Scientific method for startups: let evidence decide.
The scientific method for startups means treating your idea as hypotheses and testing each one before committing. In randomized trials, for example, founders taught this were more likely to drop ideas. Yet those who continued performed better, because evidence, not feelings, made the call.
What the Bocconi trials found.
First, at Bocconi University, Camuffo, Cordova, Gambardella and Spina ran a randomized trial published in Management Science in 2020. Specifically, founders were taught to test business assumptions before committing. As a result, they were more likely to terminate projects. Then, in 2024, the group repeated it with 759 firms in four trials, and the approach again increased terminations.
Cheap experiments that paid off.
Further startup experimentation research came from Harvard and Duke. Similarly, in a 2022 Management Science paper, Koning, Hasan and Chatterji studied a large sample of technology startups. Among the few that adopted cheap online experiments, performance improved by 30 to 100 per cent after a year. In short, testing your beliefs against reality pays.
Where the scientific method for startups has limits.
Nobody in these studies did better by feeling more confident or more frightened. Instead, the winners kept asking their idea to prove itself. However, almost everyone ran a science or technology startup in a wealthy country. So be careful about applying the results to corner shops, restaurants or farms.
Apply the scientific method for startups tonight.
First, list three beliefs your business depends on, each written as a testable claim.
Next to each one, write the evidence that would make you continue or stop.
Questions about scientific method for startups.
Does the scientific method for startups make founders quit more?
Yes, in these trials it did. However, the founders who continued performed better.
Should my fear decide whether I continue?
Your fear is one voice in the conversation. Instead, your evidence gets the deciding vote.
Do these results apply to every business?
Not necessarily, because the studies mostly covered tech startups in wealthy countries.
Read the whole chapter.
This lesson covers one section of chapter 7, “When the coward is right”. The book goes further, with the full evidence, the stories and every exercise.
Lessons in this chapter.
- 01 Knowing when to quit a startup: why it is a skill.
- 02 Entrepreneur overconfidence bias: who is most wrong.
- 03 Scientific method for startups: let evidence decide.
- 04 Business closure rate in my area: count it yourself.
- 05 Aquinas on courage: why too little fear is a mistake.
- 06 Find closed businesses on maps: fixing bad counts.
- 07 When to abandon a business idea: a four-second lesson.