A further complication to seat availability, especially to the airline, is that each seat is a potential part of many “products”, and all these products compete for that seat. The same seat on a SEA to DEN flight could be used for a Y fare for a passenger traveling from SEA to DEN, or could be part of a Q fare from SEA to BOS with a stop in DEN. Selling a seat on one flight might fill that plane, making it impossible for someone else to use that plane as part of a bigger trip that might bring more money to the airline. So many routes and many fares are all competing for seats. This makes it much more challenging for an airline to decide whether or not to offer a given booking code on a given flight. A consequence is that many airlines demand information about the entire route of a trip before providing availability for any section: this is often called O&D (origin and destination) availability. The table shows how the seat availability for a single flight can vary depending on what trip the flight is part of. When the availability of two flights depends on the passenger taking both, the flights are said to be married.
Research in network revenue management is a very popular topic in the operations research (OR) community. Most work is based on linear programming models but simulation techniques are gaining in popularity. One fundamental result is that complicated revenue management techniques only significantly increase profits when planes are flying near capacity.