Airline academics estimate that O&D revenue management can further improve an airline’s revenue performance by another 2% above leg-based revenue management. But many airlines still utilize “leg-based” revenue management systems and apply crude tools to differentiate the value of customers “by O&D.”
Certainly, on a short-haul feeder leg, say New Orleans to Dallas, American Airlines, with its mega-hub at Dallas/Fort Worth, would prefer to accommodate the $2,000 passenger connecting to Asia than the $179 passenger going onto Des Moines. Likewise, they might prefer the $179 Des Moines passenger rather than the $79 local passenger.
The conventional way to handle this hierarchy in a leg-based revenue management system is to map fares into fare classes by O&D. So, the Asia connections get mapped into the highest fare classes on the New Orleans-to-Dallas leg, the Des Moines connections into intermediate fare classes and the local fares into the lower fare classes. This results in saving seats for the longer haul connections over local passengers.
But this doesn’t always work very well. In fact, if the Asia connect fares are mapped higher, then the airline could neglect the value of two locals. In this example, the maximum revenue result could be $79 for the New Orleans to Atlanta local plus $2,000 for an Atlanta to Asia passenger. Often, the sum of two locals is higher total revenue than the connect passenger.
One airline I worked with enjoyed tremendous demand on its Bangkok flights – literally full every day. Filling up certain feeder flights with BKK passengers would block out two locals (higher total revenue) and keep the feeder flight from accommodating other high value connect passengers to other parts of the network. In this example, the airline might prefer to map BKK passengers into fare classes lower than Shanghai passengers – despite potentially lower absolute fares for SHA. It is quite complicated to develop the optimal hierarchy across the network.
O&D revenue management optimizes allocations based on forecast demand by leg – accommodating two locals when that is forecast to drive the highest network revenue and accommodating the SHA connection rather than the BKK connection when doing so is forecast to drive higher network revenue. Since demand by O&D will vary both seasonally and even over different days-of-week, any crude valuation scheme falls apart quickly – a dynamic ordering based on each individual flight’s demand forecast is logically more optimal.
Another benefit of O&D systems applies to carriers with multiple hubs. O&D revenue management can determine over which hub a particular O&D should be accommodated. Thus, the BKK connect passenger determined to be of lower value into the main hub, could be accommodated over an alternative and lower demand hub.
For any airline that operates hubs, with significant connecting passengers of varying length-of-haul and fare values, O&D revenue management has been proven to improve total system revenue…but only if the airline builds the sophistication and skills necessary to exploit the O&D RM system into its people and processes.
Managing O&D RM is a much more complex management task than managing a leg-based RM system. When RM systems are criticized for operating like a black box – non-transparent or unintuitive recommendations – O&D Systems further compound that effect. If an airline has a process for hands-on management of its leg-based RM system today – with skilled analysts regularly reviewing allocations and performance, continually adjusting forecast demand and allocations to capture more revenue – then it needs to create an O&D process and organization that features the same transparency and flexibility. Unfortunately, this is not the norm. I have worked with a number of airlines that have implemented O&D systems – recommended due to their heavy dependence on connect passengers over their hubs -- but for which O&D is not optimally managed, leading to worse performance than even under the previous leg-based system.
As is indicated on the chart, some airlines are building new analytic capabilities even without O&D revenue management. Increasingly, big data/analytics have applications beyond traditional revenue management, and which are critical for all airlines.
An O&D revenue management system is appropriate when the airline has certain network complexity – one or more hubs, many connections. However, all airlines that choose O&D RM must also have the internal analytical capability - the people, the organization, and the processes - required to exploit these highly sophisticated systems.