What is the first step in formulating an A/B test?

• A. Collect data before any hypothesis
• B. Develop a strong hypothesis about user behavior and optimization opportunities
• C. Set a fixed duration regardless of data
• D. Immediately run a test with multiple variations

Answer: (B)

The main idea is to anchor an A/B test in a clear, testable hypothesis about how a change will affect user behavior and which metric you’ll improve. Having a well-defined hypothesis gives direction: it tells you what to change, what you’re measuring, and what counts as a successful outcome. Without this, you might collect data without a specific question or test changes that don’t address a real opportunity, making results hard to interpret.

For example, you might hypothesize that changing the call-to-action color from blue to green will increase click-through rate on the product page by at least 5%. That kind of statement guides which variation you build, how you design the experiment, and how you determine the necessary sample size and analysis.

The other options don’t fit because collecting data before forming a hypothesis is not the formal first step in test design, setting a fixed duration ignores statistical power and data-driven stopping, and jumping straight into multiple variations without a guiding hypothesis risks testing things that don’t answer a specific question.