Coordinated planning in revenue management

Ruff, Volker

Vorschau

PDF
DissVR.pdf - Veröffentlichte Version
Download (1MB)

URL:	https://madoc.bib.uni-mannheim.de/37416
URN:	urn:nbn:de:bsz:180-madoc-374165
Dokumenttyp:	Dissertation
Erscheinungsjahr:	2014
Ort der Veröffentlichung:	Mannheim
Hochschule:	Universität Mannheim
Gutachter:	Fleischmann, Moritz
Datum der mündl. Prüfung:	28 November 2014
Sprache der Veröffentlichung:	Englisch
Einrichtung:	Fakultät für Betriebswirtschaftslehre > ABWL u. Logistik (Fleischmann 2009-)
Fachgebiet:	650 Management
Normierte Schlagwörter (SWD):	Revenue Management , Kundenwert , Produktdifferenzierung
Freie Schlagwörter (Englisch):	revenue management , customer relationship management , fencing
Abstract:	Revenue management has been applied in service industries for more than thirty years. Since then, revenue management has been transferred to other industries like manufacturing or e- fulfillment. Short-term revenue management decisions are taken based on other, longer-term decisions such as decisions about actual capacity, segment-based prices or the price fences in place. While optimization approaches have been developed for each of these planning tasks in isolation, existing approaches typically do not consider interactions between planning tasks. This thesis considers coordinated planning in revenue management, that is the interaction of revenue management decisions with other planning tasks. First, we provide an overview of both the literature on coordinated decision making in the context of revenue management in different industries, and the literature on existing frameworks, which aim to structure the planning tasks around revenue management. We find that the planning tasks relevant to revenue management differ across the industries considered. Moreover, planning tasks are relevant on different hierarchical levels in different industries. We discuss an approach for an industry-independent framework. Based on the relevant planning tasks identified, we investigate the long-term performance of revenue management and therefore the integration of revenue management and customer relationship management. We present a stochastic dynamic programming approach, where the firm’s allocation decision impacts future customer demands by influencing the repurchase probabilities of customers, depending on whether their request has been accepted or rejected. We show that a protection level policy is not necessarily optimal in a two-period setting. In a numerical study, we find that the value of looking ahead in time is low on average but may be substantial in some scenarios. However, the benefit from regular demand updates is considerably higher than the additional value of looking ahead in time on average. Lastly, we investigate the interaction of revenue management and fencing. We account for the trade-off between price-driven demand leakage on the one hand and costs for fencing on the other hand. We show that fencing decisions have an impact on the optimal capacity allocation, but that this is not the case vice versa as the fencing decision does not depend on the allocation decision. Taking both decisions sequentially is therefore optimal. We extend our approach in order to account for additional stock-out-based demand substitution. Then, both decisions depend on each other and firms should take both decisions simultaneously.