William R. Kerr, Ramana Nanda, and Matthew Rhodes-Kropf write,
The costs of running experiments play a big role in entrepreneurship. Technological change in the laqst decade has dramatically lowered these costs, particularly in industries that have benefited from the emergence of the Internet
Right. But then why do we see the decline in the rate of start-up formation that I have discussed in other recent posts?
Later, they write,
Large corporations…often find it difficult to terminate experiments that aren’t working out.
Right again. And government is even worse at terminating failed experiments, of course.
I have to say that I found little in this article that wasn’t obvious to me before.
One complication for the analysis is that current wisdom around computer startups is that they experiment quickly internally. Even by the standards of a small startup’s funding, the idea is that you don’t want to blow through your cash on one big experiment. Instead, do the cheapest possible experiment to see how it goes. Then, you “pivot” to the next experiment. If you are good at it, and if your fundamental idea is good enough, then your sequence of pivots will get you to an arrangement that is actually viable.
As an extreme example, before you build a web site, make a web site with static pages and try to sell *one single copy*. Process the sale manually. If you can’t get a single buyer, then you’ve already learned something rather important. Hopefully you do get a sale, at which point you can start analyzing all the costs for each step of the pipeline you need to filfill the order.
I don’t know how much this wisdom matches reality, but it’s certainly getting a lot of talk, and it would lead to a fewer total number of startups.