Getting Your Prices Right

Determining Optimal Price Levels For Core Fare Products

A Web-based, integrated decision-support prototype for a new “proactive-pricing” workflow is currently being developed that has potential to help airlines determine optimal price levels in markets subject to an airline’s desired price ranges for each core fare product.

One of the most important aspects of managing an airline is setting the right price levels for the target market segments. When setting fares for an upcoming season, airline pricing analysts typically create a wide variety of fare products with different restrictions and refundability to divide customers into traveler segments such as business (willing to pay more for flexibility, ability to book close to departure and desirable flight times) and leisure (more price sensitive but can book farther in advance and on less desirable flights).

Then the price levels for these different fare products must span a wide range to provide high prices on peak flights (and strong-demand dates of the season) but also discount fares that are low enough to generate sufficient demand to fill seats on off-peak flights (and low-demand dates of the season).

This segmentation and multiplicity of fare products helps airlines obtain higher total revenue. However, managing these fares creates ongoing challenges to ensure that current prices are consistent with the airline’s pricing strategy and are marketplace competitive.

Like any business, market prices in the airline industry change continually. Even if an airline initially does a good job of designing the right fare products and correctly setting price levels, airline pricing analysts have the ongoing work of continuously monitoring their competitor price levels and making required updates to products and price levels.

This competitor positioning is particularly important in the information-rich travel industry. Referring to the strong degree of price competition in the industry, Robert Crandall (former president and chairman of American Airlines) in a 1992 issue of Time Magazine remarked: “This industry is always in the grip of its dumbest competitors!”

Unfortunately, the sheer volume of changes in airfares (millions of published fares change daily) necessitate that the overwhelming focus of the airline pricing analyst workflow is on matching competitor fares. This information is further detailed in “Pricing decision support: Optimising fares in competitive markets,” 2009 Journal of Revenue and Pricing Management, Vol. 8, Issue 4, authored by B. Vinod, R. Ratliff and C.P. Narayan. Focusing exclusively on reactive pricing means that, over time, an airline’s price levels may drift away from the right prices needed to maximize revenue.

Higher Revenues Via Multi-fare Approach

Airlines typically use multi-fare product differentiation to extract greater revenues than a single-price strategy can provide. Comparing the demand curves in the two figures above, it is clear that the multi-fare approach (right) provides higher total revenue. In these figures, revenue = price x demand.

To help airlines find the right price levels, Sabre Airline Solutions® is working on a Web-based, integrated decision-support prototype for a new “proactive-pricing” workflow that does not exist in most airlines today.

The proactive-pricing prototype is used to determine the optimal price levels in a market subject to overall market share targets and desired price ranges for each of the airline’s core fare products (along with other practical constraints). It helps airlines answer the business question, “What is the right price to charge”?

While reactive pricing focuses on a response to a specific fare action by a competitor, strategic pricing is a proactive business process to file new price levels for a market based on the desired pricing objectives and constraints in the presence of competition.

Strategic pricing should encompass a general structure that captures the key analysis operations, automates those activities and provides an optimized pricing strategy. It should aid pricing analysts in developing an effective pricing strategy that takes into account interdependent factors, past trends, current market conditions and pricing goals.

The proposed approach to strategic pricing works by considering historical revenue associated with a previous fare season, including the observed fare levels, sales volumes and market share. Non-linear optimization models are then used against that historical season to find improved prices that would have increased the airline’s revenue outcome.

These models consider price elasticity to help users better estimate the sales and revenue impacts of fare changes together with business constraints and various competitive-response scenarios in determining the best solution. The results of the analysis can be used to help make decisions about upcoming seasonal fare levels.

The microeconomic theory underlying strategic pricing models is not complicated. The basic principles are:

  1. If price increases, sales quantities go down. If price decreases, sales quantities will go up.
  2. If relative prices change, quantities will shift to the airline with a relative price decrease.
  3. If demand is price elastic, the airline will increase revenue by lowering price and receiving a larger than proportionate increase in volume. If demand price is inelastic, the airline will increase revenue by raising price and receiving a smaller than proportionate reduction in volume.

The above holds true if there are no capacity constraints and the amount of the price change is small.

Non-price factors are another important consideration (e.g. airline brand loyalty, flight frequency and timing, on-time performance, etc.). Rather than estimate all these factors separately, Sabre Airline Solutions has developed a proprietary approach that indirectly estimates the importance that customers place on the non-price effects based on observed airline prices and market share. For the strategic-pricing decision, this approach is possible because the only factor being varied is the price itself (not the schedule or other factors).

Method For Determining The Right Prices

The pricing analyst reviews each of the competitive results to find the projected outcomes under each competitive scenario. Then, the analyst evaluates the results according to the likelihood of each scenario arising in practice. Finally, the analyst decides which price changes to make, considering all risks and rewards.

When optimizing fare levels, numerous practical considerations exist that constrain the strategic pricing solution. Examples of such constraints include:

  1. Preserving the existing price differentials (fare ratios) between the various fare products in a market (for price consistency),
  2. Limiting the degree of price change to stay within user-specified upper and lower bounds.

Such bounds are important because setting prices too high may create an undesirable price image with respect to the competition, whereas setting them too low may result in lost margins, price wars and eventual brand erosion.

Another constraint to consider is revenue-management availability across a season; it is needed in fare optimization to avoid making price drops that would over-stimulate demand relative to the available capacity.

In our experience, based on work conducted with Prof. Guillermo Gallego at Columbia University, the microeconomic model is further enhanced by considering another important tool known as game theory (the study of mathematical models of conflict and cooperation between intelligent rational decision makers). For this discussion, we will refer to the airline being optimized as the host airline. We propose a few basic assumptions:

  1. Airlines behave rationally. That is, if they can take an action that benefits them, they will do so. The objective function for airlines is to maximize revenue.
  2. We use a “conjectural variation” model, which means that the host airline optimization is based on making our best guess at the response of other airlines to any fare changes we make. If the host’s guess is accurate, it is called a consistent conjecture.
  3. Competitors respond in a manner known as “bounded rationality,” meaning that if prices are changed, the competitor response will be in the same direction of the change (ranging somewhere within a “no match,” “full match” or “partial match” continuum). Cases where competitor response is the opposite of the host price change (e.g. they increase prices in response to a price drop) are not considered.
  4. Since we are dealing with proactive price changes, the host airline moves first. Other airlines will respond to the host airline’s move.

Based on the above assumptions, it is possible to use the microeconomic model to estimate the revenue impact of fare changes for both the host airline and the competition.

However, one of the first lessons learned by airline pricing analysts in practice is that it is often difficult to predict competitor responses to fare actions. Thus, a crucial aspect of any proactive pricing-decision-support tool is to evaluate the expected outcomes under a range of different possible competitor responses.

Airline Strategic-Pricing Decision Support
  • Expects to helps airlines set the right price levels for an upcoming season.
  • Together, with recommended price changes, should provide estimated revenue changes under a range of different competitor responses (a first step toward applying game theory in practice for airline pricing).
  • Should consider other basic constraints such as:
    • Target price ranges for different fare products,
    • Price elasticity (sensitivity of demand to price),
    • Practical constraints on the degree of the price changes that can be made,
    • Available seats for the upcoming season (from airline capacity and revenue management availability).

Using our strategic-pricing framework in practice, we have found it useful to summarize the optimization model results in the following manner. The optimization model is solved once for the “most likely” case (i.e., the conjectural variation in which the host airline has made its best guess with respect to another airline’s response). This provides a base case result to compare against other, alternative competitor responses, providing:

  • Most likely — Presents what will happen if the host picks optimal prices based on the conjectural variation, and the host airline happens to be correct (i.e., under consistent conjectures).
  • Competitor no match — Represents what happens if the host airline picks optimal prices based on the conjectural variation but is wrong about other airlines’ responses, and the other airlines do not match the host fare change at all.
  • Competitor full match — Represents what happens if the host airline picks optimal prices based on the conjectural variation but is wrong about other airlines’ responses, and the other airlines fully match the host fare change.
  • Competitor optimal — Represents what happens if the host airline picks optimal prices based on the conjectural variation but is wrong about other airlines’ responses. In this scenario, other airlines choose prices according to what is known in game theory as a “best-response model” to maximize their revenue given the host airline’s new prices. This is what rational competitors would do.
  • Three-round match — Represents what happens if: 1) the host optimizes its prices, 2) the competitor responds optimally, and 3) the host matches competitor responses. This type of three-round behavior is commonly observed in practice.

Other possible scenarios could also be constructed, but the idea is to show how the expected revenue of the price changes would vary under a wide range of potential competitor responses. Understanding how the revenue changes depending on the competitive scenario is important for pricing analysts to gauge the risk of a proposed price move.

A successful strategic-pricing model should provide a programmatic approach to current manual processes used by airline pricing analysts and executives. It should also provide automated parameter estimation to facilitate setting values for items such as price elasticity, fare bounds, competitive match factors, etc.

Price Elasticity

Price elasticity (denoted Ep) is an economic measure that describes how demand changes with respect to prices. It is usually a negative number because price increases lead to demand drops (and vice versa). For example, if Ep = -1.2, then a -1 percent price reduction results in an estimated 1.2 percent increase in demand. Ep varies by market, and airline revenue-accounting data is usually used to estimate it. Ep is used to classify the demand type as follows:

  • Price inelastic demand (-1 < Ep < 0): implies that demand is relatively insensitive to price changes (e.g. business customer segments), and revenue increases if price is increased.
  • Unit elastic demand (Ep = -1): a -1 percent reduction in price results in a 1 percent increase in demand.
  • Price elastic demand (Ep < -1): implies that demand is price sensitive (e.g. leisure-customer segments), and revenue decreases if price is increased.

With these parameters, models can be used to identify potential opportunities to fine-tune price levels and estimate revenue performance across a range of possible competitor responses. Ultimately, these tools facilitate proactive pricing moves while maintaining their market constraints and policies.

In our applications experience, we have found that when these microeconomic and game-theoretic tools are applied to historical results, most markets show at least some benefit from fare optimization. The estimated revenue impacts vary by market and season, but in practice, it is typical to find opportunities ranging from ½ percent to 3 percent additional revenue.

Model-based decision-support tools are also useful in helping rank different markets in terms of expected revenue improvement so pricing analysts can spend most of their time and energy on the ones with the highest expected benefit.

Proactive pricing strategy is a new challenge to airlines to provide better insights on the impact of price change to customers, partners and competitors. Decision-support tools facilitate analysis and free up pricing analysts from tedious manual work so they can spend more valuable resources on analyzing various scenarios, making more insightful decisions and generating incremental revenues for airlines.

Table of Contents