Form Your Hypothesis

Learning Objectives

After completing this unit, you’ll be able to:

  • Explain what makes a good hypothesis.
  • List elements to include in an experiment proposal.

What Makes a Good Hypothesis?

A good hypothesis is clear, concise, and specific. When creating a hypothesis, focus on the idea you’re testing and the outcome you’re looking for.

The Oxford English Dictionary defines hypothesis as the following:

“A supposition or proposed explanation made on the basis of limited evidence as a starting point for further investigation.” 

This is how Salesforce approaches a hypothesis:

By doing {this specific thing} to {that specific thing}, {this specific outcome} can occur.

Here are some examples of a hypothesis:

  • By introducing an IVR to the Customer Support Line, we can reduce Agent-Facing Account Queries by 30%.
  • By running an extra 5 miles per day in addition to our basic training plan, we can complete a sub 20-minute 5km in 6 months.
  • By offering a Premium Subscription Service in addition to the Basic Plan, we can increase our revenues by 20%.
  • By utilizing a Knowledge Base/FAQ service on our Customer Portal, we can reduce the number of Agent Calls by 50%.

Notice that all of the above hypothesis statements are varied in nature and typically lend themselves to quantitative outcomes. We like quantitative outcomes because hard data helps confirm that we’ve proven the hypothesis. In a business experiment, you need to clearly determine what drives a “good outcome.” For instance, is your experiment yielding a good outcome if you’ve saved money or decreased customer complaints? 

The Importance of a Strong Hypothesis

As we covered in the first unit, the hypothesis is a statement that drives your entire experiment. It is specific and atomic in nature. It's easy to make your hypothesis overly complicated, so it's important that you keep your hypothesis simple, clear, and concise. Introducing too many elements and variables into your hypothesis can make the experiment unnecessarily complex. As an example of a complex hypothesis, see the following hypothesis statement.

By introducing an IVR and a Chatbot to the Customer Support Line, we can reduce Agent-Facing Account Queries by 30% while increasing upsell opportunities by 10%.

This complex hypothesis describes four experiments. When treating this as one experiment, you cannot definitely know whether it was the chatbot or IVR that resulted in the reduced query time or the increased sales. Thus, this experiment should be structured as:

[1] By introducing an IVR to the Customer Support Line, we can reduce Agent-Facing Account Queries by 30%.

[2] By introducing an IVR to the Customer Support Line, we can increase upsell opportunities by 10%.

[3] By introducing a Chatbot to the Customer Support Line, we can reduce Agent-Facing Account Queries by 30%.

[4] By introducing a Chatbot to the Customer Support Line, we can increase upsell opportunities by 10%.

The reality is that your idea is just an idea. You need to prove whether the idea has business value—your hypothesis is the first step in discovering that value.

The Experiment Proposal

Now you’re ready to create your proposal, which is a succinct overview of what your experiment entails. Keeping it simple and to the point is essential. Below are some examples of what you should typically include in your experiment proposal.

  • Experiment title: Create a unique identifier or reference.
  • Accountable executive: The sponsor and person who holds the purse strings and has committed to support every next step of the experiment.
  • Experiment owner: The person who is leading the experiment and designing and executing the experiment. Typically the person with the idea.
  • Hypothesis: The idea, in specific, clear, concise form. Remember to use this format: By doing <<this specific thing>> to <<that specific thing>>, <<this specific outcome>> can occur.
  • Background: Why are you doing this? Is it to solve a problem, or is there an opportunity here to explore?
  • Potential test method: This is an idea of how the experiment can be tested. Think of this as a micro-plan for the experiment.
  • Potential value: What value are we looking to capture? Is it de-risking a deliverable? Is it market share? Is it pure financial opportunity? If this can’t be articulated, reevaluate the experiment.
  • Outcome and next steps: Define your next steps depending on the outcome of the experiment. If the hypothesis is proven, is the experiment scaled? If disproven, do you come up with a new hypothesis?

The above list covers all of the bases and allows you to test your ideas and intentions in an informative way. 


Keep learning for
Sign up for an account to continue.
What’s in it for you?
  • Get personalized recommendations for your career goals
  • Practice your skills with hands-on challenges and quizzes
  • Track and share your progress with employers
  • Connect to mentorship and career opportunities