Whenever you're ready, walk me through how you'd approach this.
How would you design a research study to determine whether feature X is impacting metric Y?
Explain how you would define the causal question, select participants, establish a comparison group, combine qualitative and quantitative methods, and account for confounding factors. Describe how you would interpret conflicting evidence and decide whether to iterate, scale, or remove the feature.