Define and validate an airline profitability metric

Your goal is to develop a profitability measure for airline routes that can support a PowerDay-style case presentation. Recommend one primary metric plus two to three guardrails that blend revenue outcomes with operational performance. It should break down cleanly by route and month and hold up under irregular operations. The data fields you can use are: route_id, flight_date, seats_sold, fare_usd, ancillaries_usd, fuel_cost_usd, crew_cost_usd, airport_fees_usd, block_minutes, delay_minutes, cancellations, refunds_usd, rebooking_cost_usd.

Tasks: (a) Specify the formula for your primary metric (for example, Adjusted Route Profit per Block Minute) and define each guardrail (such as cancellation rate, on-time arrival rate, or an NPS proxy when available). Lay out all assumptions, including how refunds and rebooking costs should be allocated, and explain the business rationale for the metric. (b) Describe how you would validate the metric using historical data: backtest it against prior route openings and closures, run sensitivity analyses for demand shocks and fuel price spikes, and assess its correlation with long-run cash contribution. (c) Imagine management tests a new policy, such as dynamic overbooking. Design an experiment or quasi-experiment that can detect an improvement in your primary metric while adjusting for seasonality, competitor actions, and weather. Cover the randomization unit, the inputs for a power analysis, the guardrails, and how you would interpret effects that differ across routes.

Overview: This question tests skills in metric design, business analytics, causal inference, and experimental design: it asks for a decomposable airline route profitability metric, operational guardrails, historical validation through backtests and sensitivity analyses, and an experiment plan for a policy change.