How A 29-Year-Old Chemist Proved That Even In England Startups Can Thrive

Keith Davies has made Chemical Design a world leader in molecular modeling LONDON—By now we’ve all seen the numbers and heard the gloomy forecasts. Science in the United Kingdom is suffering from a dearth of funding, incentives, and political clout. The lack of commitment to R&D on the part of both the private and public sectors in Britain has sparked an alarming emigration of scientists and high-tech entrepreneurs. But for every rule there is an exception, and Keith Davies is the man to prove that even in luck, England, the age-old recipe for entrepreneurial success remains a constant. With a good idea, a little funding, a lot of commitment, and an indeterminate amount of luck, the “little guy”—yes, even a chemist—can build a company and make a mark in the world. Davies, just 29 years old,is CEO and founder of Chemical Design Ltd., the small company that in five short years has progressed from a one-man startup—funded with $145 of his own money—to a worldwide leader in molecular modeling software. With annual sales approaching $8 million and 270 installations throughout Europe and the United States—including in its list of clients Exxon, Du Pont, Glaxo, and Abbott Laboratories—Chemical Design has soared to the forefront of its field. competition coming from only four major competitors--all of them U.S. startups (see table on page 9). Chemical Design’s business is to write, produce, and support software that allows scientists to create models of molecules ranging up to 8,000 atoms. In many instances, it installs software-hardware packages: intelligent workstations that have been customized to the molecular-modeling task. These work- stations drive specialized subprograms that enable the scientist to do more than just create images of complex molecular structures; they can actually help the scientist form hypotheses concerning the dynamics of the molecules' action. For example, one subprogram can help determine how readily a novel medicine might be absorbed by the body by testing the potential reactions the molecule could engage in. The applications of such programs are so broad that chemists tend to see them as opening a new era in labwork. Only a decade ago a chemist had to construct stick-and-ball models to “see” the molecules he or she wanted to study. In fact in 1976, when Davies entered St. Peter’s College of Oxford University to study chemistry, although the notion of modeling on a computer was beginning to be developed, he had yet to hear of the concept. “I got into computing simply because I found it interesting," says Davies, who stresses that he didn’t enter into his research with the objective of starting a company or making a profit. “But then I found that I could put both chemistry and computing together and have a business opportunity, and off we went.” Well, perhaps it wasn’t quite that easy. Davies first got involved in writing modeling software during his fourth year at Oxford. which is reserved exclusively for laboratory research. He was charged with the task of designing software that could build molecular models more accurate than those devised through the traditional method of crystallography. The models that he initially created were of H2 antagonists (the active compounds in ulcer treatments) for Smith Kline & French Laboratories. Corporate scientists had already done the crystallography by conventional means but were planning to experiment with molecular modeling on-screen as part of a program to develop a second generation ulcer drug—one that was to follow up on Smith Kline’s Tagamet. It was the algorithms that Davies wrote for this project that laid the foundation for Chem-X software. “Within 12 months of beginning the research,” says Davies, “it became quite clear, through the recaption our poster presentations and brief talks were getting at U.K. conferences, that [the software] was of great interest to pharmaceutical companies. I knew that it was an up-and-coming area. The question was: Was it a business?” By this time, Davies was working on his Ph.D. and held a junior research fellowship (a three-year, nontenure post) at Lincoln College, Oxford—but the commercial potential of the work became increas- ingly enticing. He took a sabbatical and traveled to the U.S. “on the grounds that whilst Oxford was a good place to be, I didn’t think it was a good idea not to see bow other people were doing things.” He spent four months in Bob Langridge’s lab at the University of California, San Francisco, a recog- nized leader in the computer modeling field. Davies also visited many of the leading pharmaceutical companies—Abbott, Merck, Exxon, and DuPont—and became increasingly convinced that he was sitting on a potential gold mine. Many of the companies had begun to develop their own software in house, but in most cases their programs were problem-specific, rather than of use in generic modeling. What’s more, while the computer groups that developed the software could use the programs, they were of little use to the chemists at the bench, who were not as familiar with computers. The time was ripe for a user-friendly program such as Chem-X. “Keith had an advantage in that he started early in this game,” says Charles Hutchins, a theoretical chemist in the computer modeling group at Abbott Labs and current head of Chemical Design’s U.S. user group. “At that time [1982] few of the companies had any software at all, so when something was offered that was very cheap and relatively useful, he was able to make a big start right away.” John Newsam, a staff chemist in the X-ray scattering group at Exxon, concurs that timing was a key to Davies’ success, but also points out that his program filled a niche that was not being addressed by the other software available at that time. “It was quite clear that Chem-X was the best choice, especially for work with inorganic mol- ecules. The other companies were applicable only in the region of organic macromolecules, but Exxon was interested in applying the modeling software to a variety of types of molecules.” Newsam explains that his group uses the software to investigate molecules such as zeolites, compounds that are important to the petrochemical industry as catalysts and solvents. “We know the molecules work, but don’t always know exactly how or what alterations might make them more efficient. The computer enables us to take an Edisonian approach to the problem take a molecule that already works and have the computer tell us what the optimal structure would be, then actually tinker with the molecule to optimize the process.” In the fall of 1982, even before Chemical Design had been incorporated, Newsam met with Davies and was responsible for Exxon’s being the first U.S. installation site for Chem-X software. It was on this wave of success that, in February 1983, Davies returned to Oxford and established his corporation. Though he initially planned to run the company concurrently with his work at the university, Chemical Design’s activity soon became more than the academic suppqrt structure could bear. On a few memorable occasions after major conferences, calls for Davies jammed the university switchboard, “And you can imagine that the administration got a little bored with that,” he remarks. Oxford now has direct dialing. Davies explains that in those days the university was not as commercial as it is now and had no staff to look after the development of work originated in its laboratories. Also, as he had never been employed by the university directly (only by Lincoln College and indirectly by the Science and Engineering Research Council, which had funded some of his work), it had no claim on his work. “It really wasn’t very interested in taking on a piece of software that only I knew anything about. In that light, they felt it was better for me to exploit it without the university being involved.” That decision on the university’s part gave Davies his break. For, upon paying a small royalty to the National Research Development Corporation, which had funded his work, Davies found himself in possession of some intellectual property that was to have considerable initial value, and even greater longterm potential. At that point Chemical Design began doing business in earnest. Davies’ initial investment of $145 covered the costs of forming the company and getting some licenses signed. Though he didn’t open a corporate office until late in 1984, operating out of his home that first year, within 18 months of incorporation Chemical Design had received orders from Beecham, G.D. Searle, the Welicome Foundation, and Pfizer in the U.K., and American Cyanamid, Squibb, DuPont, and Amoco in the U.S. “We didn’t buy much [equipment that was new,” Davies recalls of those lean, early days. “What we did spend on was literature and administration so that the company ran efficiently. But the real secret was that we had customers out there paying real money, in some cases large amounts of money. By the time DEC [Digital Equipment Corp.] could deliver the VAX [computer] we wanted [in late 1984], we could pay for it in cash.” Davies did consider turning to venture capital for startup funding, but the companies he talked to were too greedy. They wanted over 30% of the company for $145,000 cash, which on the basis of the business plan that existed then was unreasonable, Davies says. “Perhaps if they hadn’t asked for quite so much of the company, I’d have taken one of the offers of funding, which would have probably accelerated the initial growth a little bit.” But looking back, he does not regret this decision, for 18 months after the start, sales had exceeded $350,000 and Chemical Design was realizing over 10% profit. The following year, 1985, Chemical Design opened corporate offices in Mahwah, N.J., to handle the increasing U.S. business, and incorporated hardware into its line of products. And this year Chemical Design’s earnings are projected to top $7.3 million—a figure that Davies says it would exceed were it not for constraints imposed by hardware delays. IN the meantime, Davies certainly isn’t sitting on his laurels— or his profits. Chemical Design’s staff is currently expanding ChemX to handle molecules up to 231 atoms in size—a tremendous leap from the current 8,000-atom capabilty. Also, this spring the cor- poration purchased 95% of Expert Systems International Ltd., a software house specializing in artificial intelligence. Teaming these products with Chem-X will render the system of even more use to bench chemists by providing advice on what modeling strategies to adopt, warning the user about pitfalls, and automating a lot of calculations. Still, not everyone is as sanguine about the company’s future. Exxon’s Newsam predicts that competition will drastically increase in the field, as universities begin to command a more direct role in the commercial use of their work. (He points to the fact that most of the important advances in computer modeling are still coming from academic settings.) Moreover, many chemists and modelers predict that the next critical step for the field will be to develop energy calculations, an important factor in understanding and simulating molecular behavior accurately. “Exxon wants to get a leg up by being involved in the developmental stage of these new advances,” says Newsam. “And I don’t see Chemical Design as being at the forefront of this work.” Davies acknowledges that his share of the large protein modeling market in the U.S. is only about 10%—not as large as he would like. But he stresses Chemical Design’s supremacy in the small, inorganic market, and points to the broad. stable base—in both application and geographic terms—upon which he has built his company. And, though he doesn’t like to admit it, he does see his success as somewhat of an anomaly in his country. “The press believes that the U.K: has failed to market things correctly, and I’m afraid I agree with that. My success is undoubtedly unusual in the U.K.” And undoubtedly unusual for a 29-year-old. John Stansell is editor of the Innovation Section of the Sunday Times in London.

How A 29-Year-Old Chemist Proved That Even In England Startups Can Thrive

Interested in reading more?

Become a Member of