As O&D yield management techniques take systems to unprecedented levels of capability, the real challenge for airlines will be their proper integration and use. Jackie Gallacher reports.

Like many technologies, yield management has taken time to evolve from the early systems of the 1980s to reach its current level of sophistication. Now the technology is on the verge of a leap forward that will see its improved integration into the airline sales process as it adapts to the requirements of growing international networks.

The latest generation of yield management systems is bringing into play Origin & Destination (O&D) yield management, a technique that involves taking into consideration a passenger's full itinerary when selecting passengers for flights with limited capacity and high demand. As with earlier yield management systems the aim is to optimise revenues, but these techniques go beyond anything that was possible before. 'This is a significant jump and bigger than anything since the beginning of yield management,' says Barry C Smith, senior vice president at AMR Corp's Sabre Decision Technology.

The payback will fall well short of the first systems - when first introduced a completely new technology brings much larger gains - but is nonetheless conservatively estimated at a 1 to 2 per cent annual revenue increase, in other words millions of dollars a year in extra profits for a sizeable airline. Some system vendors, including Bob Salter, vice president and managing director of Pros Strategic Solutions, cite the potential revenue increase at between 2 and 5 per cent in situations of extremely high demand and very limited capacity.

The O&D systems themselves typically cost between $100,000 and $10 million, depending on the needs of the airline. 'I know of no investment that most airlines could make that would have a higher return on investment or revenue payback,' says Donald Garvett, a vice president of consultants SH&E.

However while everyone is agreed that there are substantial spoils to be had, the picture of which carriers are currently in the enviable position of operating an O&D system is blurred by controversy. Most of the uncertainty relates to disagreements over the different versions of O&D yield management being developed by vendors and airlines. The competition among vendors intent on pushing their own systems is also very intense.

Claims to be marketing 'the only true' O&D yield management system abound, as do claims and counterclaims regarding carrier contracts. At the same time carriers' reluctance to reveal too much and confidentiality agreements with vendors make some information hard to come by.

However it appears that only a handful of carriers, including SAS, American Airlines and United Airlines, currently operate new generation O&D yield management controls. But many more carriers are close to acquiring or developing systems or are operating differing degrees of O&D control. These include the other US majors, Cathay Pacific and Lufthansa. British Airways says its current inhouse system embodies O&D yield management 'to some extent' while many other carriers around the world are taking a close look at whether they should acquire such systems.

The latest O&D yield management techniques encompass 'bid pricing,' a technique which determines the opportunity cost of adding a passenger to a flight. This represents the displacement cost of taking a lower yield passenger instead of a higher yield passenger. The bid price roughly equates to the minimum net revenue required on a flight leg and only bookings with a total net value above the bid price are normally accepted.

The bid price is applied within the context of the passenger's full itinerary. This works in two ways. First, a lower yield passenger connecting from a short-haul flight onto a long-haul flight may be given priority on the short leg over a premium class passenger, due to the added value of the long-haul flight. Second, if the bid price is comparably higher on a long-haul flight from Copenhagen to Seattle than on Copenhagen-New York, the same discounted fare may not be available on a feeder flight from Oslo to Copenhagen for the Seattle-bound passenger as for the New York passenger.

Other considerations may also be factored into the system to determine the bid price. Group pricing or group acceptance will calculate the opportunity cost with regard to filling 30 potentially empty seats with a group booking and the possible displacement of passengers who are willing to pay more. Another related technique known as net revenue control will calculate the expense associated with a passenger - such as free limousine service or travel agency commissions - in relation to sources of additional revenue. The information is filed by the airline's revenue accounting system to the yield management system. However net revenue calculations have less impact in the airline industry than in the hotel industry, where additional revenues can be taken into account from the casino, dry cleaning or room service.

The itinerary control permitted by the latest O&D yield management systems is a significant step up from the current 'third' generation systems. These primarily focus on managing discount allocation and overbookings and operate solely on a single flight leg or segment level. Garvett credits most medium to large US, European and Asian carriers with operating 'third' generation systems and moving towards the new O&D systems. However he adds that many smaller carriers need to take a better low technology approach or acquire improved systems (see box).

Peter Belobaba, a professor in the flight transportation laboratory at the Massachusetts Institute of Technology, believes the problems extend beyond smaller carriers, some of which operate good systems extremely well. 'There are a lot of airlines which do not practice good yield management. The vast majority are not even tapping into the opportunities of even the basic yield management of overbooking and discount allocation.' He adds that these airlines do not exclude some in Europe and Asia.

Belobaba is not alone in underlining the dangers of carriers jumping to O&D systems when they have yet to manage the basics of yield management. 'You must first make sure that you are blocking and tackling right, and after that additional value can be added with O&D yield management,' says Garvett. Salter at Pros concurs: 'The people who think this is the new holy grail and haven't done the other part are going to have some real problems.'

The dangers of technology running away with yield management analysts were clearly illustrated back in 1988 when American Airlines' system briefly malfunctioned, costing the carrier $50 million. However, one airline revenue manager commented: 'At least they knew they had a problem.'

In contrast SDT believes its continuous nesting O&D yield management system could enable carriers that are doing little or nothing in yield management to leapfrog more advanced carriers. 'This technology will simplify the transition from doing nothing to doing something significant,' says SDT's Smith. SDT believes its continuous nesting system is unique, though SAS says it has a similar system.

SDT's system offers carriers a specially built smaller computer - the availability processor. 'This allows an airline to move from a leg base up to an O&D level without making major changes to the guts of the reservation system,' says SDT's Smith. SDT has a patent pending and claims that other O&D systems are simply virtual nesting O&D systems. 'I believe some vendors are going in and changing the guts of the system,' says Smith.

Virtual nesting, introduced by AMR in 1986, uses approximate market class controls by clustering market and fare classes into similarly valued groupings called buckets. This means the availability is not explicitly stored in computer reservations systems. Smith describes virtual nesting as an 'approximation' to O&D control due to constraints imposed by CRSs. The latest O&D technology requires seamless availability from CRSs - the ability for all CRSs to communicate on a real time basis. This is in the process of being achieved and currently exists on a limited basis.

Smith adds that SDT is doing O&D control using virtual nesting today, but is in the process of implementing continuous nesting through its availability processor. Elements of this will become available during 1995 and it should be fully functional by 1996.

Needless to say, other vendors deny SDT's claim to supremacy. Pros claims to be developing two of the few O&D systems in the world for Cathay Pacific and Lufthansa, and says Continental already operates the biggest system. However Salter believes that virtual nesting is relevant to most systems: 'Most of the airlines that will move to O&D will probably use some form of virtual nesting.'

Another vendor comments that Continental - which began with virtual nesting and then moved to some O&D control - does not truly operate O&D bid pricing. He explains that there are different practices which give some O&D yield management control. Fare stratification - sometimes known as 'a poor man's O&D system' - refiles fares to give some O&D control across a flight leg bucket system. Point of sale takes into account currency fluctuations, which can mean up to a 50 per cent change in sales value to the airline.

Scot Hornick, senior manager of the information and technology strategy practice at Andersen Consulting, believes the differences between systems often relate to the optimisation process. 'Some of them are using conventional leg-based optimisations to calculate controls, where the optimisation is not O&D based but the implementation is.'

While carriers face a confusing task to distinguish between systems, arguments over the better system may be superfluous. The essential may be to join the ranks of the carriers that are already gaining the benefits from applying some kind of O&D control. 'A proper O&D system is any effort to go beyond basic yield management and distinguish between different O&D itinerary requirements,' summarises Belobaba.

SAS's decision in Decem ber to sell its O&D yield management system to Unisys was in part a recognition that in a very short time there will be rival systems on the market. 'We decided that we might just as well be in the market to sell,' says Peter Groenlund, SAS's VP space control services. Unisys will market the systems and pay SAS royalties on future sales. SAS will serve as a demonstration site and San Francisco-based Decision Focus Inter national, which jointly developed the system with SAS, will customise future installations.

A second important reason for the sale was a desire to put pressure on CRSs to move more quickly to seamless availability. 'Today we have only the necessary functionality from Amadeus. If there are other operators we can put requirements to other CRSs which have some functionality but not enough,' says Groenlund. Groenlund also believes users can cooperate to get cheaper development and optimise results.

The system, baptised Unisys True O&D, is currently used by SAS for around 10 per cent of its flights. The high running costs mean the system is used only for flights with high load factors which are part of a traffic flow or connecting system. An advance selection system is used to determine which flights should be included. Joern Peter Petersen, SAS's manager for yield management system development, says the portion of the network covered could go up to 15 per cent or more.

SAS has been using the system since mid-1993 and, with its traditionally low load factor of 60-61 per cent, has experienced a revenue increase of 1-2 per cent. Based on the airline's 1993 operating revenues of SKr27.5 billion ($3.7 billion) that represents a boost of up to $75 million a year. For carriers with higher load factors the benefits could be higher, says Petersen.

In the immediate future O&D yield management systems are expected to be acquired mainly by larger carriers with substantial hub operations. However, later the technology could apply to all but small point to point or commuter carriers. 'Any airline that operates more than a single isolated flight leg has opportunities for better management of traffic,' says Belobaba.

O&D yield management could also be applied within codeshare alliances. 'This is prompting a lot of the research,' says one vendor. At least one international carrier with a multiple codeshare arrangement is examining the applications to its partner. Most alliance partners cannot share inventories without falling foul of antitrust regulations, but they may be able to communicate their inventories. 'There is scope provided you keep within the regulations,' says a BA spokesman.

As they move to the next level of yield management, both large and smaller carriers need to take extra care that their systems interface sufficiently with the sales, revenue accounting and pricing departments to avoid the input of distorted data. 'Just as possessing aircraft does not make you an effective airline, buying in a good system does not guarantee the airline is practising effective revenue management,' says Garvett. The new systems should make coordination easier than in the past: 'We are putting revenue management right back in the financial stream,' says SDT's Smith.

As carriers take up their strategic positions in this latest revolution in yield management, the poorly armed airlines will need to avoid a widening of the competitive gap with their more powerful counterparts. Meanwhile consumers will be weighing the pros and cons of whether they stand to win or lose from the change in fare structures.



Most of the world's airlines are small and medium-sized - over half have under US$300 million in annual revenues - and they are often disadvantaged with respect to the greater resources of larger airlines. Yield or revenue management is one area that can make a big difference: when properly used it has a direct impact on profitability and is a powerful competitive weapon. But can small and medium-sized airlines employ the revenue management techniques used so successfully by the mega carriers?

Given the industry's very narrow profit margins, no carrier can afford to ignore this competitive tool or its potential revenue gains. Nonetheless, the organised revenue management activities of smaller airlines are, in many cases, minimal or non-existent, and in some instances their potential tangible value has not been made apparent to management. Furthermore, there are various real and apparent barriers to exploiting revenue management at these airlines.

The two most obvious barriers are costs and staff experience or expertise. A host of other barriers includes market practices - including few fare conditions and limited capacity control - as well as undisciplined travel agents, inter-departmental rivalries, and competing demands on resources. In addition, many static and seemingly viable approaches are well entrenched at such carriers.

Breaking through these barriers starts with looking at the fundamental principles and objectives of revenue management. It is important to recognise that overbooking, discount seat allocation, itinerary (O&D) control, group management and other areas are activities and processes, not just systems. Thus much can be done even without investment in high-technology systems.

The most important first step is for top management to give revenue management functions a high priority. Achieving gains will then depend on establishing departmental and individual accountability with measurable goals and feedback. Otherwise, airlines may find well motivated sales staff seeking passenger volume rather than maximum revenue per available seat km, or flight controllers not overbooking enough to avoid denied boarding complaints from airport staff.

A wide range of decision-support tools can be built on this infrastructure to provide significant benefits and a framework for further development. The key elements involve process reviews, tracking and exception reporting systems, and interdepartmental linkages. This will involve evaluating and continuously improving areas such as how flights are assigned to controllers, what decisional logic is employed, how and whether flights are reconfirmed, and what areas sales staff should focus on. This also will involve the development of simple decision-support tools such as which markets experience excessive spoilage, whether flights are booking closer-in, and how well tour operators use group space.

While these low-technology approaches will capture a lot of potential revenue, all but the smallest airlines can usually benefit from state-of-the-art yield management systems. While their cost of US$1-10 million may seem daunting, airlines can acquire high quality entry-level systems for as little as $150,000, and there is probably no other expenditure that can provide a higher return on investment than revenue management technology. A $500,000 investment which provides a net revenue gain of 1 per cent for a $50 million airline will have a payback period of just one year.


Source: Airline Business