THE IDEA
The lightest possible touch
The instinct when facing a problem is to match the size of the response to the size of the problem. Big problem, big intervention. This feels proportional. In complex systems, it’s often exactly wrong.
A minimum viable intervention is the smallest change that could plausibly shift the system in the direction you want. Not the comprehensive solution. Not the transformation programme. Not the grand plan. The smallest move that could work - deployed quickly, observed carefully, and built on if it succeeds.
The logic is simple. In a complex system, you can’t predict the response to your intervention. Larger interventions are harder to reverse, produce more side effects, take longer to deploy, and fail more expensively. Smaller interventions are faster to deploy, easier to reverse, produce less collateral damage, and generate learning at lower cost. The minimum viable intervention isn’t a compromise. It’s a strategy - one that respects complexity by keeping the bets small while the learning is large.
IN PRACTICE
Small moves, big learning
A team’s meetings are unproductive. The big intervention: hire a facilitator, redesign the meeting structure, introduce new tools, train everyone in meeting protocols. The minimum viable intervention: start every meeting with one question - “what do we need to decide today?” - and see what changes. If it works, the meeting has a focal point. If it doesn’t, you’ve lost nothing and learned something about why the meetings aren’t working.
A neighbourhood has a litter problem. The big intervention: more bins, more enforcement, a public awareness campaign, community clean-up events. The minimum viable intervention: put one additional bin at the exact spot where litter accumulates most. If litter reduces there, add more bins at other hotspots. The small move tests whether the problem is infrastructure (not enough bins) or behaviour (people don’t care), at a fraction of the cost of the comprehensive programme.
A person wants to improve their sleep. The big intervention: new mattress, blackout blinds, sleep tracking app, evening routine overhaul, no screens after 8pm. The minimum viable intervention: put the phone in a different room at bedtime. If sleep improves, the phone was the issue. If it doesn’t, the problem is elsewhere and the expensive interventions can be better targeted. The MVI costs nothing and generates the information that makes the bigger decisions smarter.
WORKING WITH THIS
Finding the smallest useful move
Ask: what’s the simplest thing we could try that would test whether our understanding of the problem is right? The answer is your minimum viable intervention. It doesn’t need to solve the problem. It needs to teach you something about the problem.
Design the intervention to generate feedback. If you change something and can’t tell whether it worked, it wasn’t a useful MVI. Define what you’ll look for before you act. How will you know if the system responded? How quickly will you see it?
Be willing to scale up. The MVI isn’t the end - it’s the beginning. If the small move works, amplify it. If it partially works, refine it. If it fails, you’ve learned something cheaply and can try a different small move. The sequence of small moves, each informed by the last, is usually more effective than one large move informed by analysis alone.
THE INSIGHT
The line to remember
The smallest intervention that could work is almost always the smartest one to try first. You can always go bigger. You can’t always undo big.
RECOGNITION
When this is in play
You need a minimum viable intervention when the temptation is to launch a large programme but nobody is sure the diagnosis is right. When the cost of being wrong is high and the cost of trying something small is low. When the system is complex and unpredictable and you’d rather learn cheaply than fail expensively. When someone says “we need a comprehensive approach” and you suspect a targeted one would teach you more, faster, for less.