What is a causal insight in the context of testing?

• A. A conclusion drawn from testing that shows how a specific variation affects customer behavior
• B. A general consumer preference opinion
• C. A correlation observed in data without manipulation
• D. A forecast based on historical data

Answer: (B)

Causal insight comes from experiments where you deliberately vary something and observe the effect on behavior, so you can show that the change in the variation caused the change in outcome. In testing, you’d manipulate a factor (like price, layout, or a call-to-action) and measure how that manipulation changes a specific metric (such as conversions), ideally using random assignment and control groups to rule out other influences. This is what lets you claim a cause-and-effect relationship, not just a correlation or a guess.

General consumer preference opinions aren’t proven causes, they’re attitudes or tastes. A correlation observed without manipulation shows an association but doesn’t prove that one thing caused the other. A forecast based on historical data predicts future values but doesn’t demonstrate that changing a specific variation will cause a different outcome.