INTRODUCTION
Origin and destination (O&D)-based airline revenue management systems have increasingly incorporated the use of bid price evaluation for the purposes of dynamically generating accept and reject decisions for seamless availability and booking transactions. These real-time evaluation requests flow through various distribution channels, the airline reservation systems, and ultimately interface with a revenue management system. The real-time evaluation process makes it possible to further improve revenue management practices as each individual transaction is eligible for scrutiny. Complex decisions can be based on individual comparisons of the calculated value of a transaction and the minimum bid price for the next unit sold.
DISTRIBUTION-CENTRIC EVALUATIONS
Real-time availability evaluation has its basis as the execution step that uses network optimisation bid price calculation and fare valuation data. This essential comparison not only allows for a revenue optimised response to be generated, but also permits for beneficial business rules processing. Based on the specific needs of a given airline, this comparison can be influenced by allowing for artificially expanded or restricted access to certain fare products. In addition, best of class revenue management solutions enable a number of business strategies to be employed in this step. The most specific of these business strategies, when considering impact to a targeted transaction, is the ability to influence by distribution channel. In practice, these distribution-centric influences enable airlines to bias availability for a given channel. Although online sales channels may have actually increased the channel power yielded by traditional Global Distribution Systems (GDS) companies, direct connect sales options and GDS alternatives may offer significant comparative savings. According to G2 SwitchWorks, the average airline-paid GDS distribution fee averages $12.50 on a per ticket basis (Trottman, 2005). When considering that direct connect and alternative sales channels allow for a fraction of these fees, focusing revenue management attention to distribution strategy challenges is a worthy endeavour.
REVENUE MANAGEMENT IMPLICATIONS OF POCKET PRICE
The quantifiable value of distribution-centric availability influences was first formally introduced by McKinsey consultants Michael Marn, Eric Roegner, and Craig Zawada in The Price Advantage (Wiley, 2004). In their vernacular, the pocket price is the amount of money a company actually 'puts in its pocket' from the sale of a good (ie net revenue). A price waterfall, or graphic depiction of the many different costs including, but not limited to, distribution costs, shows the many different costs that lead from the list price to the pocket price (ie gross revenue to net revenue). In the case of an airline, the list price should be considered a given fare or daily production fare, which reflects the calculated value of next seat availability for a specific O&D, fare class, and point of sale for a given flight. Those airlines who are able to account for these pocket price differences in their real-time availability processing step will benefit from correcting instances where a transaction would otherwise be accepted or rejected incorrectly based on an incomplete picture of the true cost to serve a given customer based on sales channel. This is an extremely important business concept to grasp as active management of an airline waterfall can yield up to several percentage points of revenue improvement.
CUSTOMER-CENTRIC EVALUATIONS
Once these distribution-centric availability corrections are made, the next logical step to consider is adding even more granularity to the price waterfall. The only way a transaction can be more accurately evaluated is to look not just at the GDS and travel agency identifier but to extend analysis to a specific customer segment level.
CUSTOMER SEGMENTATION
Customer segmentation is the formal analysis of customer data for use in a real-time availability evaluation tool. Ideally, customer segmentation will make use of various statistical techniques to help determine those attributes that are statistically significant. Analysis of variance (ANOVA), information criterion, and business expertise can serve as the bases for a segmentation study, but even this can be improved by other predictive models that can optimise a given segmentation to achieve specific goals. In terms of process, the customer segmentation identifies several attributes (eg customer lifetime value metric, frequent flyer elite miles balance, home airport, etc) and determines the attribute level of granularity (eg miles=0–9,999; 10,000–14,999, etc). This specific level of granularity per attribute type in a unique combination across all customer data points constitutes the completed customer segmentation.
CUSTOMER-CENTRIC APPLICATIONS
Once the segmentation is complete, whether using statistical models, business experience, or a combination of the two, specific actionable business strategies can be identified and utilised. Customer-centric strategies can be categorised into three types: revenue based, non-revenue based, and product or other reservation functions.
Revenue-based strategy
Revenue-based strategies may include demand stimulation activities for revenue passengers where non-revenue-based strategies focus on optimising reward booking and upgrade decisions. Many other strategies can also be employed, so long as reservations data moves through the real-time availability engine, customer-centric strategies can even include seat assignment criteria, product specific advice, and many other options. In the case of product strategies, airlines with differentiated product strategies (eg fare differences based on cost of miles, meal service, seat selection, etc) can optimise their direct channel displays based on those products most likely to achieve a high sales completion rate, also called 'win rate'.
Looking at the first use case, revenue-based strategies, customer-centric influences target specific customers and transactions by maximising incremental revenue while minimising dilution. This is achieved by receiving a unique customer identifier (eg frequent flyer number, corporate code, or other unique customer identifier) as part of the message received by the real-time availability engine. When evaluating the availability or booking request from the reservations system, the customer is matched with a pre-determined customer segment. Based on membership within a customer segment, business rules may be assigned in varying levels of applicability; searching first for very specific or small group rules and then searching for increasingly generic or universal rules. These rules serve to modify the network optimised fare value to effectively change availability levels for only segmentation identified passengers based on a variety of possible criteria including O&D, itinerary, fare class, sales channel, time based and other criteria.
Targeted sales strategy
In practice, revenue-based strategies using customer-centric influences are ideal for managing a large volume of targeted fare actions in place for a relatively short period of time. Marketing, sales, and pricing departments can envision any number of strategies but implementing these can become time consuming and impractical at a large scale. A system that enables users to manage a library of fare action strategies that can be customised by market offers the ability to expand greatly the number of targeted fare actions in the marketplace. As airline websites continue to regain distribution share versus online travel agencies (OTAs) and strive to build a direct relationship with the customer, the ability to offer customer segment specific pricing becomes increasingly important. New carrier-specific applications including Southwest Airlines 'The Ding' begin to show the possibility of customer-specific offers and how making these available in volume can impact customer loyalty and stimulate non-dilutive demand. Revenue management applications which support customer-centric pricing strategies can power shopping applications such as this. These highly customised segment-specific shopping tools can be based on complex rules including fare class, origin, destination, point of sale, and multiple other itinerary-specific criteria insuring the right offer goes to the right customer at the right time. Further applications of revenue-based strategies can also include a number of other marketing strategies including those where a customer sits at the edge of a milestone. For example, so called mileage runs, trips taken largely for the purpose of earning miles or credits toward the next elite status level, can be accounted for in the customer segmentation. In practice, customer-centric rules can be set to give an incentive for mileage run travel for only those customers who are near the next elite tier and have exhibited other consistent purchase behaviours. In this way a valuable customer segment is identified with clear activation or purchase goals in mind when establishing customer-centric rules in the evaluation response. In summary, revenue management applications that enable revenue-based customer-centric strategies offer airline organisations a significantly increased targeting capability making synergies between revenue management and customer database tools possible and profitable.
Non-revenue-based strategy
Non-revenue-based transactions include various transactions but are especially relevant to reward bookings and upgrades. When considering how rewards and upgrades are allocated, some simple segmentation is built in to most airline programmes. Using loyalty tiers, zonal miles requirements, and other criteria, airlines strive to place a value on reward and upgrade transactions that give some customers better access to these loyalty benefits versus others. An improvement to this is the use of segmentation data to offer differentiated levels of access to reward and upgrade inventory. Use of broad frequent flyer tier levels and zone-based values is replaced with a system where each customer segment can have differentiated access based on rules which may be applied based on flight information (ie O&D, itinerary, fare class, sales channel, etc) and also on customer criteria (ie year to date value metric, total miles balance, elite miles balance, etc).
Value-based reward allocations and upgrade clearance
In practice, so called non-revenue-based strategies using customer-centric influences are ideal for more effectively managing reward seat availability and upgrade clearance on a micro level. One of the most critical decisions made between revenue management and loyalty marketing can be the balance between revenue seat and reward seat availability. Specifically, how should reward seats be allocated when customer demand heavily exceeded reward seat availability? Revenue management systems with customer-centric evaluation tools can leverage customer segmentation information to create micro-tiers within existing loyalty programmes. In the case where the first level of a loyalty programme has the greatest participation, but the customers at the bottom and top of this tier have very different contribution values, micro-segmenting tiers allows for rewards programs to give preference to those customers who have a greater long-term value to the airline. Specifically, customer-centric strategies could be set to match reward seats to requests based on customer segments where criteria could include not just loyalty tier level but also customer value metrics, programme miles or elite balances, and other related attributes. This in combination with the ability to make reward decisions on the origin, destination, and point of sale level, creates a completely new and value-based way for rewarding customer loyalty in the most appropriate way. In the same way, upgrade queues can leverage segmentation advice in rules processing so that those customers with the greatest value to the organisation or other relevant segmentation attribute are given preference in upgrades instead of simply basing these decisions on tier level or first in/first out logic. In summary, revenue management applications that enable non-revenue-based customer-centric strategies offer airline organisation a significantly increased targeting capability making customer recognition programmes more clearly focused on delivering benefits to the most profitable or highest value customers.
CONCLUSION
As O&D-based airline revenue management systems evolve, the accompanying real-time evaluation systems will also evolve. Having started from the basic function of bid price evaluation, and growing into an advanced business rules engine, real-time evaluation systems are at the point where new customer-centric strategies are increasingly possible. Those organisations which are able to incorporate both distribution-centric and customer-centric strategies effectively will derive the greatest benefit. The greatest challenge, however, is not technology based but it is organisational. Although robust tools are now able to implement creative and revenue positive customer-centric strategies, the increased cooperation across revenue management, distribution planning, sales, marketing, commercial and loyalty programmes is crucial to successful implementation of revenue management-based customer-centric strategies.
References
- Michael, V. M., Roegner, E. V. and Zawada, C. C. (2004) The Price Advantage, John Wiley & Sons, Inc., Hoboken, NJ.
- Trottman, M. (2005) 'Airlines say deal will reduce ticketing costs', The Wall Street Journal (online edition). http://online.wsj.com/article/SB111456028178817759.html.
