Multi-attribute Hierarchical Revenue Planning
Date Issued
2010
Date
2010
Author(s)
Ma, Chi-Kang
Abstract
Revenue Management (RM), concerned with maximizing a firm’s revenue, is dealing with many issues such as pricing, capacity control and demand management to design the best planning. Maximizing the revenue is to sell the right price at the right place to the right customer. The airlines industry is one of the earliest industries to apply RM and increase revenue by applying dynamic pricing, seat inventory control and overbooking to sell more seats and thus make more revenue. However, most studies of RM focus on short-term demand. For example, deciding a flight’s cabin ticket prices is a short-term planning under the predetermined flight schedule. In practice, firms must consider many product attributes, such as product types and their corresponding market positions, for different stages of planning before product sale.
Pricing and capacity allocation for a product demand with multiple attributes to maximize revenue is known to be a hard problem to solve in the literature. Product pricing needs to consider product attributes. The price in turn changes the demand for each product. The capacity allocation then needs to be adjusted to meet the demand for each product type. Therefore, pricing, capacity control and multi-attribute demand planning are issues that can’t be considered separately. The complexity becomes even greater with the hierarchical planning where different planning stages: long-term, mid-term and short-term planning, are required. The shorter-term planning inherits and has to follow the longer-term planning results. It is thus important to choose right product attributes for different stages of planning. For example, a multi-attribute flight schedule consists of three ticket attributes: “season”, “ticket flexibility” and “seat class”, for long-term, mid-term and short-term capacity planning. If “season” is used first for long-term planning consideration, then the mid-term planning, concerning the “ticket flexibility” or “seat class”, has to be performed under busy or slack season. On the other hand, if the flight schedule uses the “ticket flexibility” or “seat class” first for long-term planning, then the mid-term planning with consideration of “season” has to follow whatever is planned in the longer term. This means that once a product attribute is used for longer-term planning then the shorter-term planning has to inherit the longer-term planning results and is allowed to consider only product attributes other than the one used in the longer-term planning. Therefore, right product attributes have to be chosen for different stages of planning for the multi-attribute hierarchical planning problem..
In this thesis, we consider the multi-attribute hierarchical planning problem to maximize the revenue. We also consider the sharing costs for capacity allocated in the longer-term planning and shared by different product groups. With a given pricing strategy, we use two methods, namely, Mean Proportional Capacity (MPC) and Average Selling Price (ASP), to allocate the capacity and compute the aggregated expected revenue for the longer-term product demand. By capacity allocation of different planning stages, we can determine a sequence of product attribute choices for hierarchical planning to maximize the expected revenue. We refer to such a sequence of product attributes as Revenue-based Demand Planning Hierarchy (RDPH). We use two case studies to explain the calculation of aggregated expected revenue and to demonstrate how RDPH helps the decision making. The first case study is the problem of hospital bed allocation. Three attributes, department (internal medicine or surgical), patient type (emergency or outpatient) and payment type (health insurance or not), for hospital bed are considered. We show how to compare two RDPH with different pricing strategies. Based on the RDPH comparison, we can choose a pricing strategy with the maximum revenue. The second case study is the problem of inpatient abdominal ultrasound. Three attributes, gastrointestinal/hepatobiliary, cancer or not and screening/track, are considered. We use this second case study to show that how the hospital allocates the resources of ultrasound examination for demands coming from disease types with different insurance payment plans. The hospital can compare the abdominal ultrasound RDPHs under two capacity expansion plans and choose the plan maximizing the expected revenue.
Pricing and capacity allocation for a product demand with multiple attributes to maximize revenue is known to be a hard problem to solve in the literature. Product pricing needs to consider product attributes. The price in turn changes the demand for each product. The capacity allocation then needs to be adjusted to meet the demand for each product type. Therefore, pricing, capacity control and multi-attribute demand planning are issues that can’t be considered separately. The complexity becomes even greater with the hierarchical planning where different planning stages: long-term, mid-term and short-term planning, are required. The shorter-term planning inherits and has to follow the longer-term planning results. It is thus important to choose right product attributes for different stages of planning. For example, a multi-attribute flight schedule consists of three ticket attributes: “season”, “ticket flexibility” and “seat class”, for long-term, mid-term and short-term capacity planning. If “season” is used first for long-term planning consideration, then the mid-term planning, concerning the “ticket flexibility” or “seat class”, has to be performed under busy or slack season. On the other hand, if the flight schedule uses the “ticket flexibility” or “seat class” first for long-term planning, then the mid-term planning with consideration of “season” has to follow whatever is planned in the longer term. This means that once a product attribute is used for longer-term planning then the shorter-term planning has to inherit the longer-term planning results and is allowed to consider only product attributes other than the one used in the longer-term planning. Therefore, right product attributes have to be chosen for different stages of planning for the multi-attribute hierarchical planning problem..
In this thesis, we consider the multi-attribute hierarchical planning problem to maximize the revenue. We also consider the sharing costs for capacity allocated in the longer-term planning and shared by different product groups. With a given pricing strategy, we use two methods, namely, Mean Proportional Capacity (MPC) and Average Selling Price (ASP), to allocate the capacity and compute the aggregated expected revenue for the longer-term product demand. By capacity allocation of different planning stages, we can determine a sequence of product attribute choices for hierarchical planning to maximize the expected revenue. We refer to such a sequence of product attributes as Revenue-based Demand Planning Hierarchy (RDPH). We use two case studies to explain the calculation of aggregated expected revenue and to demonstrate how RDPH helps the decision making. The first case study is the problem of hospital bed allocation. Three attributes, department (internal medicine or surgical), patient type (emergency or outpatient) and payment type (health insurance or not), for hospital bed are considered. We show how to compare two RDPH with different pricing strategies. Based on the RDPH comparison, we can choose a pricing strategy with the maximum revenue. The second case study is the problem of inpatient abdominal ultrasound. Three attributes, gastrointestinal/hepatobiliary, cancer or not and screening/track, are considered. We use this second case study to show that how the hospital allocates the resources of ultrasound examination for demands coming from disease types with different insurance payment plans. The hospital can compare the abdominal ultrasound RDPHs under two capacity expansion plans and choose the plan maximizing the expected revenue.
Subjects
multi-attribute
hierarchical planning
revenue management
SDGs
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-99-R97546030-1.pdf
Size
23.53 KB
Format
Adobe PDF
Checksum
(MD5):c33502fbbe99e79f9e25fed5fb35fd6a