Select Page

What do people mean when they refer to strategy as a Hypothesis?  Why should you treat strategy as a hypothesis? How does it help the strategy process? What are the different types of strategy hypotheses?  This article explains the thinking behind strategy as a hypothesis, and how it affects strategy development, strategy implementation and how we manage out strategy.

1 First, what is a hypothesis?

First a hypothesis is different to a theory.  A theory is provable.  It is capable of being tested under repeatable conditions.  That is why science develops and uses theories. They are a sound base, until proven otherwise and a better theory comes along.  Unfortunately the nature of organisations and their strategy is that we do not have the luxury of perfectly repeatable conditions for our strategy.  The environment is changing, the competition are moving, the clients and customers are changing.  Our strategies are not theories (though we might like to think so).  They are hypotheses.

1.1 Defining a hypothesis

A hypothesis is quite different.  Here are two useful definitions of hypothesis:

  1. A hypothesis: A supposition or proposed explanation made on the basis of limited evidence as a starting point for further investigation.
  2. A hypothesis: A proposition made as a basis for reasoning, without any assumption of its truth.

This definition has three important parts:

  • The supposition or proposed explanation (proposition) This is typically referred to as ‘the hypothesis’
  • No assumption about its truth. Limited evidence, or even no evidence.
  • A starting point for further investigation.  The hypothesis must provide a route to investigation and testing.

1.2 Hypothesis: A basis for reasoning – with no assumption of truth

The strategic hypothesis is a basis for reasoning, and exploration.  It might even be a hunch or guess or just an idea.  It may have some evidence or basis.  It is just as important to wonder and think, “Is it possible that this is true?”, as it is to say, “I see some evidence and I believe this might be true?”.  The point is that both are hypotheses. Truth should not be assumed.  Both should provide a basis for testing.

1.3 Genuinely test your hypotheses

You must genuinely test the hypothesis.  It is important that further investigation must set out to challenge the original hypothesis.  It is important to accept that the result of the tests or investigations may prove, or may disprove, the hypothesis.  Further investigation, merely to gather more data to ‘prove’ the hypothesis is not what we are talking about.  That is known as “decision justification”.  Decision justification is not testing a hypothesis.

Even with a lot of evidence our strategies are still a hypothesis that needs to be tested.  Of course you can test by gathering facts.  You can test through action and implementation.  The point is that you are treating the strategic hypothesis, not as an absolute fact, but a belief or explanation that maybe wrong or need refining.  We are talking about rigorous, intellectually honest, learning applied to our strategy.

2 The strategic hypothesis is not the whole strategy

Whilst I referred to this article as “Strategy as a hypothesis”, there are in fact several hypotheses in play here.  The hypothesis about the market or environment, the hypothesis about the strategic course or direction and the hypothesis about how to implement that strategy.

2.1 First, there is the hypothesis about the environment and the market condition.

For instance, we believe there is a need for this product and customers will by it.  Or there is a social need that is unmet and current investment in social care is not meeting it.  These are hypotheses about the environment or context.  They are about what is going on and the underlying reasons.  They can be tested.  They are associated with Rumelt’s “What is going on” and Excitant’s strategy tablet stages “What is doing on, the underlying reasons and our diagnosis”.    We are not yet at the specific strategy or policy to address the issues.

2.2 The second hypothesis is about the proposed solution to the issue.

If this is really the situation, does our proposed strategy, policy or approach actually address this issue?  Richard Rumelt calls this the Policy. We call it the strategy, or approach.  At this stage you might have alternative strategies to address the problems.  All are hypotheses that need testing. For instance, if the social care problem is not a funding issue, but one of recruitment?  What types of strategy could be adopted to address the recruitment issue?  Can staff be brought in from elsewhere?  Is it a training issue?  Can the community itself provide the resources?   Each of these strategies could address the problem.  Each are in turn a hypothesis.

2.3 There is also the strategy implementation hypothesis.

What is the best way to implement this strategy?  What set of coherent actions should we adopt?  Can we test these potential routes to see which would work best, or in what combination?   This is consistent with our strategy tablet’s (and Richard Rumelt’s) coherent actions.

How can we test whether this set of coherent actions will implement the strategy and address the underlying problem.

2.4 Recognising the three strategic hypotheses operate as a set

So, whilst “Strategy as a hypothesis” sounds like one thing, there are always multiple hypothesis in the overall complete strategy.  The large strategy consultancies tend to focus on validating the external environment, and, of course, selling on a large team of junior consultants to implement the strategy for you.  (Sorry some cynicism popped in there).

In reality, as a management team you are forming a set of hypothesis that all, together make up your strategy.  Each part is a hypothesis, has limited information and can be tested.

3 What does strategy as a hypothesis mean for strategy management?

At Excitant, we like to think about how we manage as well as the strategy itself.  The choice of how we manage strategy and our strategy management process is an aspect of this.

3.1 The assumption of large consultancies is strategy as a planning process

One problem is that whilst the model of strategy is that it is a hypothesis to be tested, the large consultancies tend to operate a model of strategic planning and annual strategy.   This means that they test their hypotheses about the strategy during the strategy development stage.  However after that, the model is generally, that the strategy is now complete. The issue is then strategy implementation or execution.

This is the strategic planning model.  If we develop a robust strategy, plan well, then the issue becomes implementation for the next few years.   All we have to do is a little strategic planning review event next year.  The assumption is that we are on the right ship and the right course.  We should stay on that ship, on that course.  This is not teh only model.  Nor is the model as appropriate today, as it was.

3.3 The alternative is treating strategy as a learning process

In contrast, looking at the whole of the strategy as a series of hypothesis, you would adopt more of a process of strategy as a learning process.    You would continuously ask:

  1. Are our assumptions about the outside world still correct?  Do we have more facts?  What information is coming in?  What are we learning about our markets and customers and environment, as we execute our strategy?
  2. Is our strategy still correct?  Is it possible our strategy is wrong or needs refining?
  3. Are we communicating, implementing and operating our strategy correctly?  Does how we are doing that need to change?

So what we have really is an overall philosophy of strategy as a learning process.  This is the basis of the strategy learning model we use at Excitant.

3.4 Strategic hypotheses does not mean strategy paralysis

The point of these hypothesis is to test ideas so that you do not implement a strategy only to find out it does not work 3 years later.  However you also do not want strategy paralysis.  Where no one dare do anything for fear of not having enough evidence.

The challenge is to test and things quickly.  To test things on it a safe fail environment (where you can fail safely, rather than avoiding failure).  Strategic hypothesis is not the opposite of agile entrepreneurial strategy.  On the contrary, entrepreneurs and investors are interested in gathering data quickly so their investment is safe, and yet they get to implementation quickly.  Good investors, get their clients to test their strategy quickly and prove that they have something that the market wants.  All we are discussing is how to develop and test your hypotheses quickly.

4 What does strategy as a hypothesis mean for strategy development?

In part good use of strategy as a hypothesis means that the type of thinking, conversation and approach to strategy changes.

4.1 A different type of conversation

When you treat strategy as a set of hypotheses, you start a different type of conversation.  You change the type of conversation to one which is more exploratory.  You explore ideas as a team and look for information to test those ideas. Here is an example where we helped a management team exploring some potential hypotheses in the banking sector.

Egos should be removed.  You are exploring ideas, independent of their source.  You are not proposing and defending.  This is evidence based decision making.

Testing and learning:  This is not a discussion about failure, nor avoiding failure.  This is discussion about what are we learning.

Learning as your competitive edge:  There is an argument that, when it all comes down to it, the only sustainable competitive advantage is to be able to learn faster than your competitors.

4.2 Testing the various strategy hypotheses

Testing of hypotheses is important.  It is not simply force of argument, or strategy by political will.  It is evidence based decision making and taking.

There are various way to test the strategy hypotheses.  here are just a few ideas:

  1. When looking at what is going on now
    • Customer, Market and industry research
    • Competitor analysis
    • Model alternatives
  2. When testing alternative strategies
    • Explore strategic options:  What other ways could this be achieved?
    • Test each strategy against various scenarios.
  3. When testing implementation hypotheses
    • Set up pilot schemes
    • Run test environments
    • Run multiple channels of communication & test what is being interpreted.
    • Trial with real customers
    • Experiment with different cultures

These are just a few ways that hypotheses about aspects of the strategy can be tested.

5.  Conclusion: Remove ego, explore ideas and learn quickly from your strategy

Treating your strategy as a hypothesis helps to get rid of ego from your strategy.  It forces rigorous, intellectually honest appraisal of ideas and learning.

Though it is traditionally seen associated with strategy as a planning process, it is actually the foundation of strategy as a learning process.

This is about improving the quality of conversation,thinking and learning about your strategy.

If you would like to learn more about improving your approach to strategy then you can