The nature of entrepreneurship: randomness, trial and error, and how intelligence can get in the way

In many ways, I am not a typical entrepreneur. Entrepreneurs are often risk taking, creative types who do not do well in the traditional education system or workplace. By contrast, I was the shy, introverted A+ student who was conservative and deferred to authority. I highly valued, and in many ways still value, the classical education I received and found beauty in the logical arguments I constructed both for my classes in college and for my clients at McKinsey.

Fortunately, somewhere along the way, I came to realize that’s not how the world really works. There is a tremendous amount of uncertainty in the world and especially in the fast moving environment of the Internet and thus management by trial and error is probably the single most effective way to become successful.

Consider this:

  • Friendster does not know why it is successful in the Philippines.
  • Orkut does not know why it is successful in Brazil.
  • Hi5 does not know why it is popular in Portugal.

In all three cases, it was not planned, it just happened – maybe because a single important user randomly decided to use that site rather than another. I am sure they have found arguments to rationalize their success since, but I bet those are only one of many possible explanations.

Also consider this: when eBay, Google, Yahoo, MySpace, Youtube and most other such companies were created, their founders did not expect them to be nearly as successful as they became.

From personal experience, I can tell you it’s far more effective to try something, discard it if it fails and continue refining it if it works than trying to theorize what is the most effective. Quick decisions – even wrong ones – if you have the humility to accept that you were wrong and are willing to rapidly change course – are much better than a long decision making process.

I don’t recommend running a startup by consensus. It takes time to get to a consensus and what the internal consensus of your company is, is frankly irrelevant. By all means involve everyone in the decision making process: get everyone’s opinions about what should be tried, but make up your mind rapidly and test the idea. In fact, I highly recommend testing multiple ideas and doing A/B testing if possible.

In other words, intelligence gets in the way because by looking for what should be the right answer, you waste valuable time you could be spending finding out what actually is the right answer.

Also note that many “breakthroughs” happen in a similar way. The Wright brothers did not just wake up one day and built a flying plane. They tested dozens of wing designs, built different types of engine. It was a grueling multi-year process of marginal refinements, the sum total of which led to a flying plane.

  • I wouldn’t refer to it as “intelligence gets in the way” –intelligence is the wrong word, I’m afraid…

    What entrepreneurs have is “street smarts” (a form of intelligence), but perhaps not “book smarts” (which is another form of intelligence).

    The minute you are a book-smart person, you try to plan too much, think too much, go by the rules–which is exactly the opposite of what start-ups require.

    I think about how I have formed all of my ventures (some nonprofit, some for profit)… All were a function of my willingness to take some risk and not think like a philosopher nonstop but just executing efficiently and effectively on my gut feelings. Mind you, I was also very good in school without even trying (I didn’t like studying so much but would still get a high enough score to compete at the national level and got into the toughest school in Turkey and arguably Europe–where this school’s students win international math competitions each year consistently!).

    So it’s safe to say most real entrepreneurs are “intelligent executers” of randomly high potential ideas and if they try long enough or by pure chance once in a while, they tend to succeed. They are street smart and see what’s missing in the world :).

  • Excellent timing Fabrice!
    I’m in the middle of a week with a thousand ‘left or right’ decisions to make, with very little information to make them with.

    Time to stop over-thinking, take a position and be prepared to keep my eyes peeled and iterate.


  • i think it can be simply summed up as this:

    Entrepreneurship is more about discovery than it is about invention.

    And I completely agree with that premise. We’ve changed tack at least 3 times now over the past year with the business model for JumpBox and we’re zeroing in on what the market wants. It’s a wonderful iterative game of making hypotheses and then confirming or disproving them. The idea of writing an exhaustive business plan up front is just silly. The artifact generated is worthless- it’s the process of asking and testing the questions that’s valuable.


  • In my former job at Concept Studio in Pitney Bowes, we used to call this philosophy “Fail fast to succeed sooner”.

    I also read a story in a game theory book a few years ago that talked about how in the olden days, when people went to war, they used cannons. Each cannon ball was heavy and expensive and soldiers had to spend minutes adjusting the cannon to try to hit the target precisely. If they missed, they just lost a lot of time and money. Today, on the other hand, we use machine guns. Tiny, cheap bullets that are fired away first and then the aim in adjusted to hit the target. A sticky line from the book was “Change your strategy from ‘Ready, Aim, Fire’ to ‘Ready, Fire, Aim’.”

  • I agree with the general idea of your post, I think what you are trying to get at is that decision making under uncertainty adheres to the law of diminishing marginal utility (sorry, I’m an econ major…).

    In other words, thinking about all the choices you have (say 20) in the very early stages weeds out a lot of the bad choices and is a good investment of time (imagine how long it would take to test 20 different scenarios), but when you’ve narrowed it down to just one or two choices, trying to figure out which one to pursue becomes an impossible task and it’s just not worth spending more time on–much better to just get out there and test them both.

  • I believe that a part of the mystery of success of entrepreneurship is also the people with whom you go into the (ad)venture… the creative alchemy, the yin-yang complement… On the one hand, your partner(s) provide checks & balances; on the other, they broaden the scope of “intelligence” by acting as a sounding board before you go to market.

  • From my masters studies in DSP: You can characterize systems with noise inputs. White statinary gaussian noise input into a system is a good start. NOISE -> SYSTEM UNDER TEST -> OUTPUT …..put it all through a recursive filter of the order you wish to use. Calculate the coefficients. Sounds pretty similar to me. You would get no where if you put a single frequency sine wave in. The analogy is likely complete.