| ||||
Excerpted from the book Done Deals, edited by Udayan Gupta, Harvard Business School Press
Technology has always been the cornerstone of venture capital investing, and university research and national laboratories such as Argonne, Ames, and Los Alamos have long been the source of those technological innovations. Indeed, for universities such as Harvard, MIT, Stanford, and the University of California, licensing technology has been a very lucrative source of revenue. Still, for all the value resident in these institutions, few venture funds have been formed to specifically mine and commercialize their technology.
Steve Lazarus founded ARCH Venture Partners in 1986 to transfer technology from the University of Chicago and Argonne National Laboratories, one of the nation's twelve national energy laboratories. Since that first $9 million fund, ARCH has expanded — both in terms of managed capital and in terms of its geographical reach.
As a fund that started in Chicago and now has partners in Seattle, New York, Austin, and Albuquerque, ARCH is attempting to institutionalize the process of technology transfer. With successes such as NEON, a developer and marketer of software for business enterprise applications, and Everyday Learning Corporation, a provider of educational material for mathematics classrooms, it has proven that seed investing, especially in the area of untested and unproven research technology, can succeed. But ARCH also has shown that such investing requires building bridges with other venture capitalists and research organizations, as well as a commitment and a passion that sometimes go beyond traditional venture capital practice.
| The 57th Street Irregulars |
| Mining the Corporation |
| A Success Out of Left Field |
| The Biggest Success to Date | Still Collegial |
| Spreading Out |
| Breaking from the Past |
| Does a Seed Fund Mean Staying Embryonic? |
So much of venture capital begins from a university perspective. Prior to World War II there was no substantial level of organized university-based research, not in the physical sciences nor in the life sciences. In the life sciences, you only had the advent of the sulfa drugs in the '30s. World War II was a watershed of enormous proportion. After World War II came the founding of American Research and Development (ARD) headed by General Georges Doriot, based at Harvard. While Professor Fred Terman had been at Stanford for some time before that, venture investing began broadly only after World War II. When William Shockley, co-inventor of the transistor, moved from the east to Palo Alto to start Shockley Electronics, the foundation was laid for a transistor-based electronics industry in northern California.
The University of Chicago evolved during the '50s and '60s as a research university doing work in both the physical and life sciences — probably at about the twentieth research dollar position in the United States. It didn't have an engineering school, so what actually occurred there over the years were a number of covert or hidden engineering organizations.
Chicago was not a Hopkins, a Columbia, a Harvard, or an MIT, but it was respectable. However, there was no organized technology transfer. The first reason for this was that there was the usual ambivalence within the faculty about doing anything that diverted from pure research or that had the flavor of a profit motive. There were arguments about this in the senate of the faculty. But one extremely important decision had been made — that the university and the faculty took ownership of all discoveries that were made by individual faculty members. Even today in 2000 that concept is still in dispute on many campuses. But it was clearly settled at the University of Chicago. The second thing that was going on in the late '70s and early '80s was a growing concern in Congress that the nation was not getting a payoff for all the investment that went into national laboratories and research universities. There are about 700 national laboratories, with the largest concentration in defense and energy. Many of the energy laboratories for decades had been operated by management designates such as the University of Chicago. The University of Chicago managed the Argonne National Laboratory on behalf of the Department of Energy, with a $500 million annual budget in basic research. A lot of ideas were sitting in that laboratory, and nobody was looking at them for commercial purposes. In the early '80s, Congress passed the Stevenson Wydler Act and the amendments to the Bayh Dole Patent Act and essentially the two pieces of legislation together gave the University of Chicago the opportunity to take title at no cost to discoveries at Argonne. So, you had a faculty that was doing productive research, and a potential at the Argonne Laboratory for a lot of valuable intellectual property.
The catalyst for all this, I believe, was the fact that the drug Erythropoietin, the red blood cell stimulating factor, was synthesized by Gene Goldwasser at the University of Chicago. Nobody protected it. It went into the public domain, and ultimately Amgen cloned it and it became a billion-dollar drug. Many people at the University of Chicago wondered why Stanford and UC San Francisco could participate in the revenue stream of the Cohn Boyer recombinant DNA inventions but Chicago got nothing as a consequence of Erythropoietin. All of these things came together in the mid-'80s, under the aegis of Walter Massey.
Walter was a high-energy particle physicist. He had been director of Argonne and was the VP of research at the University of Chicago. Later he went on to be head of the National Science Foundation and provost of the UC California system, and today is the president of Morehouse College. Walter was not only an excellent scientist, but he also had exquisite political skills. He pulled the basis for ARCH together, getting buy-in from some very powerful trustees, some important and influential people in the faculty, and some local businesspeople. He also designed ARCH as a not-for-profit corporation, wholly owned by the University of Chicago. Its main interest would not be in licensing technology, which was the conventional way of doing technology transfer at the time, but rather in starting new companies. At that time, I was retiring from a Chicago-based company called Baxter Laboratories, and I was thinking about teaching. Walter and the fellow architects of ARCH designed a job that was simultaneously a president of a corporation, albeit a small one, and associate dean of the Business School at the University of Chicago. So I came down to Hyde Park from Deerfield, and found myself the first employee of ARCH. I had a very small budget. We had the responsibility to do all patenting for both the university and the laboratory from that budget, which was enough to sustain me and a secretary, but not much else. We had a substantial amount of technology to examine.
The placement in the Business School turned out to be extremely fortunate. The student body was made up of a large number of young people who had been out in the workforce and who decided that they did not want to continue as salary men — in the Japanese sense of the phrase — but wanted to become owners. They wanted to build net worth; they wanted to control their destiny. Therefore, they wanted to go into small new enterprise, and build up from there. Several of them found me, and asked, "Is there any way we can work with you?" I said, "I can't pay you." They said, "That's okay, we'll do it voluntarily." Thus began what we first called the 57th Street Irregulars, and later — when we decided people needed to take us more seriously — the Group of ARCH Associates. I think it's coincidental, but very fortunate for me, that the two earliest members of that group, Bob Nelsen and Keith Crandell, are today my partners in ARCH Venture Partners. Clint Bybee, who came a year or two later, is the fourth general partner of ARCH Venture Partners.What the young people started to do was go into the halls of the institution and locate the stars. I realize that's an elitist comment, but it's just like salespeople qualifying sales. Who is most likely to buy? Who is most likely to be productive? There are a lot of fairly mundane ways of finding out. Who has the most grant money? Who has the most publications? Who has the most citations? Who are members of the national institutions, the scientific elite structures who elect people? And finally, who did other scientists admire? Out of that, one could narrow it down to the 10 percent of the research faculty who were likely to yield inventions that would have intrinsic, and ultimately extrinsic, economic worth. Over time we have continued with that technique and refined it. This form of triangulation truly works.
I think the second technique we evolved was learning how to first identify the unique invention that was not going to have much follow-on, and that was best licensed to a third party. After syndicating out the licensing candidates, we concentrated on the technologies and inventions that could be platforms for new companies. When we licensed an invention there was no certainty that a revenue stream would result. For example, all too often a molecule licensed to a Bristol Myers, or a Lilly, might engender initial interest, but over time would be displaced by interest in other targets. It would then stay on the shelf. Which was another reason to start creating companies over which you have a greater degree of control. There was, however, very little venture capital in the Chicago area. I'm talking specifically of seed and early-stage high-risk capital with which to start new companies.
We made several trips to both coasts, carrying our portfolio of technology. Everyone was interested in a number of the specific deals but nobody had any enthusiasm about getting on an airplane and flying to the middle of the continent to shepherd, nurture, and incubate an early-stage deal. This is one of the demanding characteristics of seed and early-stage investing.
So, quixotically, we set out to raise our first venture fund. That was in '87 and '88. I made something in excess of 100 visits to foundations, other venture funds, and university endowment people. Everybody thought that the idea had merit. Nobody was interested in investing. I was told over and over again that my track record was all ahead of me. There was no denying that — it was the absolute truth. There's a Catch-22 to getting started in venture capital. If you don't have experience it's hard to get started, and if you haven't gotten started, you don't have any experience. Fortunately, a friend suggested that I call on Jim Bates, then the vice chairman and chief investment officer of State Farm. So, on a Saturday I drove out to meet Jim, who had been playing tennis, at the International House of Pancakes for lunch. I talked without interruption for about an hour, at the end of which he said, "That sounds like an interesting idea, we'll put $4 million into it." Suddenly I had my first investor.
The university endowment matched that, and soon we had a $9 million fund. We invested that fund in twelve companies. We made horrendous mistakes in some of those investments. Five of them failed. But seven of them were successful, and a couple of them were quite successful. We had a first fund that was returning a respectable return to its investors, and we were now considered legitimate venture capitalists. Not experienced, but legitimate. The university examined that set of occurrences and said, "This is no longer an experiment. We as a university are not at all comfortable with continuing as a general partner of a venture fund, so why don't we divide the entity we have created? The technology transfer and commercialization organization will remain inside the university, with the venture capital partnership going outside." At that point, in 1992, Keith Crandell, Bob Nelsen, Clint Bybee, and I stepped outside the university, created ARCH Venture Partners, and started to raise ARCH Venture Fund Two. We still had friendly relations with the university, and an opportunity to have early examination at both the University of Chicago and the Argonne National Labs. About that same time, we were invited by Columbia University in New York to take our model and open an office at the University in New York City. Some commercial interests in Albuquerque, New Mexico, helped us take the model and open an office in Albuquerque that was focused on the two huge national laboratories there, Sandia and Los Alamos. Those two labs represented $2 billion worth of basic research every year and had never seen a venture capitalist. Clint Bybee went to Albuquerque. We hired a young M.B.A. to operate in New York with Keith Crandell frequently coming into New York to supervise the operation. Bob Nelsen, who was born and raised in the Washington area, relocated to Seattle to open an office, on the assumption that the latent technology in Seattle — physical sciences, information sciences, and life sciences — was largely unexplored. There were very few venture funds in Seattle, and therefore we had an opportunity at the University of Washington to apply our same set of techniques.
ARCH Venture Fund Two proved to be an extremely successful fund. We created twenty-two companies. Six of them were follow-on investments in Fund One companies. We realized that what we were becoming was a seed and early-stage venture fund that also did later-stage investing in companies of its own origin. This had two consequences. One, it reduced the overall level of risk in the portfolio. Two, it started to shorten the time to recovery of the investment.
We started to be written about. Other universities heard that we were active, realized we knew how to interact with universities, and invited us to come out and look at their technology. We started to do a lot of flying. We violated our early rule that you had to drive to the deal. What we found we could achieve was an operating relationship either with the technology transfer function in a university or with a local, small venture capital fund. For example, at the University of Michigan, we worked with the Enterprise Development Fund, which is located in Ann Arbor. In Boulder, Colorado, we worked with Boulder Ventures. We do that today in more and more locations around the country. We found there were certain other funds who had similar characteristics to ARCH — some of whom had been around longer than we had.
After focusing on university and laboratory technology primarily, we started to see opportunities in technologies that had been developed by corporations but were not being pursued by those corporations. In Fund Three, which is a $107 million fund, we dealt with several companies that are spin-offs of technology from company research and development locations. The techniques are quite similar to doing university research. We still take pure technology, act as entrepreneur or general manager — until one can be recruited in — and then nurture and support that concentration of technology. That appears to be a portfolio-expanding idea, getting away from just purely doing universities and laboratories, and having a different set of opportunities that balance the deal flow.
We tried to come to grips with the question of specialization versus generalization. It was being argued in venture capital circles that technology was becoming so sophisticated and so granular that you had to restrict yourself to life science or information technology, for example, or sometimes a subset of information technology. Because we were casting our net so broadly at a university, or in an industry, we felt that that specialization would be a mistake for us. What we decided was that the general partners should become reasonably knowledgeable in one or more technological areas. Then we would hire a group of very, very sharp consultants — Ph.D.-level academics or practitioners who understood the subsets of technology very, very well — and then pair the consultants and the partners together while working the deals. We felt that this was a successful approach to the generalization versus specialization problem.
In Fund Two, a $31 million fund, we recognized that we were sub-optimizing our capability for return by only raising the fund levels that were traditionally associated with seed and early-stage investing. We would do the start-up investment in the fund, the seed investment. We would invest substantially in the first round, but when the second, third, and fourth rounds were required, we would take an increasingly smaller position, and turn the control investment position over to a larger fund like Venrock. Essentially, that meant that we were doing the heavy lifting, and then inviting the larger fund in after a great deal of the risk had been washed out of the project. Although the price was somewhat higher, it wasn't all that much higher. So giving Venrock — who were great investing partners — that kind of a position was ceding to them much of the value of the deal. We determined when raising our third and subsequent funds that we would raise enough money to allow us to invest proportionately through all rounds to liquidity. We have 25 investments in both Fund Three and Fund Four, and these larger funds have enabled us not only to remain true to our seed and early-stage investment approach, but also to maintain a position that allows us to have somewhere in the vicinity of a 20 percent ownership by the time a deal goes to liquidity, either through an IPO or through a sale to a larger entity.
There was, in the mid-'80s, a project going on at the University of Chicago managed by a combination of the Department of Education and the Department of Mathematics. When you set yourself up as technology investors, you don't normally think in terms of going to the mathematics and education departments looking for a core, or a seed of a new corporation. But within this project, which was aimed at reforming the mathematics curriculum in the United States for grades kindergarten through six, there were materials being crudely published. These materials were being made available through what they called summer boot camps to mathematics coordinators all over the country. There could be twenty large urban centers represented on the campus of the University of Chicago for three or four weeks every summer, and they would want these materials. The project people were trying to operate out of the basement of the Department of Education, copying these materials and sending them out for cost. Finally, one of the professors in this project asked Bob Nelsen, then one of the 57th Street Irregulars, if there was any way the business school could help bring order to the chaos.
A couple of our young people went over and examined what was going on, and came back and told us there was money to be made. These people received royalties of $40,000 operating out of a closet last year. Furthermore, the product was getting publishing industry kinds of margins, it was a disposable product, and the essentials were there for a business. So, we organized a business, first staffed with four students, and started to recruit for a CEO. I should point out that this was before we had raised the venture fund. We were recruiting for a CEO without the venture money in hand. We were also extraordinarily lucky in our hire, because Joanne Schiller, the CEO of that company, proved to be a superb manager. We put in a little more than $250,000, and the company never needed another venture dollar. It broke even early in its second year of operation. The product was renewed each year. It actually violated all tenets of academic publication. It only produced a grade at a time, rather than spending $20 million and developing an entire curriculum and publishing it at one time as Simon & Schuster might do. But there was a great niche — the reform-minded market in the education community — that loved the product.
At the end of about five and a half years, that company was sold to the Chicago Tribune for $26 million. Joanne Schiller is today a group executive at the Tribune. ARCH's return on its $250,000 investment was something like a twenty-two multiple. The university, which had substantial ownership for having licensed that product to the company, got a return in excess of $10 million. It was an enormous success. We had started from a situation where the university had realized perhaps $200,000 each year for commercialization of products. Now they were seeing seven-figure numbers.
Our biggest success to date has been New Era of Networks Inc. (NEON), a company for which we help to bring the technology and the entrepreneurs together.
Rick Adam founded NEON, an Internet infrastructure company. Even though he had come out of the financial services industry, he thought the great need was in health care. That turned out to be a mistake. Health care was not mature enough or ready to buy. But because Rick knew the financial services marketplace, he found very quickly that there was a demand for the product, shifted resources almost on a dime, and started to penetrate the financial services market. That propelled NEON's great success. It's either knowing where your market demand driver will be when you go in with your product, or having the background and the knowledge to sense where it's going to come from. If you don't go there first, you go there second. So Rick had a track record, deep market experience, and experience in having supervised some people before, having done a lot of hiring and some firing. That's helpful, because you're going to have to fire some people in these small companies. You can't afford to hesitate and wait until they really do damage. If you've got a vice president of marketing or sales who is not penetrating with your product, you cannot afford to stay with that person very long. The general manager has to be able to make that change. We've learned some other lessons. We're going more toward formal executive searches, where the executive search companies have exhaustive records on people. Modern techniques of interviewing, a lot of reference checking, and exhaustive due diligence can go a long way toward making ideas bear fruit.
We're a highly collaborative organization. We see that as an element of great strength. We are a group of people who have now worked together intensely for 14 years, and while we don't think homogeneously, we have great respect for each other's judgment, and so are very supportive of each other. I think a 14-year partnership is becoming increasingly rare in this day and age, as people spin out from venture partnerships to form new partnerships. We're still pretty comfortable with each other, and that seems to be an important factor in our fund-raising efforts. Another part of our methodology is that although seed investing is conventionally considered enormously high-risk investing, we invest at the beginning with a teaspoon or an eyedropper. We are adherents of Tom Perkins's creed of investing early to wash a particular kind of risk out of an opportunity. I was very influenced by his story of how he invested in Genentech at the beginning. He put small packages of money in scientific institutions to replicate the Cohn Boyer science, and when he saw it was replicable — when he felt that the major risk had been put aside — he could confidently invest larger amounts. We try to do that all the time. We invest small amounts to achieve specific milestones. We cooperate in helping those milestones to be achieved, but we do not invest $5 or $10 million at the outset. Our level of investment in our companies, in the aggregate, is going to be somewhere between $7 and $15 million in each company. But we do that over several rounds. That's another part of our style.
We've found, particularly in Seattle — which has proven to be a very productive location for us — that there are underserved locations that have companies emerging in a number of disciplines, including information sciences, as you might imagine, the life sciences, and the Internet in particular. We have taken advantage of that position. Now we are working out of Albuquerque. We're journeying not too far to Texas. We've done three investments in Austin, one together with the local reigning Austin venture partnership, Austin Ventures. We mentioned Ann Arbor and Boulder before. Denver is a third. We have done a company in Ames, Iowa, where I think we are one of maybe two venture funds that have invested in that area over the last thirty years.
We also have refined a technique that we call the technology roll-up. One of the reasons that biotech investing is so difficult is that early in the game one tended to make a bet on a single molecule. In order to develop a single molecule to a product, one goes through not only the FDA gauntlet, but the general research gauntlet, which could fail at any stage. A lot of money has been lost, and one of the reasons that biotech has been slow to return its investment is because of that early style of biotech investing. What we learned to do, once we saw an interesting molecule, was range around the country to find corresponding, associative, or complementary molecules that we could bring into the same company. So, it was a roll-up in the sense of not only beginning with science at the University of Chicago, but pulling in complementary science from the University of Indiana or the University of Pennsylvania or one of the national labs. One of our companies, a microfluidics company called Caliper, has technology from nine different institutions in it. (Caliper had a very successful IPO in December 1999.) The roll-up also preempts a competitive response to your idea if you have all the relevant technology in your company. The other strategy in doing this is if you have early-stage inventions — or in the case of biotech, molecules — that are probably looking at six, seven, or eight years to move through the laboratory and testing, you want to match that with later-stage molecules in the same general discipline that are in the clinic, or at least close to the clinic. That's part of the roll-up strategy as well. All our biotech investments have this characteristic.
In the early days, we did the jobs ourselves. That proved to be too onerous a demand. You could not be acting CEO of even two companies at the same time. It was more than a seven-day, twenty-four-hours-a-day job. So we learned to recruit earlier and earlier. Our recruiting has had the same characteristic over the fourteen years we've been doing this. That is, we've been right more often than we've been wrong, but not by much. The old clich© truly pertains — the best predictor of success is track record. If you recruit a CEO or acting entrepreneur who has no relevant experience, you are taking a very long chance. I don't think that has changed very much in the fourteen years I've seen this, though I think we have become a lot better at finding people.
The first trick is to find a natural affinity, as illustrated by Joanne Schiller and Everyday Learning. Thinking back to our early interviews with Joanne Schiller, nothing in those interviews would have predicted the rich level of success that she ultimately had. She had not done it before, but she had experience in the industry. She knew the marketplace. Her personality and her stamina just fit what had to be done in that company. I hesitate to call it luck, but there was certainly a lucky aspect to the marriage of Joanne and that company. Apart from a track record, probably the most important factor is really knowing the marketplace. Not just having the Rolodex of names of the people who are in the marketplace, but really understanding what aspect of the marketplace would become a driver for your product. This is critical to long-range planning and positioning of a company.
Does a Seed Fund Mean Staying Embryonic?
I think the classic venture investing pool of capital is a hybrid. You start a company and you stay with that company to the point of liquidity. You do not hand off the entire destiny of your investment to a IVP or a Chase. Rather, you remain in position, not only to realize the level of return that allows you to go back and raise the next fund, but because you want to stay in control of the deal. Everything you've learned about that business at the beginning becomes more and more valid and valuable as the company matures. You don't want to put its destiny in the hands of someone whose background may be in an investment bank, because it's a different style of thought than the development of enterprise. I think that eventually bankers are useful, but not in driving the project. So I believe that the best model for people with personalities like ours is a fund of a size that allows you to do a certain number of investments and stay in those investments to liquidity, sale, or disposition of some sort.
Now, what's the right number when you're starting companies all the time? When I was going around the circuit and trying to learn what the masters knew about this business, Tom Perkins said that he felt that the start-up bandwidth was one, maybe one and a half companies at a time, and probably that meant one to one and a half a year. You could probably have ideas simmering on the back burner. Think in terms of a guild: If you have some really good journeymen, and some clever apprentices, the master can leverage himself or herself beyond the one and a half to some degree. But I think it's still a number that has merit, in order to calculate how many early-stage deals can you really start in a given time period.
Ultimately, we have an exploration mentality. We think the land is out there, the way is uncharted, but the reason we leave the port and sail uncharted waters is because of the belief that there is an undiscovered continent or a route to the Indies. There used to be very few real seed and very early-stage investment funds, but now the enormous and unique phenomenon of the Internet space has midwifed a gold rush generation of seed investors into existence. We're glad to have the company.
· · · ·
[ Buy the book ]
Related stories on HBSWK:
Done Deals