Run Small Experiments
Written date
Aug 24, 2025
Category
Innovation

Big plans feel safe. They come with timelines, budgets, and confident slide decks. They also hide a risky assumption: that you already know what will work.
Innovative leaders flip the approach. They treat ideas as hypotheses and run small experiments to find out.
1. Start with a question
Every experiment should answer something specific. Will customers use this? Will this process save time? A clear question keeps the test small and honest.
2. Build the smallest useful version
Eric Ries popularized the minimum viable product in The Lean Startup: the simplest version that lets you learn. It doesn’t need to be polished. It needs to be real enough to get a genuine reaction.
3. Compress the timeline
In Sprint, Jake Knapp and his colleagues lay out a five-day process for going from a big problem to a tested prototype. The lesson applies anywhere: short deadlines force focus.
4. Decide in advance what success looks like
Before you run the test, agree on the result that would make you continue, change course, or stop. It keeps everyone honest when the data comes in.
Define the question
Build the smallest test
Set a short deadline
Agree on the success signal
Bottom line
Experiments make failure small and learning fast. Run enough of them and you won’t need to guess what works. You’ll know.

