A theme park is piloting a virtual queue system intended to reduce average wait time by 20% while keeping per-capita revenue within −2% of baseline. Design the analysis and experiment: define primary/guardrail metrics (e.g., avg wait, throughput, guest spend, complaint rate), identify units of randomization and interference risks (families, ride-level spillovers), and propose an experiment design that mitigates network effects (cluster randomization by ride or time blocks). Estimate required sample size and test duration with key assumptions (arrival rates vary by hour; weather and weekends cause seasonality). Specify how you’ll monitor in-flight (sequential testing or pre-registered peeks) and your decision rules. After the test, show how you would segment effects (ride popularity, guest demographics) and compute the budget impact. If results are mixed (waits down 25% but revenue −3.5%), recommend a follow-up test (e.g., dynamic return windows or targeted upsell) and the minimal data you need to decide rollout.