The reinvention of Progeny: How one FOSS company survived the dot-com crash

177

Author: Bruce Byfield

In June 2001, Progeny Linux Systems was in crisis. Looking around, co-founder and CEO Ian Murdock realized that the company needed fundamental changes to survive. Four years later,
Progeny is back up to its former staffing levels and showing modest profits. It is also one of the few

Free/Open Source Software (FOSS)-based companies from that era to survive. Murdock’s assessment of
where the company went wrong and his story of how it reinvented itself offer some practical suggestions
for other start-ups, especially FOSS-based ones.

Progeny Linux Systems started operations early 2000 with modest funding from the Linux Capital Group, a short-lived venture
in which Bruce Perens was co-founder and president. By May, Progeny was hiring rapidly and beginning develop Linux NOW
(Network of Workstations), an updated version of Sprite, a research
operating system developed at the University of California-Berkeley that provided a single system image to a cluster
of work stations.

As a first step toward Linux NOW, Progeny also began development of Progeny Debian, a distribution based on a mixture
of the Debian stable and testing distributions with administrative tools developed by Progeny. Progeny Debian rapidly
became an end in itself, with an active mailing list and a retail version released in April 2001.

By May 2001, after a year of operation, Progeny numbered more than 20 employees. Employee morale was high, with one
employee exclaiming at a company meeting: “This is the job that I want to retire from!” As scheduled in the
business plan, it was also rapidly running out of money, and searching out venture capitalists to raise more. But what
Progeny had not anticipated was the the dot-com collapse.


Beginner’s mistakes

Like many startups, Progeny made its share of mistakes. Looking back, Murdock identifies several key
mistakes that Progeny made in its first year.

  • For starters, like many dot-com companies, Progeny hired too quickly and indiscriminately. The fact that
    many Progeny employees were Debian developers and therefore plugged into a network of potential
    employees may have increased this tendency. Murdock now believes that Progeny felt compelled to
    hire, simply because it had the cash and its officers wanted to be able to show investors that they were taking
    concrete steps to organize the company. “There was a time,” he remembers, “when we were hiring
    people when it wasn’t clear where they would fit into the overall organization. We just had a sense
    that we had to hire people, that we had to do something.” Unsurprisingly, this hiring practice did little
    to help the corporate bottom line.

  • More importantly, with the exception of Bern Galvin,
    the chief financial officer and currently a member of Progeny’s board of directors, none of Progeny’s
    officers had business experience. Most, like Murdock himself, came from programming backgrounds.
    Murdock felt his inexperience keenly, and worked hard to learn, accepting coaching from
    Bern Galvin as well as Garth Dickey,
    a local Indianapolis VC. All things considered, Murdock now says, “I think I did a reasonably good job,
    considering how little I knew.” Still, the best time to learn business fundamentals is hardly when you are trying to
    run a startup with a small, overworked staff.

    According to Murdock, this inexperience showed most in the business plan. Linux NOW was a large-scale
    project, requiring at least a couple of years of development and offering few chances for an immediate
    return. Murdock remembered that “we had a very ambitious business plan which was essentially dependent on a continued ability to raise additional capital — a pretty typical attitude for the time. There was almost a virtual guarantee that you
    could go back out to the market in six months or 12 months and get the capital you needed to take
    the next step.”

    In fact, Progeny actually turned down funding in its first year, expecting that it would always be available.
    By delaying the company’s second round of financing, Murdock believed that “the company would be
    bigger” — therefore, presumably, making it more attractive to investors — “we would have to give up less of the company
    in return for the stock sale, and the existing shareholdings would be diluted less.”

    However, by the time the company needed another injection of cash in June 2001, venture capital was
    drying up. In early 2000, the Linux Capital Group had agreed to finance the company after a single
    meeting — in part because, although Perens and Murdock had never met, they knew each other by
    reputation because they had both been involved with Debian. A year later, Murdock spent six months
    making the rounds of VCs without success. Being new to business, like many in the dot-com boom,
    he had failed to anticipate or observe the first signs of the crash until it was too late
    to do anything about them.


    Meeting the crisis

    When the crisis hit in June 2001, Progeny responded on two fronts: reducing operating costs and finding
    immediate revenue streams. Only after the immediate crisis was over did the company have time to consider its future.

    Overnight, staff was cut by half. The cuts could have been devastating, but Murdock believes that they
    were softened by high morale among the survivors. Several of those cut were also re-hired later on contract
    as funding allowed. In addition, Murdock himself, feeling responsible, went off-salary, surviving for several
    months by doing research work for the University of Utah. Although staff went through a long “period of caution,”
    in which everyone was uneasy about the future, the reaction to the crisis was not nearly as bad as might be
    expected. “If anything,” Murdock believes, having survived the crisis “pulls people closer together.”

    Having contained costs, Progeny still needed immediate funding. In theory, the retail release of Progeny gave the
    company its first revenue stream, but, in practice, the stream amounted to a few drops. Although over 50,000 users
    downloaded Progeny Debian, according to Murdock, “less than 2%” of the user base actually purchased a copy.
    Faced with the crisis, Murdock approached several companies with whom Progeny had been having “long-
    term business development conversations” and asked them for business.

    While agreeing that the request did not
    put him in the best bargaining position, Murdock added: “It’s not as bad as it sounds, because you don’t really have much choice. But it’s definitely humbling. For months
    you’ve been trying to puff your chest out and impress your peers at all these other companies, and suddenly you
    have to go then hat in hand and say, ‘We’re facing a crisis and we need your help.’ ”

    Fortunately, several of the companies responded favorably. They included Hewlett-Packard Co., for whom Progeny began
    developing a Debian port to the Itanium platform, as well as working on other Debian-related products. By August 2001,
    the company was actually showing a profit for the first time.

    The immediate crisis weathered, Progeny began the long process of reinventing itself. In October 2001, the
    company announced the end of Progeny Debian. A month later, the end of Linux NOW development was officially
    announced. These announcements, made after a long silence, might have seemed as if the company continued to
    be in trouble. In fact, Progeny was becoming a custom software developer, selling services to other companies rather
    than marketing products to the public.

    Not that Progeny Debian was a failure in the end, Murdock hurries to add. Admittedly, the product failed in the stores.
    However, the simple fact that the company had built the distribution provided proof that it understood Debian and
    could develop a product that would be downloaded by thousands of people. If Progeny had not developed Progeny
    Debian, he now believes, then the company would have had no tangible proof of its skills to secure new custom
    development contracts.

    Throughout the rest of 2001 and 2002, Progeny continued to reinvent itself. The company expanded from its core
    Debian competency into Red Hat development as well. It also began development of its
    platform services
    an effort to reduce the time to market of custom distributions for its clients. In keeping
    with the company’s policy of maintaining its FOSS credentials, most of the components of the company’s
    platform services have been open-sourced, including work on
    discover, a tool that gives the new
    Debian installer automatic hardware detection.

    During this same period, the company expanded slowly. “We would only hire someone,” Murdock says, “if we had
    work for them.” Murdock himself returned to the payroll, this time as chief technical officer. Although as a
    company founder, he continues to be involved with business strategy, Murdock found himself more comfortable
    and competent in a technical role on a daily basis. Talking about the change, Murdock says, “You really have to find
    the one thing you’re good at and that you like doing. And then find people who are good at things that you don’t
    want to do or can’t.”

    In keeping with this philosophy, in October 2002, Murdock completed the transformation of Progeny by hiring his mentor
    Garth Dickey as Chief Executive Officer. Dickey had been one of the VCs whom Murdock approached for funding. Now,
    Murdock likes to say: “I didn’t get his money, but I got something better — I got him.”

    Since then, the company has broke even or made a small profit each month. The profits have been reinvested in the
    company, and Progeny is as large as it has ever been. If the company is not the runaway
    success that every startup of the dot-com era dreamed about, it is something almost as rare — a survivor


    Parting advice

    Asked what advice he would give to executives of new companies entering the FOSS market, Murdock responds
    with three suggestions.

  • First, remembering how Progeny turned down funding when it was offered, Murdock advises, “Don’t turn down
    money.” Referring to VCs’ continued caution about investing in tech companies in general and FOSS-based companies
    in particular, he adds, “I don’t think you have to say that much any more.”

  • Second, Murdock cautions against entrepreneurs trying to do themselves. Murdock agrees that the Type A people
    who are likely to found companies are used to a hands-on approach to business. However, he concludes that this
    attitude is impractical in a growing company. Instead, he advises, “Figure out what you do well, like to do, what no
    one else can do. Then surround yourself with good people to complement you.”

  • Most importantly, Murdock says, “Spend a lot of time putting together a good business plan.” A solid business plan,
    of course, is a basic necessity in today’s conservative business climate for companies looking for funding.
    However, for Murdock, a business plan is also a
    guide to running the company on a daily basis. A business plan, he says, “Is just as much about getting it all
    down, looking for flaws and ideas, communicating the ideas to the people you’re going to hire, and serving as
    a rationale as to why you do things and don’t do things. I can’t emphasize that enough.”

    To achieve these ends, he adds, a business plan has to be continually rewritten. Looking back at his own experience,
    he explains, “You have to be flexible and you have to be fast. The world around you can change.”

    Finally, a business plan should have the goal of developing a revenue stream as quickly as possible. To achieve this
    goal, Murdock advises, “Keep it simple. Don’t try to boil the ocean. Have a business plan that is achievable without
    enormous sums of money.”

    Summing up what he has learned over the last five years, Murdock concludes, “I’m more practical than I was five years ago.”

    Bruce Byfield was a Progeny employee from May 2000 to May 2001. He is currently freelance course designer
    and instructor and a technical journalist.

  • Category:

    • Free Software