Overbooking, forecasting and optimization methods and systems

Abstract

A computer based system for maximizing revenue by determining an optimal quantity of a product to be sold is disclosed. The system determines the optimal number of seats to be sold for a flight based upon the flight's capacity and forecasted costs associated with the flight. The forecasting is based upon probabilistic distribution models takes into account passenger itinerary data, passenger and market historical data, whether a passenger has flown on a previous leg of an itinerary, the ripple denied boarding effect of reaccommodating a denied passenger. The system evaluates the potential effect of double selling a unit of inventory (e.g., seats). Downstream inventory control, revenue management and reservations systems may use the optimization data to affect the operation of the airline.

1 . A system comprising:
 a network interface communicating with a memory;   the memory communicating with a processor for enabling revenue maximization;   the processor, when executing a computer program, executes operations comprising:
 forecasting, by the processor, a spoiled seat (SS) cost for each seat in a plurality of seats associated with a flight; 
 forecasting, by the processor, a denied boarding (DB) cost for the flight; 
 determining, by the processor, a booked passenger no-show forecast (NSF) for each booked passenger associated with the flight, wherein the booked passenger NSF is a based upon whether the respective passenger flew on a previous leg of a passenger itinerary; 
 aggregating, by the processor, the booked passenger NSF and an unhooked passenger NSF to create a flight NSF; 
 determining, by the processor, an authorized seat allocation (AU) for the flight that minimizes an overbooking cost, wherein the overbooking cost is based upon an accumulation of each SS cost, the DB cost and the flight NSF. 
   
     
     
         2 . The system of  claim 1 , wherein at least one of a reservation system and a revenue management system uses the AU to determine a number of additional seats to be sold for the flight and a respective price for each additional seat. 
     
     
         3 . The system of  claim 1 , wherein the SS cost is based upon at least one of a current selling class of the flight and historic fares associated with the flight. 
     
     
         4 . The system of  claim 1 , wherein the forecasting the DB cost comprises determining a first probability that a first DB passenger will be involuntary or voluntary. 
     
     
         5 . The system of  claim 4 , wherein the forecasting the DB cost further comprises determining a second probability that a second DB passenger will be involuntary or voluntary, wherein the second probability is based upon the first probability. 
     
     
         6 . The system of  claim 4 , wherein the DB cost is based upon at least one of a non-compensation factor, a voucher amount, a voucher breakage factor, an expected percentage of volunteers, an ill-will factor, an involuntary DB cost, an expected accommodations cost, a double DB factor and the probability of voluntary denial. 
     
     
         7 . The system of  claim 1 , wherein the booked passenger NSF is further based upon at least one of the complete passenger itinerary of each respective passenger and an adjustment factor, wherein the adjustment factor is determined based upon historical NSF data. 
     
     
         8 . The system of  claim 1 , Wherein the determining the hooked passenger NSF comprises determining, for at least a subset of booked passengers, a conditional probability of a passenger showing for a second leg given the passenger flew a first leg. 
     
     
         9 . The system of  claim 8 , further comprising adjusting the booked passenger NSF in response to determining, for at least a subset of booked passengers, whether the first leg and the second leg are scheduled for the same day. 
     
     
         10 . The system of  claim 1 , further comprising determining a next active leg (NAL) for a first booked passenger, wherein the first booked passenger is one of a plurality of passengers associated with the flight. 
     
     
         11 . The system of  claim 10 , further comprising determining a first hooked passenger NSF as a probability that the first booked passenger will show for the NAL, wherein the booked passenger NSF is based upon the first hooked passenger NSF. 
     
     
         12 . The system of  claim 11 , wherein the NAL does not correspond to the flight. 
     
     
         13 . The system of  claim 1 , wherein the booked passenger NSF is based upon at least one of a previous leg, a subsequent leg and direction of travel for each booked passenger. 
     
     
         14 . The system of  claim 1 , the operations further comprising calculating a first expected marginal seat revenue (EMSR) for each first class seat on the flight and comparing the respective EMSR to a second EMSR for a potential sale of an additional coach seat. 
     
     
         15 . The system of  claim 14 , wherein the comparing the respective EMSR to a second EMSR comprises adjusting for the risk of double sell. 
     
     
         16 . The system of  claim 15 , the operations further comprising determining a coach upgrade parameter based upon the comparing. 
     
     
         17 . The system of  claim 16 , wherein at least one of a reservation system and a revenue management system uses the coach upgrade parameter to determine a number of additional seats to be offered for sale for the flight and a respective price for each additional seat. 
     
     
         18 . The system of  claim 16 , further comprising determining an optimal time for the additional seats to be offered for sale and making the optimal time available to the at least one of the reservation system and the revenue management system. 
     
     
         19 . The system of  claim 18 , wherein the determining the optimal time comprises determining that a coach achievable demand is less than the AU plus the coach upgrade parameter. 
     
     
         20 . A computer-based method, comprising:
 forecasting, by a computer for enabling revenue maximization, a spoiled seat (SS) cost for each seat in a plurality of seats associated with a flight;   forecasting, by the computer, a denied boarding (DE) cost for the flight;   determining, by the computer, a booked passenger no-show forecast (NSF) for each booked passenger associated with the flight, wherein the booked passenger NSF is a based upon data indicating whether the respective passenger flew on a previous leg of a passenger itinerary;   aggregating, by the computer, the booked passenger NSF and an unhooked passenger NSF to create a flight NSF;   determining, by the computer, an authorized seat allocation for the flight by minimizing an overbooking cost, wherein the overbooking cost is based upon an accumulation of each spoiled seat cost, the denied boarding cost and the flight NSF; and   updating, by the computer, an authorization parameter for the flight based upon the authorized seat allocation.