By Matthias Viehmann, Travel in Motion AG
The airline industry is undergoing one of its most significant transformations in decades. The shift toward Modern Airline Retailing (MAR) is no longer a distant vision—it’s happening now. For years, the transition to Offers and Orders has been discussed, but the pace has accelerated. Airlines have started the transition from legacy Passenger Service Systems (PSS) to modern Offer-Order Management Systems (OOMS) to enable dynamic, personalised retailing.
The transition has profound implications for Revenue Management (RM). Retailing enables dynamic, context-specific offers and a more granular trip-specific customer segmentation. To support this, RM must evolve from managing static fare classes to dynamic offer optimisation, including real-time offer curation and contextualised pricing.
Traditional RM: From enabler to obstacle
For decades, Revenue Management (RM) has been the backbone of airline profitability— managing inventories to balance supply and demand while leveraging customers’ willingness-to-pay. This was achieved through forecasting class-based demand and optimising booking class availability. Historically, these mechanisms worked well to segment demand and differentiate products and prices, but today they limit airlines from creating customer-centric, journey-specific offers. Other industries have long surpassed airlines in modern e-commerce capabilities.
Retailing-friendly RM: What will it take?
Separate product from price
As booking classes and filed fares will become obsolete, airlines gain the freedom to dynamically bundle flights and ancillary services in context-specific offers, and to price them without being constrained to specific price points. Modern product management must evolve to manage components and bundling logic, rather than relying on pre-defined static bundles.
Without classes and filed fares, prices must be optimised for dynamic bundles and à-la-carte ancillaries. Forecasting and optimisation models need to handle contextualisation and dynamic bundle construction. Real-time pricing modules are essential—offline pre-calculation will no longer suffice in such a dynamic world.
Balance “the new” supply and demand for price optimisation
Despite changes to product and price, RM still needs to balance supply and demand and optimise the passenger mix. The optimised price must reflect capacity constraints, whether for seats or limited supplies of certain ancillary services.
Solutions will still need to determine the opportunity cost of the next unit, or, as revenue managers call it, the bid price. The logic extends from seats to capacity-constrained ancillaries. In the future, opportunity cost needs to additionally reflect post-booking ancillary sales of potential future customers and marginal cost of items offered.
Price optimisation will split into two components:
- a capacity-centric view to determine opportunity cost for seats and capacity-constrained ancillary services, and
- a customer-centric view to build and price contextualized dynamic offers.
While the former averages across all demand segments for a resource and can run offline, the latter requires evaluating customer, request, and market context in real-time. As offer creation moves from distribution partners to the airlines, processing requests in real time with high shopping volumes, e.g., from meta-searchers, becomes a challenge that must be addressed for any new-generation price optimization to scale.
Up-level demand forecasting models
Forecasting demand at increasingly granular levels has always been challenging, also for analysts to comprehending, validating, and influencing system forecasts. As demand segmentation becomes more detailed and context-driven in the new world, this will become even more pronounced. Splitting price optimisation into a capacity and customer perspective might allow forecasters and analysts to work on natural aggregates, i.e., leg-level and market-level demand.
Estimating market demand needs to evolve from discrete demand across classes to segment-specific class-less willingness-to-pay (WTP) distributions. This eliminates the restriction of 26 letters of the alphabet and allows unlimited price points.
Adopt a new view of willingness-to-pay and elasticity
While airlines have extensive experience in pricing airfare bundles, dynamic pricing for ancillaries is particularly challenging due to limited historical variability. Techniques such as reinforcement learning can help generate the necessary data while simultaneously exploiting customers’ willingness-to-pay. The potential revenue and profit uplifts justify the effort. Several airlines have reported double-digit uplifts when dynamically pricing ancillaries.
Pricing dynamic bundles is still being researched in the scientific and industry communities. Historically, estimating cross-price elasticities from booking data was virtually unfeasible. Dynamically adding components to bundles and optimising contextualised prices requires addressing cross-elasticities. Additionally, bundle and ancillary prices need to be consistent to avoid confusing customers. For example, a superior bundle including an additional ancillary must not be more expensive than an inferior bundle with the ancillary offered à-la-carte. Improved solutions will capitalise on advances in machine learning, combined with a stronger data foundation moving from traditional bookings to Orders consolidating all purchase information.
Additional data streams, such as competitive insights or shopping sessions, can now be integrated to optimise offers. New data sources will emerge in the future, and solutions must be flexible to incorporate them with limited effort. For example, ski resorts in Switzerland and Austria already today factor in weather forecasts to dynamically price ski passes.
Prepare for organisational changes
The change beyond technology is not to be underestimated. Revenue managers have long thought in terms of controlling the availability of static fares supplied by pricing managers. Pricing has been structured around fares and fare ladders for decades. Ancillary services are usually handled separately from flights. Moving to Modern Airline Retailing will require rethinking organisations and redesigning established processes to break down siloed decision-making into a holistic market/customer perspective and a capacity/flight perspective. This split also helps re-define clear-cut responsibilities in the organisation.
The way forward
While the potential is enormous—with the promise to drive profits and customer loyalty—the transformation process is, without a doubt, complex and challenging. Luckily, vendors and front-running airlines have already developed promising solutions. Airline leaders must start the transition and follow a stepwise approach as systems become available and more sophisticated.
Progress is already happening, at an accelerated pace year over year. For example, continuous pricing is already practiced by many airlines; solutions are available from various vendors. Even with legacy PSS and RM systems, simple interpolation between filed fares is a starting point. Once solutions are capable of determining an optimal price point, discounting filed fares based on context is the next step.
Sequentially, class-less forecasting and optimisation models will replace legacy RM and pave the road to offers in the absence of filed fares. In parallel, modules for ancillary services can be deployed to gain experience and benefit from uplift potential.
Airlines starting the transition on the IT side need to address organizational changes in parallel and can gain valuable experience with newly designed processes.
During the transition, one more challenge will persist for quite some time: airline and distribution partners might not move as fast and still require legacy processes requiring translation layers connecting the old and new world.
In summary, the future of Revenue Management will be a fascinating, complex, and (sometimes) messy process as our industry continues its journey toward MAR.
Join us at World Aviation Festival 2026 to discuss the ongoing transition to Offers and Orders.
For more like this, see:
