Leadership is a Constant Experiment

why curiosity beats certainty, and how small tests make teams stronger.

Most leaders talk about certainty like it’s a job requirement. Teams look to you for answers, executives expect you to chart a path, and a whole industry sells confidence as a leadership virtue, usually in hardcover.

But in real work, confidence isn’t the superpower. Curiosity is.

The longer I lead, the more convinced I am that leadership works a lot more like a science lab than a command center. You form a hypothesis, you try something, you observe the results, and you learn. Then you do it again. And again. And again. It’s never finished and never clean. Half the time you only discover the hypothesis after the experiment fails on you.

One reason I’ve read dozens of leadership books is to stockpile inputs for my own experiments.

Leadership isn’t consistency. It’s empiricism.

I didn’t understand this early in my career. I thought good leaders made decisions quickly and stuck with them. That was the job: stay the course, project steadiness, pretend certainty even when you don’t have it. In hindsight, it was just fear of admitting I was learning.

The better path is obvious once you see it. It’s also harder because it exposes you. When you treat leadership as a series of experiments, you’re admitting you don’t always know what will happen. You’re saying out loud: let’s see.

People trust you more when you stop pretending you have the answers.

Models Are Maps, Not Magic

I love frameworks. My last few newsletters covered a good chunk of my favorites. Frameworks help you predict where work might break and give you language to explain why something feels off. But models are only maps. You still have to walk the terrain.

Most teams fail in the space between a framework and the humans who are supposed to live inside it. That gap is where experiments matter. They make theories real.

Try shortening your planning cycles. Try sharing decisions earlier. Try a weekly “what did we learn” ritual. Try removing a process step you’re not sure you need. Try treating accountability as something teammates extend to each other instead of something you impose. Try anything that shifts how the team works, not only just what it produces.

Some experiments will land. Some won’t. The failures teach more. They always do.

Experiments Reduce Risk Better Than Certainty Does

A lot of leaders avoid experimentation because it feels risky. But the real risk is clinging to untested assumptions.

Douglas Hubbard makes this point in How to Measure Anything: uncertainty shrinks when you collect small bits of information. You don’t need perfect data—you need directional signals. Experiments are how you collect those signals. They turn guesses into grounded decisions.

Contrast that with the leader who insists on certainty. They delay, polish, and overthink. When they finally act, the world has already moved. The team feels the lag even if no one says it out loud. The silence is the signal.

In software engineering we talk about reducing batch size. Same concept. A smaller experiment fails faster, costs less, and teaches more. Big-bang plans feel efficient until you’re sweeping up the pieces.

I once had an idea for a new, “more efficient” planning process that I was sure would change everything. It looked perfect in my head and on my whiteboard. Nowhere else. Luckily, I tested it on one small part of the team before forcing everyone to pretend it made sense. Within a week they found more issues than I thought existed: clunky handoffs, confusing documentation, and an artifact that probably deserved its own user manual. Because the experiment was small, the embarrassment was contained. The next version didn’t need a rescue team.

And yes, I shared the bad idea, and the results, with the whole org. Learning is learning, even when it stings a little.

People Learn More From What You Try Than What You Say

When leaders experiment, teams follow. They start running their own tests. They stop waiting for permission. They share ideas that aren’t finished yet. They speak honestly about what’s not working.

That’s psychological safety in practice. Amy Edmondson’s research shows that teams with strong safety don’t make fewer mistakes. They surface mistakes faster. Experimentation creates the same effect. When everything is a learning loop, mistakes aren’t indictments. They’re inputs.

This is why experimental leaders end up with healthier cultures. Not because they’re smarter or luckier, but because they show how learning actually happens.

One of my favorite lines from Thinking in Systems is Donella Meadows’ reminder that small changes in the right place create outsized results. Experiments reveal those leverage points. They show you where a slight shift in a conversation, a boundary, or a process might unlock something bigger.

Leadership Becomes Lighter When You Stop Performing Certainty

The performance of certainty is exhausting. It forces you into a defensive posture because once you claim you know the answer, you’d better be right. Experimentation is easier on the ego. You don’t have to be right. You have to be learning.

Anyone who’s been in one of my organizations knows I carry a steady supply of half-baked ideas and share them freely. Sometimes it’s a rough draft of a process. Other times it’s a thought tossed into a team meeting or a question in Slack. I often remind my teams of this quote:

The way you get good ideas is to get a lot of ideas and then throw the bad ones away.
- Linus Pauling

I have a lot of bad ideas.

This changes how you show up. Instead of presenting polished solutions, you share ideas in draft form. Instead of protecting your authority, you ask better questions. Instead of holding tension alone, you bring the team into the ambiguity.

Three useful questions for any experiment:

What outcome do we want to influence?
What’s the smallest change we can make to learn something?
How will we know if the change helped?

That’s enough structure. Anything more and the experiment becomes a project plan.

One subtle benefit: experimental leaders don’t take failure personally. They don’t attach their identity to the result of each test. This makes them more present, more patient, and far less defensive. Their teams feel it instantly.

Experiments Build a More Capable Team

There’s a long-running myth that teams want leaders who are decisive, unwavering, and certain. In my experience, teams want leaders who help them grow.

Experimentation is a growth engine.

When you try something new and learn from it in public, you’re modeling the behavior you want from everyone else. You’re also giving them permission to do the same. That’s how you build a team that learns faster than the problems they face.

This is at the heart of Modern Testing’s fourth principle: we coach, guide, and nurture quality culture instead of guarding it. Experimentation is coaching at scale. It distributes knowledge. It breaks bottlenecks. It turns leadership into a shared responsibility instead of a positional one.

When teams experiment, they stop asking, “What does the leader want?”
They start asking, “What will make the work better?”

Those are very different questions. Only one moves the business forward.

I’ve worked in a few organizations where far too many people spent their time trying to please the leader and their whims rather than doing what was right for the business and the customer. In every one of those cases, employee engagement was weak and customers were frustrated.

You Don’t Need Better Answers. You Need Better Tests.

When I talk to newer leaders, they worry most about saying the wrong thing, choosing the wrong direction, or not having the full picture. The truth is you’ll never have the full picture. None of us do.

But you can always run the next experiment.

Try a lighter-weight process. Try a faster feedback loop. Try sharing context earlier. Try clarifying decision rights. Try giving someone else the pen. Try changing how you show up in meetings. Try setting one less goal. Try something that scares you a little.

If it works, keep it. If it doesn’t, learn and drop it.

Leadership becomes easier to carry when you stop trying to be perfect and start trying to be curious. Teams don’t need a hero. They need someone willing to learn out loud.

The Work Is the Learning

Everything in leadership is a draft. You update your beliefs as you go. You try again tomorrow. You gather new information. You make smaller bets. You build systems that learn with you.

Certainty is a trap. Experimentation is a path.

And once you step onto that path, leadership stops being a performance. It becomes something closer to the work you actually signed up for: helping people do their best work in a world that refuses to sit still.