Uber is running marketing ads on YouTube and wants to understand whether the campaign creates incremental business value, not just whether users watched or clicked the ads. Answer the following: 1. What success metrics would you use to evaluate the campaign? Include both marketing-funnel metrics and business-outcome metrics. 2. How would you design a measurement approach to estimate the causal effect of the YouTube ads? 3. If you compare treatment and control groups using a t-test, what assumptions must hold? 4. How would you conduct a power analysis before launching the test? 5. Suppose average order value has extreme outliers and the distribution is highly right-skewed. Can you still use a t-test? If not always, what alternatives or adjustments would you consider?