This week, I went to the AWS (Amazon Web Services) Lean Cloud event. At these events, I often find the examples are about startups with funding behind them and a team of programmers. You try to think about how the examples could apply to you and it doesn’t match your reality. Eric Ries had great examples and comments that instantly connected.
You’re running a scientific experiment
A startup is an experiment in a scientific sense. You have a hypothesis that needs to be tested.
This really resonated with the scientist within me and it applies to any new business (even one that has been around for a few years). hy·poth·e·sis – noun
- A supposition or proposed explanation made on the basis of limited evidence as a starting point for further investigation
- A proposition made as a basis for reasoning, without any assumption of its truth
Think of your business as a proposed explanation or trial solution, which (if you’re honest) is based on limited evidence and needs to be tested, just like a scientist tests their hypotheses when conducting an experiment. Experimenters may test and reject several hypotheses before solving the problem. As a scientist, you’ll think, well I proved that hypothesis wrong, so I’ll move onto the next one. You might apply a few different tests to make sure, but after a few tests have come up wrong, you know you have to develop a new hypothesis. Just as a scientist throws away many hypotheses when testing solutions for a problem, most startups fail. However, scientists seem to be better at moving on than most business owners or entrepreneurs who keep trying to prove that their hypothesis is true.
Make many small experiments
Do micro scale experiments to see if you’re getting closer to the product market fit.
Instead of completely building everything then going live to see nothing happen, do little steps and test reactions. Set up accountability measures early so you have numbers to test. Listen early to customers and their suggestions. Move on to your next micro scale experiment and test the outcome.
Putting your scientific hat on:
What is the hypothesis your business is testing? Fill in the blanks: If _____[I do this independent action] _____, then _____[this dependent outcome]_____ will happen. Is it testable in terms of being able to measure what you do and what will happen? What small experiments are you doing to test whether your hypothesis is accurate or not?