|
Article Excerpt
1. Introduction
1.1. Motivation and Research Questions
In recent years, name-your-own-price (NYOP) retailers, exemplified by Priceline[R], have emerged as an online alternative to other traditional channels through which service providers such as airlines, hotels, and car rental companies offer their products to customers. Under its patented system, Priceline collects individual customer offers (price bids guaranteed by a credit card) and communicates the information directly to participating sellers or to their private databases. It operates on a commission plus the difference between the consumer bid and the price it pays the service provider (Kannan and Kopalle 2001). Unlike most other auction models, the rules restrict consumers to a single bid for the service.
Our research employs a stylized model to identify and understand key trade-offs driving the decision by a service provider in the travel industry to employ an NYOP channel, assuming that such a channel is available. The decision to distribute through the NYOP channel requires the existence of forces that counteract the adverse consequences of cannibalization of sales through traditional posted-price channels. Our analysis provides some insight into these forces, and examines the following issues:
* The conditions under which it is profitable for the service provider to contract with the NYOP retailer;
* The prices the service provider should set for the posted-price channel and as the "wholesale price" for the NYOP channel; and
* The impact of vertically integrating the activities of the NYOP retailer on the posted-price decision.
The first two issues address the questions of when a service provider should contract with an NYOP retailer and what prices to set when it is profitable to do so, whereas the third considers the implications of the service provider owning the NYOP channel.
1.2. Overview of Modeling Approach and Key Results
Although in practice an NYOP channel can be implemented in various forms, we employ a stylized, two-stage, game theoretic model with some plausible characteristics of such a channel to capture the behavior of a monopoly service provider and a population of rational consumers of two "types," whom we label as leisure and business travelers. Leisure travelers have the flexibility to plan their travel well in advance of the service date (in the first stage of the game), and are heterogeneous in their willingness to pay for the service. Business travelers finalize their travel plans closer to the date of departure (in the second stage) and have a willingness to pay that exceeds the average reservation price of leisure travelers. The number of business travelers is uncertain until business demand is actually realized in the second stage. Leisure travelers can either buy advance-purchase tickets in the first stage or they can wait for the second stage to place bids with the NYOP retailer (if contracted by the service provider in the first stage for service in the second stage). Our key results, based on the Bayesian Nash equilibrium solution for this game, are as follows:
* Engaging an NYOP retailer is profitable for the service provider only when the available capacity is not too large relative to the expected number of business travelers and their willingness to pay. The relative profitability of adding the NYOP retailer to the service provider's own posted-price channel increases in the expected number of business travelers and their willingness to pay.
* When the service provider contracts with an NYOP retailer, the uncertainty of obtaining the service if leisure travelers postpone purchase to the second stage deflates consumer surplus and is sufficient to separate the low- and high-valuation consumers.
* The service provider is better off setting the wholesale price in the NYOP channel before he is absolutely certain of the extent of business travel demand, because less precise information helps to avoid significant cuts in the wholesale price.
Contracting with the NYOP channel facilitates market segmentation and price discrimination, and allows for disposal of excess capacity after meeting business travel demand. However, this flexibility implies that the service provider can no longer credibly pre-commit to maintaining high prices when there is unsold capacity. Furthermore, if the NYOP retailer is an independent entity, a portion of the revenue generated from NYOP consumers (bid price minus wholesale price) is retained by the retailer.
A key insight from our analysis is that it is the uncertainty in business travel demand that provides the economic rationale for contracting with an NYOP retailer, not the expectation of excess capacity. Indeed, all else equal, the larger the capacity, the less likely it is that contracting with an NYOP retailer is the right decision on the part of the service provider. As discussed in our conclusion, this insight offers a plausible explanation for some observed outcomes with regard to Priceline.
1.3. Related Literature
Fay (2004), Hann and Terwiesch (2003), and Terwiesch et at. (2005) focus on the NYOP retailer's decision with regard to setting a minimum acceptable price and/or bidding rules. Fay (2008) uses a Hotelling model with two service providers to investigate the rationale for the existence of opaque products, although the model assumes that the retailer of the opaque product sells at a posted price, rather than employing a bidding mechanism characteristic of the NYOP pricing model. Amaldoss and Jain (2008) examine whether asking consumers to place a joint bid for multiple items (rather than the current practice of bidding on one item at a time) can increase NYOP retailer's profits.
From the consumers' perspective, Spann and Tellis (2006) test consumers' rationality with reference to their bidding behavior using data of bid sequences for airline tickets at a European NYOP retailer (allowing multiple bids, unlike Priceline). In a different vein, Ding et al. (2005) incorporate the effect of emotion in modeling bidders' behavior in Priceline-like channels, considering the excitement or frustration of having the bid accepted or rejected.
From the service provider's perspective, the emergence of the NYOP retailer provides the opportunity of adding a new channel to the existing posted-price channel. Adding a channel can be risky for the service provider, with potential for cannibalization (Balasubramanian 1998, Chiang et al. 2003). The cannibalization issue in the context of adding an online channel has received recent research attention (Brynjolfsson and Smith 2000, Geyskens et al. 2002, Lal and Sarvary 1999, Zettelmeyer 2000). Because the products offered on the posted-price and NYOP channels are essentially differentiated in terms of their expected value (due to the uncertainty of obtaining the service), our work is more closely related to the theoretical literature on product line decisions (Balachander and Srinivasan 1994, Desai 2001, Moorthy 1984, Mussa and Rosen 1978, Villas-Boas 1998). However, our setting differs from the traditional quality-differentiated product line framework in that in our case the inferior product is offered under a "name-your-own-price" (bidding) mechanism, with a sequential (two-stage) process. Distinct and, as it turns out, critical--to our model is the role played by the uncertainty in demand of the business travel segment in the first stage.
Turning to extant research on yield management, the operation research literature (Belobaba 1989, Belobaba and Wilson 1997) typically treats prices as exogenous. In the economics literature, Dana (1999) and Gale and Holmes (1993) consider the timing of sales in the airline industry. However, all sales are via the service provider's direct channel, and the option of using a reseller is not considered. More recently, a stream of literature has emerged in marketing that focuses on profit maximization for capacity-constrained services. Desiraju and Shugan (1999) use an analytical (two-period, two-segment) model to investigate pricing strategies based on yield management systems (YMSs), including early discounting, limiting early sales, and overbooking. Such strategies can apply to services for which price-insensitive customers buy later than price-sensitive ones, as in the travel industry, which is the focus of our research. In contrast to our results, their results show that in the traditional yield management setting, the larger the capacity, the more attractive a YMS approach (with early discounting and possible limits to early discounted sales) is relative to a single-price strategy. Furthermore, the relative attractiveness of YMSs depends on the size of the price-sensitive (leisure customer) segment.
Other profit enhancing mechanisms may be employed, such as overbooking in the presence of "no shows." Biyalogorsky et al. (1999a) show that by deliberately overselling capacity, service providers can improve profit by accommodating late-arriving high-valuation customers with opportunistic cancellations on low-paying customers. Xie and Gerstner (2007) demonstrate that service providers can improve capacity utilization and increase profit by offering refunds for service cancellation even with late-arriving low-valuation consumers. Other related work includes Biyalogorsky and Gerstner (2004) on contingent pricing and Biyalogorsky et al. (1999b) on service upgrades. In contrast to the above research, we focus on the possible role for a separate NYOP channel in a yield management system.
Next, we develop the model in [section]2. Our analysis and results at equilibrium are...
|
