Musings on Entrepreneurship and Innovation
Sunday, May 20, 2007The Cost of Failure Falling...Success Remains Elusive
William Goldman, the playwright and author, once said of Hollywood, "Nobody knows nothing." A little farther up the California coast, some Silicon Valley veterans are publicly admitting as much about the hit-and-miss nature of Web 2.0 startups. In a recent profile in The Wall Street Journal [1] technology evangelist Guy Kawasaki notes:
If you raise $2 million from VCs, you have to pretend like you "know" all this stuff. The truth is whether it's $12,000 or $2 million, you really don't know. The only difference is what you think you can admit.
Quoted in Business 2.0 [2], Blogger founder Evan Williams chimes in:
More and more of us are becoming aware that you don't necessarily know what is going to be successful.
Why the sudden outbreak of candor? For starters, there seems to be a growing realization that the returns to a Web 2.0 company are wildly uncertain. Investor and entrepreneur Naval Ravikant asserts:
The Web is the most hit-driven business the world has ever seen.
Possibly, but the hit-and-miss nature of Web 2.0 companies is not unique. The market outcomes of a host of other businesses, including movies, music, and books are subject to long-tail economics, where a small number of offerings account for the bulk of revenues:
In fact, research shows that a remarkably broad range of consumer products, from food to sporting goods, can be similarly described [3].
Secondly, the cost of failure is falling. As Kawasaki puts it,
During the dot-com bubble, you needed $5 million to do stupid ideas. Now you can do stupid ideas for 12 grand.
The implication for these entrepreneurs is to predict less and experiment more. To do that, Ravikant has launched Hit Forge:
This is like a movie studio. It's about milestone-based development, piloting concepts, access to distribution...The engineers have the freedom to experiment, but they have 90 days to ship a product. The product has to grow orginically without any marketing.
Those that survive get more funding and access to distribution.
We are going to build as many as 20 companies a year. We need to find one hit to succeed. We can do that.
Possibly, but the challenges are considerable:
- It's tough to manage 20 experiments in parallel. That becomes particularly true when one or more of the experiments begin to exhibit some market traction and one must face the various challenges of scalability.
- Given confirmation and escalation bias, it is tempting to throw most of one's eggs into one or two promising baskets, which is very risky.
- On the other hand, absent substantial commitment to a product or project, it's very difficult to defend an early market lead.
Though it may be difficult to effectively manage a dynamic portfolio of real options, it beats the fool's game of making big, irreversible predictions in the face of unanalyzable uncertainty. Jim Armstrong of Clearstone Venture Partners may be right when he asserts, "The old way of building companies is broken, and [Ravikant's] way could be the future."
Nevertheless, I can't help but think of a recent interview of Ingmar Stenmark, the most successful ski racer in World Cup history. When asked about the impact of modern equipment on ski racing, he said that while skiing has gotten easier, winning is as hard as ever. Similarly, although it may be easier and cheaper than ever to play in the startup and new product games, it is not likely to become any easier to win. I'm confident that we can proactively tilt the odds in our favor, but a little (or lot) of luck is welcome.
[1] Gomes, Lee; In New Net Economy, Everyone Gets to Be Stupid for 15 Minutes; The Wall Street Journal; May 16, 2007.
[2] Copeland, Michael V.; Return of the Startup Factory; Business 2.0; May 7, 2007.
[3] Kohli, Rajeev and Sah, Raaj; Market Shares: Some Power Law Results and Observations; Revised November 21, 2003.
