|
|
Monday – May 14
|
|
|
15:00
– 18:00
|
Registration
|
|
18:00
– 21:00
|
Conference
Opening Ceremony
Welcome
Reception & Dinner in Jeju KAL Hotel
Sponsored by Korean Air and PROS
|
|
Tuesday – May 15
|
|
|
07:15
– 08:15
|
Breakfast in Jeju KAL Hotel
|
|
08:15
– 08:40
|
Welcome Address
Hak Jin Park - Korean Air
|
|
08:40
– 09:05
|
Conference Overview and Airline
Updates
Gina Morello - AGIFORS RM Study
Group
|
|
09:05
– 09:45
|
Keynote Presentation
What’s Wrong with
RM, and What Can We Do About It?
Scott
Nason – American Airlines
|
|
09:45
– 10:25
|
RM performance measurement
with a dependent demand model
Scott
Chandler and Sunny Ja - American Airlines
|
|
10:25
– 10:40
|
Coffee break
|
|
10:40
– 11:20
|
Comparison
of RM Methods for Reversing Spiral Down
Peter
Belobaba - MIT
International Center for Air Transportation
|
|
11:20
– 12:00
|
Methods for Estimating Sell-Up: Part II
Craig Hopperstad - Hopperstad Consulting, Inc.
|
|
12:00
– 13:00
|
Lunch in Jeju KAL Hotel
Sponsored by ITA
|
|
13:00
– 13:40
|
Keynote Presentation
A
Look Back at Revenue Management Development and Some of the Bumps along
the Way
Barry
Smith – Sabre Holdings
|
|
13:40
– 14:20
|
Revenue
Management under the Multinomial Logit Model: Incorporating
Competitive, Recapture, and Restriction-free Pricing Effects into
Single-leg EMSR Optimization
Richard Ratliff -
Sabre Holdings Inc.
|
|
14:20
– 15:00
|
Conditional
Revenue Optimization (CRO): United Airlines’ Experience and Future
Challenges
Burak Ozdaryal - United
Airlines
|
|
15:00
– 15:15
|
Coffee break
|
|
15:15
– 15:55
|
Cargo Rate Optimization
Ben Vinod - Sabre Holdings Inc.
|
|
15:55
– 16:35
|
Demand
and supply driven complexities of Air Cargo Revenue Management
Bjoern Becker - Lufthansa Cargo
|
|
16:35
– 17:15
|
Air cargo overbooking based on the shipment
information record - feasibility, definitions and avenues for further
research
Bjoern Becker - Lufthansa Cargo
|
|
18:00
– 21:00
|
Evening Networking program
Dinner
Sponsored by Sabre Airline Solutions
|
|
Wednesday – May 16
|
|
|
07:30
– 08:30
|
Breakfast in Jeju KAL Hotel
|
|
08:30
– 09:10
|
Use
of Competitive Data and Simulations
Royce
Kallesen - PROS RM
|
|
09:10
– 09:50
|
The
Continuing Evolution of Revenue Management: Customer Centric Revenue
Management
Ben
Vinod - Sabre Holdings Inc.
|
|
09:50
– 10:30
|
From
RTDP (Real-Time Dynamic Pricing) to RTCM (Real-Time Customer
Management): Using Customer-Centric Segmentation
Garth
Hoff - PROS RM
|
|
10:30
– 10:45
|
Coffee break
|
|
10:45
– 11:25
|
Impact
of the Internet on Airline Fares II
Bill Brunger - Continental Airlines
|
|
11:25
– 12:05
|
A
seasonal index for demand forecasting
Dan Muzich and Jeff McLellan
- USAirways
|
|
12:05
– 13:00
|
Lunch in Jeju KAL Hotel
|
|
13:00
– 13:40
|
SAS'
O&D forecasting system Odyssey
Thomas Fiig - Scandinavian
Airlines System
|
|
13:40
– 14:20
|
A
more detailed integration of pricing into revenue management
forecasting
Natascha Jung - Lufthansa Systems
|
|
14:20
– 15:00
|
Linear
Approximation Model for Network Revenue Management
Moon
Gil Yoon - Korea
Aerospace University
|
|
15:00
– 15:15
|
Coffee break
|
|
15:15
– 15:55
|
Positioning Revenue
Management within the Organizational Structure
Adam Dudar - Air Canada
Cargo
|
|
15:55
– 16:35
|
Future Opportunities &
Directions for Air Cargo Revenue Management
Ricardo V. Pilon - IBS Software Services (P)
Inc.
|
|
16:35
– 17:15
|
Vendor
Presentation
Amadeus' Altéa
suite of airline IT solutions
Damian Hickey - Amadeus
|
|
18.00
– 21:00
|
Evening Networking program
Dinner in Seogwipo KAL Hotel
Sponsored by Amadeus
|
|
Thursday – May 17
|
|
|
07:20
– 08:20
|
Breakfast in Jeju KAL Hotel
|
|
08:20
– 09:00
|
Dynamic Pricing
of Perishable Assets under Competition
Guillermo Gallego – Columbia University
|
|
09:00
– 09:40
|
Forecasting
on Market Sensitivity
Catherine
Cleophas - International Graduate
School of Dynamic
Intelligent Systems
Michael Frank - Deutsche Lufthansa AG
|
|
09:40
– 10:20
|
Agent
Based Modeling of Passenger Behavior
Richard Lonsdale
- Boeing
|
|
10:20
– 10:35
|
Coffee break
|
|
10:35
– 11:15
|
Basics
of Dynamic Programming for Revenue Management
Jean
Michel Chapuis - University of French
Polynesia
|
|
11:15
– 11:30
|
Conference
Awards and Closing Ceremony
Gina Morello - AGIFORS RM Study
Group
|
|
12:00
– 16:00
|
Lunch & Half-Day Networking
program
Jeju
Island Local Tour (Lunch on the bus)
Sponsored by AGIFORS
|
|
|
|
Keynote Speakers
What’s
Wrong with RM, and What Can We Do About It?
Scott Nason, VP of Revenue Management
American Airlines
This presentation will discuss shortcomings in
existing RM systems and the thoughts on what we need to change in the
next 5-10 years.
A Look Back at
Revenue Management Development and Some of the Bumps along the Way
Barry Smith, Chief Scientist
Sabre Holdings
The development of revenue management has been driven
by the evolution of the airline marketplace. Over the past 30
years, changes in airline business practices have forced the revenue
management community to reinvent models and procedures to maintain and increase
their value. We’ll look back on some of the most significant
changes that affected revenue management development, review the current
landscape and look forward to future developments.
Technical Program
RM
Presentations ( Based on submitted abstracts to date* )
Comparison of RM
Methods for Reversing Spiral Down
Peter Belobaba
MIT International Center for Air Transportation
The “spiral-down” effect in fare structures without strong
segmentation restrictions (e.g. minimum stay) can lead to revenue losses
of 5-10% with traditional RM forecasting and optimization models.
The current RM challenge is to find ways to reverse this effect in order
to reclaim all or part of this revenue loss, and make existing RM systems
more effective. This presentation will summarize our PODS research
over the past year, as we compare the performance of previously
documented methods like EMSR Sellup models with more recent advances in
Fare Adjustment methods and Hybrid forecasting of price- vs.
product-oriented demand.
Methods for
Estimating Sell-Up: Part II
Craig Hopperstad
Hopperstad Consulting, Inc.
Sell-up estimation continues to be an important topic
in PODS (Passenger O/D Simulator) research. The environment for
estimation has been expanded to include semi-restricted fare
products. Several new methods will be described and PODS results
provided as to their effectiveness.
RM performance measurement with a dependent
demand model
Scott Chandler and Shau-Shiang Ja
American Airlines
Performance measurement of an airline’s revenue
management system is important to understanding weaknesses in the system
and focusing efforts on where improvements can be made. Some
measures that are very familiar, such as RASM and Market share, can
easily be affected by competitive actions and other conditions outside
the control of the airline’s RM system. One measure that hopes to
remove the impact of these outside influences is the Revenue Opportunity
Model. This model measures the actual revenue performance in
relation to first-come, first-serve (no RM) and a simulated maximum
achievable (perfect RM) to see what % of the revenue opportunity was
captured. This allows for a reasonable assessment of the RM system
and a consistent performance comparison across markets and time periods.
Unfortunately, one of the traditional assumptions in the Revenue
Opportunity Model is that the pax demand for
different fare classes are independent of each other and the controls
that are in place. This paper will focus how the erosion of fare
rules, which has created the need for a revolution in RM optimization and
forecasting models, has also had an effect on
the traditional Revenue Opportunity Model.
Revenue Management
under the Multinomial Logit Model: Incorporating Competitive, Recapture,
and Restriction-free Pricing Effects into Single-leg EMSR Optimization
Guillermo Gallego, Lin Li, and Richard Ratliff
Columbia University and Sabre Holding Inc.
New heuristics for single-leg, EMSR-based airline
optimization problems are described based on multinomial logit
choice demand models including downsell, upsell and competitor
effects. Extensive simulation analysis was performed across a wide
range of market conditions for the two-class case. Revenue performance of
numerous different heuristics is compared against optimal booking
controls. Implications for multi-class and O&D problems
are discussed.
Dynamic Pricing of Perishable Assets
under Competition
Guillermo Gallego
Columbia University
Internet fare transparency and restriction-free
pricing present a challenge to fare based revenue management solutions.
Carriers need systems based on consumer choice models that take into
account competitive prices in real-time. We provide a choice-based model
under competition and formulate it as a stochastic control problem with
prices as decision variables. We provide sufficient conditions for the existence
and uniqueness of open-loop Nash equilibria of
the corresponding differential game. We show that pricing heuristics
suggested by the open-loop Nash equilibria are
asymptotically epsilon-Nash equilibria for the
stochastic game. We then present results indicating how carriers should
react to price changes initiated by competitors and show that it is
rarely optimal to match price changes. Some numerical examples are
provided to illustrate the nature of competitive pricing.
Basics of Dynamic Programming for
Revenue Management
Jean Michel Chapuis
University of French Polynesia
The Revenue Management (RM), namely the pricing and
the inventory control of a perishable product, is usually used to improve
services marketing efficiency. While booking a flight, the manager has to
allocate seats to various fare classes. Then, he has to assess the
consequence of a current decision on the future stream of revenue, i.e.
accept an certain incoming reservation or wait
for a possible higher fare demand, but later. Since its practice becomes
omnipresent this last decade, this paper presents some basics of Dynamic
Programming (DP) through the most common model, the dynamic discrete
allocation of a resource to n fare classes. The properties of the
opportunity cost of using a unit of a given capacity, the key of any RM
optimizations, are studied in details.
The Continuing
Evolution of Revenue Management: Customer Centric Revenue Management
B. Vinod
Sabre Holding Inc.
This presentation discusses the evolving trend of
customer centric revenue management over the competitive landscape. In an effort to retain profitable
customers, airlines are investing in advanced analytics to gain insights
into customer traits and preferences.
Central to the data analytics and data mining initiatives is the
revenue management and inventory control process for recommending the
right offer to a customer and sales channel. Six key customer centric initiatives
are discussed for acquiring new customers and retaining existing
customers and its potential impacts on pricing and revenue management.
Impact of the
Internet on Airline Fares II
Bill Brunger
Continental Airlines
Bill will update his description of his doctoral
research focused on how customers have understood the transition from
traditional to Internet distribution, and how that change has affected
realized airline fares.
Conditional Revenue
Optimization (CRO): United Airlines’ Experience and Future Challenges
Burak Ozdaryal
United Airlines
Major changes have taken place in the airline industry resulting in 97%
of United’s domestic revenue being generated in markets with
non-traditional fare structures. Reductions in fare restrictions have
resulted in revenue dilution due to buy-down. United has been enhancing
its O&D revenue management system (ORION) with conditional revenue
optimization (CRO) working closely with PROS Revenue Management. The
objective of CRO project is to improve revenue performance by forecasting
passengers’ price sensitivity and leveraging buy-up probabilities in a
new optimization subsystem. Some results from our initial tests of the
enhancements will be shared. This talk will also go into the details of
some of the shortcomings of the current product and the challenges
inherent in overcoming them.
A seasonal index for
demand forecasting
Dan Muzich and Jeff McLellan
USAirways
Improving the accuracy of airline demand forecasts
results in a better allocation of inventory and increased revenues. For
airlines that use seasonality as part of their forecasting methodology,
accuracy is greatly affected by the seasonal multipliers that are applied
to weekly historical data. Holidays do not fall in the same week every
year, complicating the calculation of the multipliers. This presentation
develops a methodology that uses trend filtering, outlier detection and
weighting to improve the quality of weekly seasonal multipliers. In
addition, a method is proposed to respond to holiday movement by
explicitly modeling their residual effects. These methods have been
implemented at a major US
carrier and are expected to result in a 9% reduction in the mean absolute
percentage error of the forecast.
Forecasting on
Market Sensitivity
Catherine Cleophas, International Graduate
School of Dynamic
Intelligent Systems
Michael
Frank, Deutsche Lufthansa AG
Assuming independent demand distributions, forecasts
for revenue management have been implemented and applied successfully in
the past. However, with the advent of restriction-free fares, customer
behavior changed and is now more dependent on prices than on
restrictions. Internet portals enable the effortless comparison of prices
at diverse airlines. This presentation gives an overview over new forecast
approaches incorporating the examination of price-elasticity. This
includes methods of regression, discrete choice analysis, and neural
networks. It points out approaches to integrating information about
market competition as well as upsell behavior. Further more, it offers a
concept of comparing and evaluating results from different forecast methods
on a common data basis.
Agent Based Modeling
of Passenger Behavior
Roger A Parker and Richard Lonsdale
Boeing
Agent-based simulation modeling is a method for analyzing complex
phenomenon by simulating the behavior of individual agents in a complex
system. Recent research in Boeing Marketing has applied these methods to
the dynamics of airline passenger booking behavior in the presence of
airline revenue management systems. This presentation will discuss the
nature and scope of the research. The central concepts of agent-based
modeling will be presented, including the definition of agent and agency,
their relationship to complex adaptive systems, computer implementation
issues, and how agent models are validated. An example will be described
showing how agents representing airline passengers can be modeled.
Ongoing research and future applications will also be described.
Use of Competitive
Data and Simulations
Royce Kallesen and Dan Zhang
PROS RM
Using competitive data within revenue management
science represents a change in demand models and forecasting and
optimization algorithms. We present ways in which this information can be
utilized. Simulations in a competitive environment will also be
presented.
From RTDP (Real-Time
Dynamic Pricing) to RTCM (Real-Time Customer Management): Using
Customer-Centric Segmentation
Garth Hoff
PROS RM
Many airline O&D revenue management systems rely
upon a dynamic bid price evaluation that serves the basic function of
making a comparison between the minimum acceptable price per fare class
and the optimized bid price per fare class. Over time these dynamic
bid price evaluation systems have evolved to accomplish an ever
increasing number of roles beyond the basic evaluation. These
expanded features generally include accounting for transaction value
differences by distribution channel, implementing non-published fare
rules restrictions, checking for booking agent policy compliance, and
reporting on the impact of these strategic rules and strategies.
What is not widely used for the purposes of dynamic
bid price evaluation is customer-centric data including “customer
attributes” and “product attributes” in developing a complex
segmentation. Airlines are increasingly using customer-centric
strategies to offer their customers a differentiated pricing and product
offering. With the rise of supplier.com sites and direct connect
bookings, airlines are increasing their control of product distribution
and customer interaction, making it possible to encourage customer self
identification for the purposes of gaining access to these preferred or
targeted pricing and product offerings optimized to the customers
specific profile. Those organizations that are able both adapt
organizationally to an environment where revenue management, pricing,
sales and marketing are in close communication and implement a system for
managing complex customer-centric segmentation decisions will find revenue
benefits through increased “win rates” and positive customer good will.
This presentation will articulate the value
proposition and examine several high-value business cases that can be
addressed with an RTCM module. An overview of concepts will include
customer attributes, product attributes, attribute buckets, and
segmentation. Additional time will also be spent examining the
customer management process flow as part of an expanded real-time
availability module implementation.
SAS' O&D forecasting
system Odyssey
Thomas Fiig
Scandinavian Airlines System
The Odyssey forecasting project has been developed to provide improved
O&D forecasts to the O&D Optimization System and to provide
analysts with information and interventions at traffic flow level. The
O&D forecasts are created for the SAS Group including regional
partners. The central element in the system is a traffic flow database,
which contains information such as booked and fares by traffic flow. The
system allows for automatic handling of schedule changes; flight number
changes; special periods such as Christmas and Easter. A graphical user
interface has been developed that allows the users to make interventions
to adjust historical bookings and forecasts; create new destinations;
perform airport changes; and booking class changes. The Odyssey system
includes estimation of the price elasticity by traffic flow for
unrestricted fare products.
A more detailed
integration of pricing into revenue management forecasting
Natascha
Jung and Silvia Riedel
Lufthansa Systems
Currently the forecasting within the revenue
management process in the airline business considers only the booking
classes of the own carrier without any information about the underlying
price product. Consequently it is neither possible to depict purely the
dependency between the own booking classes in terms of sell-up and
buy-down effects nor the dependency between competitive booking classes
in terms of migration effects. By a detailed analysis of the pricing data
like ATPCo-data it is possible to identify the underlying price product
of own and competitive booking classes. We propose an unconstraining
procedure that uses booking data in connection with competitor
information like MIDT-data and considers the real dependency between own
and competitive booking classes. It produces also a reliable base for the
estimation and computation of price elasticity.
Linear Approximation
Model for Network Revenue Management
Moon Gil Yoon and Hwi
Young Lee
Korea
Aerospace University
In this paper, we propose an approximation model for
solving airline seat inventory control problem in network environments.
Using Linear Approximation technique, we will transform our problem into
a concave piecewise LP model. Based on the optimal solution of ours, we
suggest how to implement it for airline inventory control policies such
as booking limits, bid-price controls and virtual nesting controls.
Cargo
Presentations ( Based on submitted abstracts to date* )
Cargo Rate Optimization
Hari Subramanian and
Ben Vinod
Sabre Holding Inc.
The
cargo industry is faced with declining yields over the past three
decades. This talk focuses on the value proposition of optimal
pricing. A pricing decision support framework to pro-actively
determine the optimal cargo rate structure will be presented.
Demand and supply driven complexities of Air
Cargo Revenue Management
Bjoern Becker, Lufthansa Cargo and European Business School
Nadja Dill, Universidad Pontificia Comillas ICADE, Madrid and
Deutsche Lufthansa AG
Dr. Andreas Wald, European Business
School
The special characteristics of
the air cargo business lead to complexities within all the processes of
an air cargo carrier. Based on these complexities, especially the Air
Cargo Revenue Management processes face some major challenges compared to
revenue management applications within other industries. As part of a
research project at the European
Business School,
Competence Centre Aviation Management, the
characteristics of Air Cargo were clustered, complexities analyzed and
approaches developed how to manage those complexities.
Air cargo
overbooking based on the shipment information record - feasibility,
definitions and avenues for further research
Bjoern Becker, Lufthansa Cargo and European Business School
Dr. Andreas Wald, European Business
School
Overbooking models on flight
event level have shown limitations because of the special characteristics
of the cargo business. Within a research project at the European Business
School, Competence
Centre Aviation Management, a new approach for overbooking is currently
developed based on the information given in the shipment record. It has
been analyzed whether this approach - derived from the PNR-based
overbooking on the passenger side - is feasible for air cargo. Basic terms
had to be defined for future research and the data structure of the
shipment information record has been analyzed for possible parameters of
the show up rate. Those possible parameters have been enquired with an
expert questionnaire in order to get a basis for future research. The
results of this enquiry are now presented.
Positioning Revenue Management within the
Organizational Structure
Adam Dudar
Air Canada
Cargo
In recent years, the air cargo industry, like
passenger travel, has become an increasingly commoditized market. To extract the greatest value, focus
has been directed onto the Revenue Management team, and their ability to
command any marginal revenue where possible. Little focus has been placed on how
does the RM team operates within the organization. The formal and informal role of the
team can dictate the success of systems and processes. Leveraging the experience and knowledge
of the team, RM can become the incubator for change, facilitating the
implementation of change in tandem with the various departments of the
organization.
Through this presentation, the role of the revenue
management team will be reviewed, identifying internal and external
forces that can contribute to the success or failure of the practice, as
well as the degree of centralization.
Future Opportunities & Directions for Air
Cargo Revenue Management
Dr. Ricardo V.
Pilon
IBS Software
Services (P) Inc.
Over the long term, air cargo is a reliable growth
industry. However, due to an unpredictable and volatile global economy,
trade flows and thus air cargo movements fluctuate erratically. An
increased need thus exists to ensure that contribution margins are
optimized across a carrier’s entire cargo route network in the short as
well as medium-term. While initial solutions in cargo revenue management
have produced significant results in maximizing profits using short-term
margin maximization decision-support tools, important opportunities exist
with regards to the integration of short-term opportunity costs (bid
prices) and longer-term customer value assessments across an entire
network (value-based CRM and pricing), rather than on an individual
O&D basis.
In this presentation, a number of current practices
are reviewed and used as a foundation for the identification of other
opportunities for further study in an industry that faces sizeable gains
from further automation in cargo margin management and network profit
optimization.
Vendor
Presentations ( Based on submitted abstracts to date* )
Amadeus' Altéa suite of
airline IT solutions
Damian Hickey
Amadeus
Amadeus will outline its view on the current external
and internal pressures facing airlines today, and how this is increasing
the demands placed on their IT infrastructure. They will share with
us how they have worked with many of the worlds leading airlines as their
technology partner in helping them make the transition from legacy to
open platform customer management systems and thus helped them address
many of such pressures. They will also share with us how this has
shaped the evolution of its own leading community based Customer
Management System, Altea.
Official Sponsors
  
 
|
