By Dr. Ricardo Pilon
Introduction
Deregulation and increased competition provided the impetus for passenger airlines to adopt a more sophisticated approach towards marketing ever since the 1980s. The closing stages of regulated pricing brought about an initial era of fierce pricing competition; resulting in the apprehension that profitability could only be instigated through erudite market segmentation aimed at identifying those market segments that would have a higher willingness to pay while stimulating demand at discounted fare levels in other segments (Cross, 1997). Born was the new discipline of revenue management (RM), embraced by pioneers such as American Airlines, who applied differential pricing and yield-class based inventory control as part of the first generation of revenue management systems (RMS).
The success of revenue management has been documented widely and is often claimed to generate incremental revenues of between 3-8 per cent (Belobaba, 1987; Cross, 1997). Following airlines, other hospitality businesses such as hotels, car rental companies and cruise lines soon followed suit (Donaghy et al., 1997). However, the air cargo industry has generally been progressing unhurriedly with respect to automation and convoluted business technology. Despite the actuality that the cargo industry enjoys all the characteristics of industries that qualify for revenue management application, many obstacles and hurdles are often quoted as impeding its implementation (Jonker, 2006).
This Whitepaper proposes a new perspective of cargo revenue management using supporting tools that will allow passenger combination and all-cargo carriers to re-instate profitability through enhanced analytics and sophisticated pricing and capacity optimization.
The first part of this Whitepaper outlines the often-cited obstacles to cargo revenue management. The following section describes the requirements for cargo RM, followed by a proposed innovative approach to innovative revenue optimization business technology in the air cargo industry. A conclusion follows to recapitulate the feasibility and benefits in embracing revenue management for a sustainable cargo business.
Background
A number of general business realities as well as daily operational limitations hamper cargo revenue management. The air cargo activities of a combination carrier are multi-faceted and sternly entwined in a multitude of other elements of the carrier’s organization and operations.
The predominant focus many combination carriers put on their passenger business is oftentimes the root cause of a hindrance to achieving the full potential of cargo market and revenue opportunities. But the cargo value contribution could potentially exceed the average 15% of carrier revenues it generates today (ACW, 2005), notably in bottom-line results. Generally, divergent demand patterns between air cargo and passenger demand could help stabilize system-wide capacity utilization (be it main deck for combination aircraft or belly-hold capacity). So what tends to coerce this conflict and what is the foundation of air cargo still being considered a by-product today?
Some fundamental challenges to cargo revenue management are evident when comparing the attributes of capacity, demand, and profit optimization between the passenger and the cargo business.
· Demand: Unlike the passenger business where demand tends to be relatively stable and regular when corrected for seasonality, air cargo demand is highly unstable and irregular. Also, unlike the one-dimensional seat, the demand for cargo space is multi-dimensional with length, height, width, that is volume/weight and Unit Load Devices (ULD) position requirements. In addition, contrary to the passenger itinerary-based travel, air cargo is indifferent as to the itinerary traveled or routing as long as service-level agreements are respected. A key factor that distinguishes air cargo demand as well is the fact that by and large the market is accounted for by relatively few customers (intermediaries). This leads to shipment consolidation and has significant impact on contract negotiations and dependency on the part of air cargo operators. Finally, passengers – when they show up – tend to be flown as booked in terms of capacity requirements; whereas under- and over-tendering with regards to volume and weight tends to occur regularly, requiring precise overbooking.
· Capacity: The cargo capacity characteristics are comparatively far more complex and multifaceted than the single-dimensional seat requirements of passengers. The multi-dimensional cargo capacity characteristics include length, height, width, ULD positions whereas the overall capacity depends on many other factors that impact on its available volume and weight. These include the passenger, baggage, mail, fuel, crew and catering loads in addition to unforeseen weather conditions, potential ad-hoc passenger connection requirements, and so forth. The highly variable nature of cargo capacity is further underscored by different aircraft types, flown missions, the availability of all-cargo and charter aircraft, or belly-hold space in combination carriers, or procurable trucked air cargo capacity.
Returns on investment in systems are lower in absolute terms for cargo. While some carriers have taken the cargo business seriously, revenues from air cargo typically represent only a fraction of the size of the overall business, so a revenue uplift of 3-8 percent means less. For the cargo business, installing a revenue management system generally requires the re-engineering of a myriad of complex (manual) processes and inter-dependent systems and thus costs much more, in absolute terms, and returns much less in incremental revenue (Kliewer et al, 2002). Also, air cargo managers have typically gained remarkable experience and make intuitive decisions that have for the most part gone unformulated and undocumented. In sum, frequently quoted as being “too complex”, some argue that the implementation of revenue management systems for air cargo optimization is simply not feasible.
Cargo Revenue Management Requirements
The previous statement is flawed as air cargo enjoys all the characteristics of industries that qualify for revenue management application. Cargo capacity is a perishable commodity as it cannot be stored. This notion of time inherently represents value in a market where many segments depend on service level agreements as part of a larger supply-chain propelled by global trade. Also, owing to relatively low variable costs (fuel requirements, handling costs), yield enhancements generated from improved pricing and capacity control almost solely ends up in bottom-line profits.
However, the success of RM in air cargo centers on systemic process re-engineering in marketing and sales, rate establishment and capacity control. Only by achieving a clear understanding of the demand for air cargo products or services at a micro-market level can a carrier maximize its revenue by optimizing price and availability of the product. It thus represents an area in which marketing and technology, supported by mathematics is closely tied together. The need for systems and technology arises as the amount of information that is available and needs to be processed increases over time. The amount of data would be impossible to process manually.
RM is increasingly focusing on dynamic pricing, where product prices are established in accordance with demand from multiple customer segments to maximize revenue or profit. As a result, prices are adjusted dynamically, depending on the inventory availability and time left in the selling season. This compels companies to be able to change prices quickly to reflect new business conditions. Systems are required to use real-time information and formulas to calculate prices. Prices need to be set across products, markets, and channels in order to meet demand. The above works well for spot market capacity (free-sell), and requires air cargo carriers to review the revenue effectiveness of the common practice of allocating space in advance under contract (so called “allotments”). However, clear opportunities exist for decision-support tools allowing optimal rate determination for allotments as well.
A full-scale cargo revenue management system would allow companies to avoid missing out on revenues and profits, by preventing:
· Discounting rates when unnecessary
· Neglecting up-selling opportunities to customers who want to buy value-added services
· Wasting money on irrelevant promotions to the wrong customers
· Misallocating inventory to unprofitable products, segments, or channels
· Failing to create product packages that enhance value (differentiated products)
· Maintaining excessive inventory levels blocked for sale (reserved space, allotments)
· Unprofitable behavior driven around long-term contracts
· Neglecting high-value customers.
In order to be successful, cargo RM needs to apply disciplined tactics that predict consumer behavior at the micro-market level, optimizing product availability and price to maximize revenue growth. The technique allows for the setting of the price of goods and services based on factors such as current product demand, anticipated demand, real-time inventory, and the customer’s willingness to pay.
As a result, the three main areas of cargo revenue management around which business technology should be built are:
1. Clear market segmentation around differentiated products
2. Forecasting demand for different products, points-of-sales, and (reaction to) price levels
3. Adjusting capacity availability or price to accept the highest-yielding mix of products across the network at any given point in time.
Business Technology for Air Cargo Revenue Optimization
So as to develop an effective solution for cargo revenue management requirements, it is of crucial importance to relate the “what” to the “how to”. The “what” requires a clear definition and often depends on data availability of the reservation systems that are used as well as current business processes. Innovative RMS’ also likely require business process reengineering, and may thus impact profoundly on the organization. Oftentimes, the amount of data that is required does not represent valid information necessary to make (automated) decisions. One must also take into account that customers can adopt new and ever-changing behaviors depending on the market or business environment.
The overall systems objective is to control reservations and sales through traffic-flow or network control (local demand versus overall network optimization supported by demand forecasting and rate optimization, including discount allocation and availability restrictions (protect inventory for higher-yielding demand), as well as overbooking (allocating more inventory than physically available in order to account for cancellations and no-shows).
Designing an Innovative Cargo Revenue Management System
From a user standpoint, the following modules could be envisioned as part of an all-encompassing cargo revenue management solution:
· Historical Performance Review: Allows revenue analysts to conduct a full analysis of past economic performance at a very detailed level. This would include past network and leg capacity utilization (weight, volume), booking cycles and average yield information per origin-destination (OD), product category, commodity, customer, etc. Rate-demand curves further allow users to obtain valuable inputs into rate elasticity that aids pricing analysts to optimize rate lines.
· Capacity Forecasting: Cargo capacity forecasting is based on passenger load forecasts, baggage and mail load estimations, fuel requirements, as well as crew, catering and other configurable loads. Passenger final load forecasts could be obtained from the passenger revenue management system, while baggage and mail loads could be estimated using historical data obtained from departure control systems for each flight number and calendar date/day. Fuel requirements and updated cargo capacity in terms of weight can be calculated dynamically as per the cargo demand forecaster, since fuel requirements depend on total take-off weight and sector length. Evidently, capacity forecasting is sensitive to aircraft type and mission flown and could be adjusted for contingency factors such as the propensity for ad-hoc passenger requirements (probabilistic forecasting of misconnections per season, etc.).
· Demand Forecasting: Using a variety of forecasting methods, demand for air cargo space would be forecast from the ground up, i.e. per category, commodity, product, customer/point of sale, SCC, all based on historical data. A future feature could be to obtain rate-elasticity based demand forecasts. Weighting factors could be applied to the data of multiple previous years, while demand forecast updates during the booking cycle of a flight could also be configured in terms of the weight of historical versus current booking data. Overbooking levels could be set based on user-configurable risk-averse levels related to historically-observed no-show rates as well as cancellations, under- and over-tendering behaviour for each station. In addition, the probability of shipment acceptance without booking is also forecast and included in the final demand forecast.
· Tariff Management: Users would be able to review historical rate performance using price-demand curves and use what-if analyses to simulate the potential impact of new rate lines on overall revenue performance. Different than today’s industry practice of using commodity rates with weight brackets for date ranges, cargo carriers would be able to implement calendar day/date-based pricing (at the individual flight level). This represents a significant revenue enhancement and could only be performed using automated business technology. Using information obtained from the optimizer, which sets hurdle or dynamic bid prices, the system would also determine whether rates established by users satisfy the hurdle rate criterion.
· Schedule Allotment Creation: Schedule allotments are created based on the historical economic performance as well as yield-based demand forecast of each customer/station for each category and commodity. Allotments are assigned subject to a number of conditions such as the regularity of supply from the station for the category and commodity, the observed show-up rate as well as under- and over-tendering, minimum yield requirements, as well as overall system contribution (network optimization). For differentiated products (time-sensitive, live animals, etc.), a yield-class based approach is suggested and as such, allotments are allocated as such.
· Revenue Planning: This function would allow users to perform a full revenue plan based on an optimized mix of allotments vs. spot capacity for each category and commodity across the cargo network. Inherently, sales targets could be derived from the revenue plan based on the minimum yield levels that have been established by the demand forecaster and optimizer, although they could be adjusted manually. Prior to being released as the active version, revenue plans can be prepared and adjusted for macro-economic factors such as exchange rate fluctuations, commodity price indices, competitive reaction, as well as anticipated GDP and market growth.
· Flight Capacity Optimization: This module would be part of revenue and booking analysts’ day-to-day work list. It allows the analysis, acceptance or rejection of all shipment requests that have been received in an automated as well as manual fashion through spot rate requests that require user intervention. The suggested optimization approach is based on the concept of each request’s “shipment value”, which is a weighted-average value based on shipment yield, revenue, and customer importance. Furthermore, automated acceptance would be based on a validation of the shipment value against the yield-performance profile of each flight, while respecting the service level agreements. The over-arching objective of flight capacity optimization is to ensure that low-value shipments are assigned to lower-demand flights and that capacity on high-demand flights is protected for higher-value shipments. Requests for which the spot rate does not satisfy the minimum yield requirement are queued for user follow-up with supported alert functions (priority list, based on departure date, customer, or any other configurable parameter).
· Dynamic Allotment Management: Booking behavior of allocated allotments that concern non-contracted space could be monitored for potential release and redistribution to other stations based on updated forecasts and re-optimized network space allocation. This would occur on the basis of initial allotment versus current and projected booking levels relative to forecast threshold values, allowing the estimation of unutilized capacity at flight departure.
· Queue Management & Alerts: The queue management function would permit the handling of shipments on the standby list, disruption list, but also allow users to identify shipments that can be considered for pre-carriage in order to maximize capacity that otherwise is anticipated to go unutilized. The logic used by the system to handle and assign shipments to flights would be similar to the flight capacity optimization approach described earlier. That is to say, the system would identify flights for which the shipment value matches the economic value of the flight while still respecting (contracted or requested) service level agreements. Furthermore, alerts can be generated by the system so as to signal users that the flight (booking) performance is such that it merits attention (“warning”) or immediate action (“critical”). Alerts would go hand in hand with rule-based controls that users could set up with regards to capacity control and dynamic pricing.
Conclusion
While the implementation of automated revenue management business technology is still relatively scarce in the air cargo industry, a noteworthy opportunity exists to introduce a systematic approach to revenue optimization supported by business technology consisting of decision-making as well as automated optimization tools. This Whitepaper has described a vision towards next generation cargo revenue management that goes beyond today’s practices and includes potential future requirements that would be financially rewarding for both combination as well as all-cargo carriers.
The rewards of cargo revenue management systems would include greater transparency of cargo’s contribution to corporate results as well as significantly improved cargo revenue from increased yields, better utilized uplift capacity and enhanced sales performance. Notable additional benefits would include improved customer satisfaction from better services and network over and above an expected cost reduction from streamlined processes.
The current generation of legacy cargo reservation systems and general lack of sophisticated cargo revenue management systems falls short of supporting the above goals. In conclusion, now that the industry’s financial results appear to be strengthening, the timing is ripe to invest in sophisticated cargo revenue management solutions so as to enjoy maximized profits.
Bibliography
ACW (2005), Air Cargo World online, www.aircargoworld, October 2005.
Belobaba, P. (1987), Air Travel Demand and Airline Seat Inventory Management, Ph.D. dissertation, Massachusetts Institute of Technology, Boston, USA.
Cross, R. G. (1997), Revenue Management, Broadway Books, New York, USA.
Donaghy, K., McMahon-Beattie, U. and McDowell, D. (1997), ‘Yield Management Practices’, in Yeoman, I. and Ingold, A. (eds), Yield Management Strategies for the Services Industries, Casell, London, pp. 183-201.
Garvett, D. and Hilton, K. (1999), ‘What drives airline profits? A first look’, in The Handbook of Airline Finance, McGraw-Hill, New York, pp. 181.
Jonker, E. (2006), Internship report, Business Mathematics and Informatics, Vrije Universiteit, Faculteit der Exacte Wetenschappen, Amsterdam, The Netherlands.
Kimes, S.E. (1989), ‘Yield management: a tool for capacity-constrained service firms’, Journal of Operations Management, 8, 4, pp. 348-363.
Kliewer, G., Grothklags, S. and Weber, K. (2002), ‘Improving revenue by system integration and cooperative optimization’, Proceedings of the AGIFORS RYMSG Annual Meeting 2002, 16-19 April, AGIFORS, Berlin, Germany.