High average load factors are leading US airlines to clamp down on multiple bookings which play havoc with their yield mix. But some more advantageous solutions should be considered. By Bill McKnight, Geoff Murray and Patrick Meynial.

Plan ahead. Leave yourself plenty of time. Always check in at the airport at least one hour before the scheduled departure time.

For many professionals all this is easier said than done. Often they cannot finalise their itinerary until minutes before heading to the airport. Meetings overrun, are cut short or have to be completely rescheduled.

Recognising this many airlines offer multiple flights in many high-volume markets. For instance, both United Airlines and American offer hourly service between New York/La Guardia and Chicago/O'Hare. A traveller can simply select one of these hourly flights, rather than adjust his or her schedule to fit an airline timetable. High-fare passengers pay dearly for the luxury of such flexibility.

There is a problem, however. In 1996, airlines saw their load factor increase to 70 per cent, from 62 per cent in 1993. As a result flights are heavily booked and seats may only be available in advance. The problem is significantly worse on Fridays, when average load factors are 4.5 percentage points higher, according to estimates by Salomon Brothers.

In some markets, such as between New York, Washington and Boston, airline shuttle services are geared towards last minute passengers. Shuttle operations are expensive since they require additional crews. However they use identical equipment and serve a small set of cities so that downline equipment routing difficulties are unlikely.

Where such shuttle operations don't exist, a simple solution is for passengers to book themselves on multiple flights around the time they plan to depart. A passenger intending to make the 5:30 flight from LaGuardia to Chicago may also book the 4:30 and the 6:30, thus covering the options if a meeting ends early or runs late.

Travel agents used to accommodate such uncertain passengers by booking them onto several flights and elite members of frequent flyer programmes were occasionally afforded that benefit when booking seats in airline club class lounges. Such 'double booking' was not officially permitted by airlines, but they rarely pursued violators.

No longer. Last year's big profits and high load factors have changed the landscape. Airlines are cracking down on multiple bookings and travel agencies are unlikely to recommend them, much less actually make them. In recent months travel agencies have seen an increase in debit memos resulting from multiple bookings and airlines have formally notified travel agencies that they will systematically charge a booking fee for such reservations.

In an 18 April article, the Wall Street Journal revealed that airlines intend to crack down on the number of double bookings placed by the public and travel agencies by cancelling duplicate reservations and routinely issuing debit memos to offending agencies.

The alternatives for the traveller are limited. Passengers can book seats on different airlines or arduously create single reservations for each of the flights they want bookings on. They can also stand by for flights, knowing that they may not get a seat or that they could end up in a back row centre seat. The odds that such a bet will pay off are decreasing - the denied boarding rate for passengers holding a reservation increased 15 percent a year from 1992 to 1994 and is expected to accelerate, leaving even less chance for standby passengers to get on board.

The ill-advised double bookings are mostly made by full-fare passengers. In the US they number 73 million, representing 14 per cent of all US domestic passengers. The majority of these travellers are frequent flyers with at least one airline.

Typically these passengers and travel agencies need to make speculative bookings on between 10 and 50 per cent of their flights. An analysis of O&D data collected by the US Transportation Department, shows that roughly 25 per cent of US majors' domestic traffic is between city pairs where at least one major airline offers a minimum of five flights per day. This means that roughly between 2 and 7 million enplaned passengers may be affected by the multiple booking issue.

It is relatively easy for airlines to issue debit memos, cancel duplicate reservations and pursue agencies that make such reservations. But could airlines take more advantage of the need for multiple reservations?

In these years of profitable growth, the impact of the current approach is probably limited for the airlines concerned. Despite a slightly negative image for the industry as a whole and the loss of a few passengers to competing airlines, the 'bottom-line' impact should be minimal.

Airline perspective

Before the advent of computer reservation systems, flights would become closed to reservations as soon as the number of reservations equalled the number of seats on the aircraft. If a passenger with a reservation did not show up, the flight would depart with an empty seat and the airline would lose the revenue associated with a potential additional passenger. Airlines later compensated for such uncertainty in the 1970s with the advent of yield management systems and overbooking models.

Collectively, yield management and overbooking models manage the capacity of seats on flights and between city pairs. The ultimate objective is to fill an aircraft with the most profitable mix of passengers without refusing boarding to anyone. This mix is often made up of low-yield passengers, who book a non-refundable ticket long before the departure date, and a smaller number of people who book unrestricted tickets at the last minute. A yield management system continuously calculates the optimal mix of each flight based on probabilities and expectations, while an overbooking model compensates for passengers who may not show up. However a fundamental assumption of these models is that a passenger will only book one flight - or sequence of flights - rather than a set of flights.

Multiple booking is therefore viewed by the airlines as an undesired complexity in the calculation of both overbooking levels and fare class limits. Multiple bookings increase the requirement for overbooking, therefore diminishing the accuracy of the forecast. This loss in accuracy may be significant and will carry a cost. Airlines could forego revenue on three fronts: they are more likely to refuse boarding to passengers holding a reservation, they are more likely to fly with empty seats, and they are more likely to fly with a poor mix of high and low-fare passengers.

New options

In spite of the complexity, there may be a better alternative to the 'cancel and punish' approach to multiple bookings. Airlines may be able to either:

1 Guarantee a seat on any flight to all full-fare passengers, much like the East coast shuttle flights do.

2 Permit multiple bookings and cover the associated expense of such bookings with across-the-board fare increases.

3 Charge a premium for passengers to book on multiple flights, above the regular coach air fare.

While all of these alternatives have associated costs and benefits, the third option may be the most feasible and attractive.

Airlines could allow speculative bookings by requiring a passenger to book a full-fare ticket on the flight they actually intend to travel on while permitting additional bookings, perhaps in a lower fare class, on other flights. The passenger would pay a premium, indexed to the cost of the full-fare ticket, for the privilege of making such a booking. Airlines could also create a new sub-class of ticket - probably within 'Y' class - that would enable the passenger to make a set number of bookings in any 24-hour period. This class of ticket would sell at a premium above the normal full economy fare.

For example, in the New York/LaGuardia-Chicago/ O'Hare market, a one-way ticket costs approximately $480 and an advance purchase ticket is $110, the latter based on a roundtrip. When making a booking, a passenger would book in 'Y' class - at the $480 rate -- for his/her actual intended flight, and could create additional bookings on other flights in 'Q' class. The most obvious disadvantage in this example is the restriction to 'Q' class flights. A traveller requesting multiple bookings usually does so on flights that are likely to be full, that is non 'Q' class flights. The restriction to 'Q' class flights reduces the value of the multiple booking so the passenger may have to trade the comfort of a forward aisle or window seat for that dreaded centre seat in the last row.

How much would travellers be ready to pay for that convenience? In cases where the travellers want unrestricted multiple bookings, both the cost to the airline and the value to the traveller require complex calculations. One possible approach to calculate the cost to the airline is through yield management and overbooking models, where each aircraft is assigned a continuously updated expected marginal seat revenue or expected net revenue (ENR), according to a 1994 study published by Paris-based Institut des Transports Aériens. Using the ENR methodology, A T Kearney estimates that an airline should charge a 6 percent premium for the right to double book.


How will customers react to the introduction of double booking 'rights'?

Some passengers are clearly already placing multiple bookings and airlines will need to quantify the percentage of passengers who would take advantage of the programme. Estimates could be gleaned from an airline offering double booking in a particular market, perhaps a relatively low-yield, high-frequency market where the potential for lost revenue associated with overbookings or no-shows is low. An excellent market for such an experiment would be low-yield, low-load and high-frequency, but fortunately few of those exist.

From a systems perspective, implementing such a programme would be similar to that for coach class upgrades to first class. Delta Air Lines has introduced a fare into its system which permits passengers to pay a premium and obtain a confirmed seat in first class, instead of waiting at the gate for an upgrade. A new fare category with a separate set of rules would form the foundation for the implementation of a multiple booking class.

On full-fare tickets, passengers routinely pay a premium of at least $50 for the convenience of a more comfortable first class seat. Therefore passengers are likely to pay more than an extra 6 per cent for the convenience of additional reservations.

By refining some of these ideas, airlines may find compelling arguments to change their strategy towards multiple bookings, and collect additional revenues and profits while providing ever improved services.

Source: Airline Business