Improbable != Irrational

I want to start a company.  Therefore, I must be nuts.  Right?

With high probability, whatever company I might start will fail.  Even many conceivable “successes” will result in a smaller economic return than could be obtained from an ordinary job.  Bad financial results coupled with long hours and lots of stress?  Yes, my desire to start a company seems indicative of some kind of insanity.


Just how likely am I to fail?  Scott Shane (in his book on angel investing) claims that U.S. Census Data for 1996 indicates that 510,654 new U.S. businesses were formed in 1996.  Six years later 260,970 (just over half) had completely failed.  Only 1.6% achieved revenues of at least $5M/yr.  A $5M/yr. business is something of which one could be proud, but it’s hardly the next Google.  A mere 474 (about 1-in-1000) had achieved revenues of at least $50M/yr.  Even a $50M/yr. company is only medium-sized.

Of course, I can argue that my odds are better than Shane’s numbers suggest.  Shane’s data covers all businesses, not the kind of high-tech business I would like to start.  A small local restaurant, with no plans to become a chain, cannot possibly reach revenues of $50M/yr.  So, maybe the fraction of high-tech businesses started in 2014 that will reach $50M/yr. is larger than Shane’s 1-in-1000.  My previous company, CodeSourcery, reached $8M/yr. in revenue before being acquired, so maybe I know something.  But, does it really matter?  Whether the odds are 1-in-1000, 2-in-1000, or even 1-in-100, the conclusion is the same: any company I start is likely to fail.

As Daniel Kahneman makes clear in his excellent book Thinking, Fast and Slow (stay tuned for my book review!), people tend to neglect base rate information when they think about a specific situation.  Entrepreneurs, in particular, greatly overestimate their own odds of success, even when they are reminded how many similar businesses have failed.  Paradoxically, the fact that there are so many unknowns — including unknown unknowns — contributes to entrepreneurial overconfidence.  It’s hard to foresee the competitors that are about to arise, the new technology that will eliminate the market, and the unexpected health problems of one of the co-founders.  Would-be CEOs focus on their perceived advantages (technology, business plan, founding team, etc.) and ignore the fact that most businesses fail, even though those businesses had advantages similar to theirs.  

I do not dismiss the base rate information.  I know I’m likely to fail.  Full stop.

So, is it crazy to try?

I don’t think so.

From a financial perspective, I can certainly earn a good living from a conventional job, but I make more in dividends from my boring portfolio of index funds than any sane company will pay a senior manager.  So, taking a job isn’t going to have a large impact on my standard of living.  And, while I will certainly invest capital in a new venture, I will never risk enough to put my home, retirement, or kids’ college funds in jeopardy.  On the other hand, the potential financial upside from a new business is significant.  So, by playing the game, I have a chance to win a prize which is otherwise completely unattainable.

Beyond the financial considerations, I like the idea that I will have no-one but myself to blame for what happens.  It would be a mistake to say that as a founder, I will be in control of my own destiny; the world is far too random.  But, it is true that I will not be able to complain that someone else made it impossible for me to succeed.  I like being responsible, and, as a founder, the buck will stop with me.

But most importantly, starting, operating, and growing a business is fun.  I’ve done it before and I liked it.  I’m looking forward to doing it again!

Image courtesy of Carlos Porto /