Death by detail: When data is dangerous
We live in a time when building a startup is increasingly becoming a science. The world is awash with ‘lean’ methodologies, ruthless A/B and multivariate testing and communities where ‘talking to customers’ is known as ‘customer discovery’. Is that signalling goodbye to the good ol’ ways of building a company? I’d argue not.
Data, of course, is a great leveler. Suddenly, every idea is an opinion that needs to be proved or disproved. It keeps the team on one page without hierarchical bias. Warm and fuzzy feelings of anecdotal growth are irrelevant.
In its most extreme, anal and obsessive avatars however, this culture can be fatal.
- You talk to 25 customers and ‘measure’ what they don’t like? That’s wrong - in all matters of science and arts.
- You change the color of a signup button after 10 days of your website launch to see conversion rates? Nope, you can’t optimize before building.
- You dive into five levels of detail before attacking obvious problems and die a death by detail?
A better way to understand data is that it helps you get comfortable with your counter intuition. To arrive at a mental state where the exact opposite of your expected result is equally obvious. Go beyond just objectivity and start telling your intuition that it’s ignored brother named counter-intuition is also important. It’s almost a deep, philosophical level where useful contrarians exist.
Let the science underpin your art. Not the other way round.