Time series and forecasting · insight-0248
Model spill and unconstrained demand before estimating gains from added capacity
- Evidence type
- practitioner heuristic
- Difficulty
- hard
- Roles
- Data scientist, Data analyst
- Industries
- aviation
Summary: situation, action, reason, boundary
Situation
A planner asks for passenger gains if the aircraft on a route changes from 160 to 180 seats.
Action
Forecast unconstrained demand per departure, compute expected extra passengers as the expected demand above 160 capped at 180, and account for how fare controls would change with more seats.
Reason
Extra seats help only on departures where demand exceeds the old capacity, so the gain depends on the upper tail of the demand distribution, not its mean.
Boundary
Capacity planning in airlines, hotels, events, staffing and cloud provisioning.
Symptom or risk
Using a forecast of passengers flown, which was capped at 160, implies no gain or a naive gain of 20 on every flight.
Exceptions
If spill is small on almost all departures, a quick tail estimate from historical load factors may suffice.
Trade-offs
Unconstraining assumptions drive the answer; show results under several of them.
Pitfalls
Ignoring that more capacity can change pricing and competitor responses, which turns the question causal.
Example
Hypothetical: if 15% of departures would have demand above 160, gains come only from those departures.