at the CIO level. Now we are finding bright IT
graduates, just out of college lots of times, who have been evangelizing the CIO
sometimes once or twice a week, saying, "there are better ways to do
these things." We rarely fail to make sales in these environments.
Q:You've pointed out that the chaotic state of the semiconductor
market was one reason for declining to make predictions about your
A: That's true.
Q: Frankly, is the semiconductor market a long-term market for
you? Or were most of those projects just that; projects -- one-offs?
A: It's both. Yes, a lot of these are individual projects.
They have come to us for a variety of tools that we have developed for them.
These have been individual projects, but they are projects to the same
set of customers over and over again. We really have the expertise --
we're the only guys with the capability they need. We make money on
individual projects and by providing support over a year, 18 months, and by making
Q: What are your long-term expectations of this segment as a
A: It remains important. We expect to continue to generate direct
revenue from individual projects and from annual support.
Q: Any sense as to when your new database offering should begin to generate a
A: Not immediately.
Q: How does this complicate your relationship with Oracle or
A: This isn't aimed to compete with Oracle or DB2. This is focused
on small to medium businesses, departmental users, the lower end of the
database market. We're focused on a market and looking at $2,400
Q: There's been some carping that the dB is really Postgres,
and a certain outfit based in Washington State is suggesting this is
a good example of rampant cannibalization among Open Source
businesses? How do you respond to that?
A: We build on the belief that the best way to deliver to our
customer is to deliver complete access to the code. We use the same philosophy
regarding our database offering that we take to the OS. We'll continue to pick
the best softwares, bring them to maturity if that's needed and
Q: What's the company expecting to generate overseas?
A: We expect to see overseas revenue come in at around 30%. That's
only marginally different than the 29% we've been doing. We see particularly
strong opportunities in Japan. I know that Japan's economy is an issue,
but the Japanese companies we've been dealing with see the high value
and low cost, especially on the embedded side. The Japanese have been more
willing to invest. Europe's also growing at a pretty decent rate.
We have modest expectations there, but we think Germany, the UK, Italy
and Spain ARE key. It's worth noting that all our international
business is break even or better.
Q: It sounds like you don't plan to give the analyst community any
numbers for the remainder of the year -- or is it longer than that?
A: Right now a lot depends on markets like the chip marketplace.
We've developed a lot of credibility with the street -- we've beat street
estimates consistently and we didn't want to speculate and give guidance lacking
certainty. Right now I don't know if it will be Q3 or Q4. No one
expects a turnaround in the IT market until next year. We don't want to put out
guidance or damage our credibility. We may wait, and go until our next
Fiscal Year -- go into March.
Q: Isn't all this uncertainty really a just another sign of the
immaturity of the Open Source marketplace?
A: Yes. It's still very accurate to say we are in a very early
stage in this marketplace. We're at a very early stage in adoption. At the
same time, we're in a very different place than we were in October of '99.
We're no longer at the point of determining if this is a viable or stable
business model. And it really hasn't been a factor ... it didn't affect my
ability to give guidance [to the analyst community]. If anything the market is
growing larger and faster than we thought. But no fundamentally new
technology gets adopted overnight.
Q: Given your uncertainties regarding revenues, you must have
several contingencies in place.
A: Absolutely. But we've also been very disciplined. We track our
revenue streams every day. We've also been very conservative, in
that revenue leads spending.
Q: How do you know if you're not over-invested [in
sales/development, etc.] given the uncertainty concerning revenues?
A: There is no over-spending hoping that revenue grows to match our
investments, in the way some have done. We will not get surprised and
generate large losses. It's not our approach.
Q: How are you addressing employee concerns about revenue
A: This is one luxury we have. The engineers and staff at Red Hat
work here because of their belief in Open Source -- and if you believe in
Open Source there's no place but Red Hat where you'd want to work. Our
people believe in the vision and work hard -- we've had very low turnover rate;
in the low single digits. We're not concerned about turnover as a
result of the stock price. That's not to say we ignore that price or our
employees. We work hard at giving our employees a good package and incentives.
Q: You bear the burden, along with VA Linux, of proving whether
Linux and Open Source make for a viable business model. How do you see Open
Source, and in particular the GPL, working for you or against you as a
business enabler or as a threat to revenues.
A: We were built around Open Source as a business model. We thought
that long-term the value of software by itself was zero ... that the
real value was the value of services, and that there was a particular value
around providing the customer with access to code and the ability to modify
that code and then supporting it. We've proved that this is a viable
business and removed a lot of folklore around Open Source as a business model.
We've proved it works, now we need to prove it can scale. The question is how
long it takes to develop the market and how fast it will grow.
Q: Ransom Love among others have tried to suggest that Linux
and open-code developers should strike a compromise, form a half-way
covenant with the community that allows developers to generate
code that may be open but not GPL. How does Red Hat view those
A: We support Open Source licenses. There's a thousand softwares
packaged in Red Hat Linux offerings, and those software come with a variety of
licenses ... GPL, community, BSD, etc. We defend Open Source, push GPL, but third
parties may and do release software under a variety of licenses. That's their
decision. We support them. We think all Open Source licenses are important. For
example, the Postgres license is a BSD license. That's how they chose to release
Q: You're the chief financial forecaster for the company. What has
surprised you about the Red Hat and the Open Source business so far?
A: From a macro-economic viewpoint -- the length of the slowdown.
There's really been a lack of predictability. Part of the problem is
that IT spending grew 39% last year versus a growth rate of around 13% [this year].
There was a lot of purchasing where systems weren't implemented, and that's
still working its way through companies. It's taken longer for the excess
inventory to shake out of the system, especially on the hardware side. On the
other side, the slowdown has really driven interest on the enterprise side,
as companies have been forced by the slowdown to really justify price to
performance. We think it's easy to do that for Linux companies, but the slowdown
meant that we at Red Hat were surprised. We didn't expect to see the
level of enterprise interest we've got for another six months. It's one
reason why we've accelerated our enterprise initiatives.
Q: Do you have advice for [VA Linux CFO] Todd Schull?
A: Buy Red Hat Stock. It's a bargain.