Then the next question is how to come up with a good hypothesis
- Start with understanding the problem:
As Albert Einstein once said: “If I had an hour to solve a problem, I’d spend 55 minutes thinking about the problem and five minutes thinking about solutions.”, first and most importantly, we need to start with understanding the problem, clearly defining it, and not dive into the solution directly, because we cannot solve what we don’t understand.
Luckily, we have the Double Diamond framework, also known as the convergence and divergence model, for problem solving that can help us identify the right problems and then come up with the right solutions.
The model illustrates and emphasizes the importance of taking time for the problem exploration, rather than directly converging upon a solution by explicitly showing separate problem and solution spaces as two different phases.
In the problem space, where our aim is to understand the people that we are trying to create solutions for, we start with questioning the problem we initially had, expand its scope and diverge a lot to find all the underlying problems. Then we converge until we find our real problem statement. After we are done with the problem space, we start with diverging to all possible solutions by ideation and then again converging to one proposed solution which we believe meets our users’ needs best.
In essence, hypothesis is the combination of the proposed solution for the defined problem we get out of the Double Diamond exercise, together with our rationale and the expected change on the metrics and user behavior. Then, we test the proposed solutions via A/B testing to validate our hypothesis.
In the end, if the proposed solution works better, then it’s great that the problem and the proposed solution matches, if not we can use our problem statement again to explore new potential solutions and iterate.
2. Create the hypothesis based on research, analysis and data:
Our hypothesis should be informed by as much knowledge as we have which is backed up with research, analysis and data since the key difference between a well-formulated hypothesis and a guess is data.
As it is emphasized in the discover part of the problem space in Double Diamond, we should use the data and insights to identify problems.
We should validate our problem statement by using the insights we gain from qualitative and quantitative research. Often, it can be tempting to make assumptions when thinking about the user experience. However, it’s important to use data to confirm the issue we are trying to solve.
The best is to combine qualitative and quantitative research insights (like insights from user interviews, heatmaps, surveys, customer feedbacks, analytics data etc.) and to generate solid, data-driven and research-based problem statements and hypotheses.