What is hypothesis-driven testing?

Hypothesis-driven testing is a CRO approach where changes are made and tested based on a specific, evidence-based hypothesis rather than assumptions or personal preferences. The goal is to determine whether a proposed change produces a measurable improvement in user behavior or business outcomes.

The process generally involves:

1. Identify a problem: Use analytics, user feedback, heatmaps, or other research to find a potential conversion issue.
2. Create a hypothesis: Explain why the problem may be occurring and what change could address it.
3. Define the expected outcome: Specify the metric that should improve.
4. Run a test: Use methods such as A/B testing to compare the existing experience with the proposed change.
5. Analyze the results: Determine whether the evidence supports the hypothesis.
6. Apply the learning: Implement the change when appropriate or use the results to develop another test.

For example, if a B2B website has a low Request a Quote rate, a hypothesis might be that the form is too lengthy. The company could test a shorter form and measure quote-request conversion rate and lead quality.

This approach helps CRO teams make decisions based on evidence and measurable results rather than subjective design opinions.