The Hypothesis Validation Board is best used when a single experiment is being run at a time. To start designing a startup hypothesis validation experiment define either the customer segment you are targeting, or value proposition of a solution you have in mind. When starting with a solution in mind, you must be able to make a guess about the target customer segment. This is not true vice versa. When you have an idea about the customer segment or the value proposition, try to formulate a hypothesis on the customer jobs, i.e. the thing the customer is trying to get done and the problems that stop them getting it done.

At the very start, one of the above three will do well enough as riskiest assumptions. However, as you progress you may need to consider more implicit assumptions you are making regarding customer motivation, behaviours, and value creators.

Now it’s time to define the heart of the experiment; the methods by which you will gain data to validate or invalidate your hypotheses. This might include interviews, prototypes, landing pages, or any other mechanism which gives you valid and reliable data; preferably first hand. Specify quantities when sensible, such as 20 interviews. This will make sense once you address the next dimensions of the experiment; metrics that matter and minimum success criteria.

When conducting interviews, for example, metrics that matter may include certain answers given, words spontaneously used, or the level of emotion in voice and gestures rated by the interviewer on a scale of 0 to 5. Knowing what your measurements are before going out will help a lot when it comes to make sense of your data. When you’ve figured out the metrics for the test method, then decide on the minimum success criteria. Deciding, in advance, what constitutes strong enough evidence to validate a hypothesis, will allow you to be more objective when the data is in. An example will illustrate the idea. Suppose for the method we choose 20 interviews, and decide to measure the emotion interviewees display when talking about a topic. We define, in advance, that we are right (about the hypothesis) when 10 of the interviewees show a level 3 or more on emotion.

At this point you’ll conduct the methods to gather the data. Once you’ve done that, and analysed the results, you can close the experiment. Note observations, insights, & lessons learnt regarding the outcomes of the experiment. What was the result of the metrics measured? Did you uncover unexpected information? What did you learn about the content and the process? And last but not least, (the whole point of the experiment), conclude in decisions & actions if the hypotheses are validated or invalidated, if you will persevere along the path, iterate the experiment, or pivot, and the major changes your insight demand you make to your business model.

Don’t forget to update your business model canvas!