Most airlines have embraced the practices of differential pricing and yield management over the past decade. By offering a range of fare options at different price levels on the same flight, airlines attempt to segment the total demand for air travel according to the different sensitivities to price and passengers' differing need for travel flexibility. To ensure that low-fare leisure passengers do not consume all the seats on high-demand flights, airlines employ yield management (YM) systems that forecast demand and calculate the number of seats to be made available to each fare type, with the goal of maximising total flight revenues.
Revenue gains of around 2-5 per cent are commonly quoted for an airline that implements a basic YM system. These revenue gains have been estimated through the use of simulation and real-world experimentation, and have focused on the booking patterns and YM capabilities of a single airline, independent of its competitors. This study is the first known effort to estimate the impact of airline yield management in competitive markets, taking into account the YM capabilities of competing carriers.
The simulation employed in this study represents an integration of Boeing's Passenger Origin-Destination Simulator (Pods) and the MIT Flight Transportation Laboratory's yield management simulation capabilities. The Boeing/MIT simulator was developed to help understand the interactions between passenger choice and airline yield management actions. It provides a realistic test environment for comparing the impacts of different YM capabilities, both on individual airlines and on the overall market.
The simulator performs many iterations of a complete booking process for a series of departure dates, which can most easily be interpreted as a large number of departures of the same flights, at the same times, on the same day of the week.
The fare product structure is intended to be representative of a typical US domestic airline market, with four fare options ranging from the highest priced unrestricted 'full coach fare' to a 14-day advance purchase non-refundable excursion fare (see table). In the simulation, the differing sensitivities of business and leisure travellers to prices and restrictions, along with different rates of booking during the period immediately before departure, determine the proportion of each group that will choose each fare product.
The simulation uses a YM system which reflects the most commonly used techniques for forecasting demand and optimising booking limits. In the absence of YM, an airline is assumed to use a 'first come, first served' (FCFS) passenger acceptance rule, under which it accepts all booking requests until the total capacity of the flight leg is reached. Note that, even under FCFS, some passengers will end up purchasing higher priced fare types, since the advance purchase and minimum stay conditions of lower fares might prevent these passengers from taking advantage of the lowest available fares.
Our analysis was limited to a single nonstop market in which only two airlines compete. The level of demand in the market was set initially to 90 per cent of the total daily seating capacity offered by both carriers combined - high enough to ensure that the use of YM capabilities would have an impact on the acceptance of passenger booking requests.
This 90 per cent 'demand factor' translates into aircraft load factors of about 80 per cent because some passengers cannot find a combination of flight and fare that meets their preferences and therefore choose not to fly on that day. As a result of this, the simulation scenarios are representative of a typical peak day of the week or peak seasonal period.
The two competitors are assumed to be identical in all respects, other than in their use (or not) of a YM system and, in later scenarios, in the flight frequencies and departure times they offer. As the base case, we assume that neither carrier makes use of a YM system. At the next level, we allow one of the carriers to use a YM system, while the other continues to operate without yield management. Finally, both carriers are given identical YM systems.
Identical schedules
In the first and simplest scenario, each of the two identical airlines offers one daily flight at the same peak-period departure time (eg 1700). Both airlines operate identical 100-seat aircraft. In this completely symmetrical environment, the only difference between the two competitors that could affect passenger choice and, in turn, load and revenue performance, is seat availability.
Graph 1 summarises the revenue impacts of YM system implementation over the base case in which neither carrier has yield management capabilities. When one of the carriers (Alpha Airlines) is given a YM system, that carrier realises a revenue gain of an average of 8.2 per cent. This value is not out of line with previous single-carrier simulations of the benefits of yield management, given the relatively high demand factor assumed. However it is more significant that the carrier without a YM system (Beta Airlines) saw its total revenue decrease by 1.4 per cent. The total market revenue increased by 3.4 per cent.
These results illustrate two distinct impacts of YM system implementation by one of the carriers in a competitive airline market. The innovating airline increases its revenue by protecting seats for later-booking high fare passengers and by limiting the number of seats sold at the lower fares. On the other hand, the airline without YM is forced to accept not only the low-fare passengers that it carried previously, but also the passengers that cannot find low-fare seats on the carrier that is now practising yield management.
Furthermore, the carrier without YM now has more departures on which it runs out of space entirely, preventing it from carrying its share of late booking high-fare passengers. Alpha Airlines, with its YM seat protection capabilities, is able to accept these late booking passengers.
When both carriers are given the same YM capabilities, both carriers see very similar increases in their total revenues, and in turn the total market revenue increases by 8.3 per cent.
When the daily demand factor changes from 90 per cent, differences arise in the relative revenue benefits of YM (Graph 2). At a market demand factor of 80 per cent, the benefit to Alpha Airlines as the sole user of a YM system was simulated to be 3.1 per cent - lower than the simulated gain at a 90 per cent demand factor, as should be expected. The simulated revenue gain at a much higher 100 per cent demand factor was also much higher for the single YM innovator, at 21.8 per cent over the base case.
The most interesting differences in revenue gains occurred when both carriers were given YM systems, at the different demand factors. At an 80 per cent demand factor, the revenue gain to each carrier was actually greater than that realised by a single carrier with YM. This is because at lower demand factors average load factors are lower, allowing more passengers to obtain Q-class seats on the competing airline that does not have YM. Alpha Airlines still sees a revenue gain of 3.1 per cent, but also loses some passengers who can find Q-class seats on Beta Airlines (which has no YM). When both carriers have YM systems, both carriers can limit the sale of Q-class seats more effectively, forcing more passengers to purchase a higher fare class.
There is a different result at the much higher 100 per cent demand factor. At 21 per cent, Alpha's revenue gain is greater when it alone uses a YM system, and decreases slightly to 17.6 per cent when both carriers use YM. At very high demand levels, the carrier that uses YM alone can shift much more Q-class passengers to its competitor, and can preclude the competitor from carrying higher fare class passengers to a greater extent. In turn, Alpha Airlines can use its YM system to protect seats for late-booking high fare passengers and improve its own fare class mix even more.
When both carriers have YM systems in such a case, this ability to fill the competitor's plane with low-fare passengers does not exist, since both YM systems are working to protect seats for higher-fare passengers.
These findings indicate that the revenue benefits of a YM system to an airline can differ not only by demand factor, but also as a function of the YM capabilities of the airline's competitors. Although it is clear that use of a YM system results in revenue gains to the implementing airline over a range of demand factors, our simulation results suggest that it is very difficult to make broad generalisations about the specific percentage revenue benefits of a YM system, without taking into account the YM capabilities of all airlines in the market.
Frequency advantage
In the second scenario we allowed one of the airlines, the 'stronger' carrier, to provide two daily flights, versus only one flight per day for the 'weaker' airline. The stronger carrier is able to offer flights at the morning and evening peak periods of demand, 0900 and 1700.
We then examined the impacts of two different placements of the weaker carrier's single daily departure: under the 'distinct schedule', the weaker carrier departs at 1300, exactly half-way between the departures of the stronger carrier, while under the 'overlap schedule' the weaker carrier places its departure at the same time as one of the competitor's peak flights (in this case as it happens, also at 1700).
There were significant interactions between YM capabilities and schedule. Under the 'distinct schedule', giving either of the carriers a YM system caused its own revenues to increase, by 9.3 per cent for the stronger carrier and by 8.5 per cent for the weaker (Graph 3). The carrier with the frequency advantage gains to a slightly greater degree.
At the same time, the negative revenue impact of not having a YM system is greater for the carrier with the lower frequency. When both carriers are given the same YM capabilities, their revenues, along with those of the total market, increase by about 7.5 per cent over the base case.
The results are substantially more dramatic, and much more interesting, for the 'overlap schedule' scenario (Graph 4). In this case, giving the stronger carrier a YM system caused its revenues to increase by 26.8 per cent and those of its weaker competitor to decrease by 4.8 per cent. On the other hand, the use of YM by the weaker carrier increased its revenues by about the same as when it operated a distinct schedule, and hurt the stronger airline similarly in both models.
When both carriers are given the same YM system, their revenue gains are dramatically different: 26.9 per cent for the stronger carrier, but only 2.2 per cent for the weaker carrier.
The placement of the weaker competitor's flight departures explains the differences in the revenue gains. When the stronger carrier implements YM, its schedule advantage can translate into higher revenue gains for itself and greater revenue losses for its competitor. With two departures 'bracketing' the single departure of the weaker carrier, the stronger airline can 'dump' Q-class passengers from both its flights onto the airline with no YM, and save seats for late-booking high-fare passengers on its peak departures, for which there is (by definition) a high demand.
When the weaker carrier is given YM on its own, the 'dumping' occurs towards both of its competitor's flights, with a much less obvious negative revenue impact for the stronger carrier. In the 'distinct schedule' scenario, each of the three daily departures effectively has its own 'base' level of demand that will prefer to fly at that time of day, which also helps to reduce some of the negative revenue impacts of not having a YM system.
The story is very different for the 'overlap schedule'. Here, the stronger carrier is much more effective at 'dumping' unwanted Q-class passengers directly onto the flight of the weaker competitor without YM, since this flight is scheduled at the same time. In the other peak (0900), the stronger carrier can make very effective use of YM tools without losing much traffic to the flight of the weaker competitor, which happens to be full of low-fare passengers anyway.
When the weaker carrier is the only airline with a YM system, it can similarly 'dump' unwanted Q-class passengers on the matching 1700 departure offered by the stronger competitor, but this does not have much impact on its competitor's 0900 departure.
The carrier with the stronger schedule, and a distinct period of the day in which it provides the only departure, gains more from YM implementation and is hurt less by not having YM than its weaker competitor. The implementation of YM still benefits the carrier with a weaker schedule, even though it might not hurt the stronger carrier as much.
It is especially important for the weaker carrier not to be both in a position of schedule inferiority and at a YM system disadvantage, as both these two conditions can interact to create substantial negative revenue impacts.
The results of our simulations led to clear and consistent findings - there is a substantial revenue advantage for the airline that initiates yield management first in a competitive market. The gains come from carrying a better fare mix of traffic and, especially, from protecting seats for late-booking high fare passengers.
More interesting was the finding that the airline without YM can be hurt in terms of revenue by not having one. The carrier with a YM system is able to 'dump' unwanted low-fare passengers onto the flights of the carrier without YM, filling up that carrier's capacity and leaving it without adequate last-minute seat availability for the highest-fare passengers.
Revenue increases
When both competitors were given YM capabilities, both were able to achieve revenue increases, indicating that yield management is not a 'zero-sum game'. These revenue gains come from forcing consumers to pay fares closer to their willingness to pay, rather than from any market-share shifts between the various carriers.
Finally, we found that there are important interactions between yield management and airline flight schedules in a market. In all cases, the weaker carrier in terms of schedule still benefits from using YM, but it can also be hurt more by not having YM than the stronger carrier in the market.
In the end, it is very clear that even the best YM cannot overcome major schedule disadvantages; it is a 'tactical' tool. However, it is just as clear that an airline, whether strong or weak, which operates without a yield management system against a competitors using YM is not just missing out on revenue potential but could well find its own revenues reduced by its inaction.
Source: Airline Business
