Remember back in grade school when we learned about the the scientific method?

The scientific method is a proven, step-by-step technique for discovery and validation. It uses questions and experiments to help us solve problems. It turns guesses and trends into data, and it’s effective across diverse environments.

And it’s a wonder that more of us don’t use it as a tool for business strategy.

We know that if we want to succeed in business, we can’t stand still. Strategy is what defines how a company will grow and move forward. It’s a plan for achieving our goals. However, since the future is uncertain, any strategy is really just a guess. After all, if we knew for certain what to do, we would just do it!

So although it’s well-informed and carefully considered, there’s a chance our strategic guess could be wrong. We know our market and our customers well. But because we can’t see into the future, our plan is still just a guess that has yet to be proven.

To validate our guess, it’s helpful to think of it as a “question in need of an answer”. What do our customers want? Where do we see the market going? What problem does our company need to solve?

Using the scientific method, our predicted answer to this question becomes our hypothesis. A hypothesis is just our best idea for how we can be a better company, a few months, quarters, or years from now. And any action we take against this hypothesis is an experiment.

Most executives are not accustomed to thinking of their strategic plans as experiments. After all, experiments connote uncertainty. Leaders are expected to show confidence in their plans. Implementing strategy is expensive. So it brings up some interesting challenges: What if we’re wrong? What if our guess is invalidated? How will we know?

As leaders, we need to balance having confidence — and conveying a vision for a better future — with creating the right environment for experimentation. The scientific method gives organizations a framework for doing this.

For example, how can we know as early as possible whether our strategy — our hypothesis — is a good one? It’s always easier to define lagging indicators than leading ones, but past data doesn’t always predict the future. Experiments trade blind execution for early detection. Running experiments means figuring out ways to “sense” — within a quarter or two, rather than a year or two — whether we’re on the right track. So what could we notice earlier, inside our business, that would tell us whether our strategy is working?

Experiments also help us distinguish between working IN the business and working ON the business. Working IN the business means meeting the commitments we’ve already made (such as, “we’re going to generate 10% more revenue”) whereas working ON the business means transforming ourselves and our processes in a way that improves outcomes (“how might we become a business that is capable of generating 10% more revenue?”)

A strategy that works ON the business will include experiments that attempt to transform the organization in some way. Maybe we’ll work on our people. Or maybe we’ll work on our structure. Let’s say a fast-growing company has just hired its first CMO. This too is an experiment! Does the leadership team know what results they want to see from their new hire? How will the company measure the impact of their new executive?

Having a clearly defined — and widely understood — strategy is key to running useful experiments, so that we’re all operating with the same hypothesis and conditions. It’s often true that a CEO can articulate the company’s strategy, but no one else can. What seems like a crisp explanation being broadly shared is actually a vague proposition being inconsistently socialized. The result is an organization where the CEO thinks that everyone understands the strategy, but in fact very few do.

So we need to articulate our strategy in a very clear way and then repeat, repeat, and repeat it again. Take the example of a CEO who decides that his company strategy will be to focus on only three priorities for the coming quarter. Do employees share an understanding of the CEO’s hypothesis, and the results they should expect from focusing on just three priorities? Do they see that focusing on only three things means NOT working on everything else?

There are great tools available to help us codify and share our strategy — including True NorthMother Strategies, and A3 problem statements. These tools help us “build a backlog” of experiments that our executive team needs to manage going forward. And the bigger the company, the more complex this work becomes.

Though the scientific method is akin to approaches like Lean Startup, agile, and design thinking, most executives with whom I’ve worked aren’t concerned with whether it’s part of a specific process model. They don’t really care what this approach to strategy is called. They like it for this one reason.

It works.

We can apply the scientific method within any organization — no lab coat required. In each step we’ll get answers to a fundamental question:


1. Make an observation

2. Ask a question

3. Form a hypothesis

4. Run an experiment

5. Analyze the results

6. Share the results


1. Actively engage in knowing your business

2. What’s the problem we’re trying to solve?

3. What could we do differently to solve the problem?

4. Take action (make a plan and execute)

5. Did we solve it? Did it amplify or dampen?

6. Bring the learning to the whole organization

In future posts, I’ll cover how companies can actively engage in this process, frame the right questions, decide on experiments, evaluate the data, and draw clear conclusions. We’ll also investigate approaches for when the leadership team doesn’t agree on the hypothesis.

What ideas do you have about strategy and science? Share your thoughts in the comments!