Smallest Single Unit of Innovation
Run one tiny reversible experiment before attempting a sweeping change
- Difficulty
- Starter
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 99%
Start with the change you hope to create, then deliberately shrink it to the smallest unit that can produce real evidence. The test should be brief, visible in actual work, and easy to reverse. Run it once, observe whether it improves the experience or outcome, and either repeat it or stop without treating the result as a personal failure. Godin's example is a new 30-second podcast feature: large enough for listeners to experience, yet small enough to remove if it fails. The mechanism replaces abstract ambition with responsible agency. Because the downside is bounded, people can move beyond scripts and discover possibility without waiting for a leader to announce a transformation program.
Origin
Seth Godin asks for the smallest single unit of innovation and illustrates it with a 30-second feature added to one podcast episode.
Core principles
- 01Innovation begins with the smallest meaningful change
- 02A reversible test lowers fear without eliminating learning
- 03Success earns repetition; failure earns a cheap stop
- 04Agency grows through responsible action rather than permission waiting
How to run it
- 1
Choose the desired change
Specify whose experience should improve and what better would look like.
Pro tip Choose an outcome someone can notice, not a vague desire to innovate.
Watch out Do not begin with a large solution before defining the change.
- 2
Shrink the intervention
Reduce the idea to the smallest version that can still create useful evidence.
Pro tip Ask what could fit into 15 minutes, one interaction, or one release.
Watch out A test that cannot affect anyone teaches little.
- 3
Bound the downside
Set a short duration and make the change reversible. Preserve any approval or safety controls the context requires.
Pro tip Write the stopping condition before running the test.
Watch out Small does not automatically mean safe in regulated or high-stakes work.
- 4
Run it in reality
Put the small change in front of the people it is meant to help and notice what happens.
Pro tip Collect a direct reaction rather than judging only from your own effort.
Watch out Do not polish indefinitely before obtaining evidence.
- 5
Repeat or retire
If it creates value, do it again and improve it. If it does not, stop and retain the learning.
Pro tip Treat a cleanly stopped test as useful progress.
Watch out Do not continue merely because you invented the idea.
In the wild
A host adds one original 30-second feature at the end of the next episode. Listeners experience the feature in a real release, but the rest of the show remains unchanged. The host reviews direct responses and completion behavior before deciding whether to repeat it.
→ The host learns whether the idea creates value without betting the whole format on it.
Common mistakes
Starting with the grand redesign
A transformation-sized proposal raises fear, cost, and permission barriers before any evidence exists.
Making the test meaningless
A tiny activity that no intended user can experience is not a useful unit of innovation.
Ignoring the result
Repeating a failed test by default turns a reversible experiment into a habit.
Is it for you?
Best for
People who want to improve work but feel blocked by hierarchy, perfectionism, or the scale of the change.
Not ideal for
Irreversible decisions or safety-critical changes that require formal review and controlled validation.
From the transcript
“What's the smallest single unit of innovation you could bring to your work?”
“And if it worked, that would be great cuz you could do it again. And if it didn't work, you wouldn't have to do it…”
From the episode
Seth Godin: Practice of Creativity
Seth Godin