I was thinking about an article I came across a long time ago, in which a new employee at a detergent company was tasked with designing a better soap nozzle that would increase production and much to his surprise a whole room filled with engineers only managed to crop up 10-20% improvements over current design. Having a background in evolutionary science he wondered…could he make the nozzle design ”evolve”? could random changes really be better than a dozen PhD’s?
45 random iterations or ”mutations” later he had a nozzle that was not 50% nor 80% but was performing a full 100 fold better!!! and the trick is…nobody knows why! yes, that’s right, to this day we don’t know what makes it so damn good and no engineer could have ever conceived a shape this odd for it.
Working on the #SuperAwesomeMicroProject aka full size lego car I had a similar challenge which sounded like this: make an engine out of tiny plastic pieces that can move half a ton of stuff down the road using nothing but air . So just like the soapy PhD’s I went at it, spending months and months with a slide ruler drawing up mediocre improvements.
This is the model I was thought, this is how you’re supposed to do it, right? WRONG! changing my work model to rapid iteration, hack & slash, trial & error is the only reason that lego car ever got built, it simply would have been impossible to get the 100 fold improvement otherwise. The car was evolved into existence
So what do detergent nozzles, lego cars and startups have in common? they are all made up of unpredictable elements and operate in an environment governed by chaos theory, if you want your company to survive and succeed you must evolve. The big don’t always eat the small but the fast always eat the slow.
I leave you with this 2 questions:
- How do you implement iteration at every level/stage of your startup?
- How far can we push the model before it breaks? how fast can we evolve? is there a breaking point? a kind of entrepreneurial singularity?