New Cringely
Bob's new article goes into the newest company he wishes he could make chic and show how influential he is in the computer business. But what he said about startups in general was particularly interesting, and even more interesting when he responded to a question I sent him - are more startup companies better off as research ventures instead of incorporated ventures?
For example, take the effect of transaction costs on a company - when transaction costs are high in any one market, it is extremely hard to sell a product. When transaction costs are low, products can be sold easily, but the market forces decreasing differentiation (specificity) and thus profits diminish. A transaction is one in which there is mutual profit for both parties.
So much VC money was wasted with the .Com rush that it is actually disgusting some of the stories that came from it. Bob claims, quite rightly, that when VC money gets tight, a lot of startups die. For example, he points to Israel, where terrorist bombings are presenting such a great risk that VCs are not travelling there out of fear and thus, capital is drying up and companies are dying.
Bob attempts to save one Israeli company by pointing out the cool technology they have. But why should one company be saved, when it could not survive on its own? Obviously, we should not let terrorists get in the way of capitalism, but when is a company a company, and a company a research venture?
A research venture is a technology under development that has no viability as a product on the open market. Transaction costs are so high, or a market does not exist, that the technology would not sell. In hindsight, so many of the dot com startups looked the part.
Bob's statement that when money dries up, startups die points exactly to the problem - many of these ventures are developing new technologies on VC money exclusively, and are too far away from first revenues. Bob has a rule that seems to point to this - people tend to over-estimate change in the short term. Yet he doesn't see how this affects companies - companies estimate time-to-revenue far too soon, so they are using up all their VC money.
Instead, why are these companies not research projects? I asked Bob that very question - he responded that the people in these startups want to get rich. Well, duh. But the number of startups that have ever made their founders rich are as populous as the number of artists that have ever got rich, or the number of aspiring actors who have ever become famous - the number is incredibly small.
With that in mind, it seems entirely unrealistic that these companies should continue to do what they do, and VCs get suckered into spending money because they ignore the lesson they should have engraved on their brain - people over-estimate how fast potential customers will change from potential to signed.
So these companies are actually better off as research ventures. They should step back, multiply their time-to-revenue by three, and re-work their business plans to see if they can last that long. If they cannot last that long on prudent spending and sustained development, then they should not be a company, they should be a research venture.
THAT alone will reduce the number of ventures by a good order of magnitude, and in turn reduce failure rates of startups, and increase success rates of research projects. The concept is Simmer, not Boil. Once research is complete and a workable product or service is ready, THAT is the time for venture capital to hire, market, and present a viable product or service to the market.
Comments
starting up
Having been involved with 5 startups and met many more, and talked to probably hundreds of VCs & bankers, I'd have to say this is unrealistic. A "research venture" is an oxymoron, the only way they can exist is as playthings of the rich (Allen, Woz, Ellison etc).
I've seen lots of companies which don't deserve to survive, prolonging their agony at simmer is always a mistake.
Research Ventures in Academia
Venture is perhaps the wrong word, but there certainly is a place for research-oriented work. Mmmm...perhaps still a wrong mix of words. As Greg points out, these entities can't exist in the real world without some kind of philanthropy or external sources of funding.
It would perhaps make more sense to incubate in the sheltered world of academia, sucking on research grants and/or funding by corporate sponsors. This is, after all, how Google got started. Just make it clear that it is, indeed, research, until such time as you leave the academic womb and turn it into a business. After that, you're on your own, and survival is your problem.
Funding too early, then.
Perhaps the best thing to say is that too many of these companies get funding in the early stages, or the funding comes too much from the people running these startups. Since funding is a first, crucial test of the viability of a new technology or service, self-funding a venture can serve to obscure the market viability of a new startup's technology.
As for the term "research venture", the semantics do work if you look at it from an academic view of entrepreneurship. Or, at least, from the school I come from. Everything is a venture, from that viewpoint. Research efforts are ventures of their own - the difference comes down to the focus, which could be on bringing a product to market, or simply advancing some area of science or technology. I'm thinking of research strictly in the sense of researching a technology to bring to market; in that definition, all startups are research ventures, researching and/or developing products to bring to market.