Thompson
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Thompson has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Michael J. Dixon
Jon M. Huntsman School of Business, Utah State University, Logan, Utah, USA
2 shared publications Heeju Sohn
School of Engineering at Cornell University
1 shared publication- John C. Goodale
Lundquist College of Business, University of Oregon, Eugene, OR 97403-1208, United States
1 shared publication - Rohit Verma
Kellstadt Graduate School of Business at DePaul University
1 shared publication - Sheryl E. Kimes
Cornell University School of Hotel Administration
1 shared publication
Latest Publications
- 18 March 2019
- Published by Wiley in Decision Sciences
- Vol. 50 (5) , 948-984
The Impact of Timing and Bundling Flexibility on Affect‐Based Service Package Design
Past research has provided evidence that customers’ affective responses to service packages, or bundles, include aspects related to the content of the bundle and the timing of bundle elements, or events. Service design efforts that attempt to maximize these affective responses may be inhibited by the degree to which an event is flexible across time and across contextual fit. This article uses numerical simulation of an artificial problem to test hypotheses about the impact of timing flexibility and contextual‐fit flexibility on affect‐based design efforts and provides insight about how differing levels of flexibility constrain or enable designers’ ability to create bundles and schedules that adhere to behavioral‐based design attributes such as the peak, end, trend, and spread effects. Our findings suggest that different kinds of flexibility enable affect‐based schedules in different manners, and managers should consider the impact that flexibility will have on successful affect‐based scheduling efforts.
- 18 March 2019
- Published by Wiley in Decision Sciences
- Vol. 50 (5) , 948-984
The Impact of Timing and Bundling Flexibility on Affect‐Based Service Package Design
Past research has provided evidence that customers’ affective responses to service packages, or bundles, include aspects related to the content of the bundle and the timing of bundle elements, or events. Service design efforts that attempt to maximize these affective responses may be inhibited by the degree to which an event is flexible across time and across contextual fit. This article uses numerical simulation of an artificial problem to test hypotheses about the impact of timing flexibility and contextual‐fit flexibility on affect‐based design efforts and provides insight about how differing levels of flexibility constrain or enable designers’ ability to create bundles and schedules that adhere to behavioral‐based design attributes such as the peak, end, trend, and spread effects. Our findings suggest that different kinds of flexibility enable affect‐based schedules in different manners, and managers should consider the impact that flexibility will have on successful affect‐based scheduling efforts.
- 15 October 2018
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 60 (4) , 378-388
The Value of Timing Flexibility in Restaurant Reservations
A recent paper presented and evaluated 10-integer programming models for restaurant reservations, finding that the models that pooled reservations by same-size tables were superior to models that matched reservations to specific tables. An assumption in all the models was that demand timing was inflexible and the evaluation of the models assumed that all customers arrived exactly at the designated reservation time. Although restaurant customers may have an ideal dining time, many customers have some flexibility in when they would accept a reservation. To address this, we extend the pooled-table models to allow for demand timing flexibility and evaluate the models with a range of differences between customer arrival times and their designated reservation time. With the highest flexibility in demand timing, a top-performing model increased revenue by over 21% compared with rigid demand timing. Fortunately for restaurateurs, the increased revenue came at the expense of only a small deterioration in customer service, as measured by the percentage of parties that need to wait for a table upon arrival.
- 15 October 2018
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 60 (4) , 378-388
The Value of Timing Flexibility in Restaurant Reservations
A recent paper presented and evaluated 10-integer programming models for restaurant reservations, finding that the models that pooled reservations by same-size tables were superior to models that matched reservations to specific tables. An assumption in all the models was that demand timing was inflexible and the evaluation of the models assumed that all customers arrived exactly at the designated reservation time. Although restaurant customers may have an ideal dining time, many customers have some flexibility in when they would accept a reservation. To address this, we extend the pooled-table models to allow for demand timing flexibility and evaluate the models with a range of differences between customer arrival times and their designated reservation time. With the highest flexibility in demand timing, a top-performing model increased revenue by over 21% compared with rigid demand timing. Fortunately for restaurateurs, the increased revenue came at the expense of only a small deterioration in customer service, as measured by the percentage of parties that need to wait for a table upon arrival.
- Open Access
- 10 June 2015
- Published by SAGE Publications in Production and Operations Management
- Vol. 25 (1) , 36-55
Bundling and Scheduling Service Packages with Customer Behavior: Model and Heuristic
Past researchers have found evidence that customers consider the sequence of event utility when evaluating past and future service experiences. Specifically, the evidence confirms that the placement of a peak event, the utility of the last event, and the slope of event utility over time all affect customer behavior and perception. We formulate an optimization problem with a focus on optimizing schedule sequence characteristics in order to maximize customer experiences. We discuss possible contexts in which this type of scheduling might be considered and, as an example, present a particularly complex model of a world‐renowned performing arts venue. We solve the problem with a simulated annealing algorithm and further discuss the complexity and opportunities associated with this type of scheduling effort.
- 10 June 2015
- Published by SAGE Publications in Production and Operations Management
- Vol. 25 (1) , 36-55
Open AccessBundling and Scheduling Service Packages with Customer Behavior: Model and Heuristic
Past researchers have found evidence that customers consider the sequence of event utility when evaluating past and future service experiences. Specifically, the evidence confirms that the placement of a peak event, the utility of the last event, and the slope of event utility over time all affect customer behavior and perception. We formulate an optimization problem with a focus on optimizing schedule sequence characteristics in order to maximize customer experiences. We discuss possible contexts in which this type of scheduling might be considered and, as an example, present a particularly complex model of a world‐renowned performing arts venue. We solve the problem with a simulated annealing algorithm and further discuss the complexity and opportunities associated with this type of scheduling effort.
- 27 January 2015
- Published by SAGE Publications in Journal of Service Research
- Vol. 18 (4) , 498-512
Deciding Whether to Offer “Early-Bird” or “Night-Owl” Specials in Restaurants
In a long history of capacity and demand management research in services, it has often been suggested that pricing discounts and specials can increase demand in off-peak periods. We examine this issue in the contexts of restaurants, where the practices of offering discounts to restaurant patrons for dining early or dining late—commonly known as “early-bird” and “night-owl” specials, respectively—exist throughout the world. These specials bridge marketing and operations—marketing from the goal of increasing customer demand in the off-peak periods and operations from the perspective of having to serve those customers. The effectiveness of these specials has yet to be examined. While simulation would be an ideal tool for predicting the specials’ net revenue benefits, it might be impractical for many restaurateurs, so we develop three simple “back-of-the-envelope” type calculations. Restaurateurs could use these calculations when deciding whether to offer a special. In the eight large simulation-based experiments we conducted, we find that it is important to estimate revenue cannibalization from full-fare customers. The calculations prove to be far more accurate for night-owl specials than for early-bird specials. This has important implications for decisions about offering the specials and raises a flag regarding a potential marketing-operations conflict.
- 27 January 2015
- Published by SAGE Publications in Journal of Service Research
- Vol. 18 (4) , 498-512
Deciding Whether to Offer “Early-Bird” or “Night-Owl” Specials in Restaurants
In a long history of capacity and demand management research in services, it has often been suggested that pricing discounts and specials can increase demand in off-peak periods. We examine this issue in the contexts of restaurants, where the practices of offering discounts to restaurant patrons for dining early or dining late—commonly known as “early-bird” and “night-owl” specials, respectively—exist throughout the world. These specials bridge marketing and operations—marketing from the goal of increasing customer demand in the off-peak periods and operations from the perspective of having to serve those customers. The effectiveness of these specials has yet to be examined. While simulation would be an ideal tool for predicting the specials’ net revenue benefits, it might be impractical for many restaurateurs, so we develop three simple “back-of-the-envelope” type calculations. Restaurateurs could use these calculations when deciding whether to offer a special. In the eight large simulation-based experiments we conducted, we find that it is important to estimate revenue cannibalization from full-fare customers. The calculations prove to be far more accurate for night-owl specials than for early-bird specials. This has important implications for decisions about offering the specials and raises a flag regarding a potential marketing-operations conflict.
- 3 March 2014
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 56 (1) , 91-105
An Evaluation of Rules for Assigning Tables to Walk-in Parties in Restaurants
A survey of 276 restaurant and food-service managers found that, when it comes to seating walk-in guests, just over half maintain a policy of seating the party that has waited the longest. About one-fifth of the respondents seated the largest waiting party and the remainder used some combination of the two approaches. With that in mind, this paper uses an extensive simulation to evaluate the relative revenue outcomes for nine rules for assigning walk-in parties to tables in restaurants, taking into account such variables as restaurant size and matching party size to the table size. I test three variants of each of the “largest party” and “longest wait” rules, along with two rules that blend largest and longest. I also include the outcome of allowing guests to seat themselves. I found that one of the blended rule using party size and waiting time performed the best across all levels of all of the experimental factors, but among the simple rules, the longest wait approach was effective in most cases. Except in the largest of restaurants, allowing guests to seat themselves is the least efficient approach.
- 3 March 2014
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 56 (1) , 91-105
An Evaluation of Rules for Assigning Tables to Walk-in Parties in Restaurants
A survey of 276 restaurant and food-service managers found that, when it comes to seating walk-in guests, just over half maintain a policy of seating the party that has waited the longest. About one-fifth of the respondents seated the largest waiting party and the remainder used some combination of the two approaches. With that in mind, this paper uses an extensive simulation to evaluate the relative revenue outcomes for nine rules for assigning walk-in parties to tables in restaurants, taking into account such variables as restaurant size and matching party size to the table size. I test three variants of each of the “largest party” and “longest wait” rules, along with two rules that blend largest and longest. I also include the outcome of allowing guests to seat themselves. I found that one of the blended rule using party size and waiting time performed the best across all levels of all of the experimental factors, but among the simple rules, the longest wait approach was effective in most cases. Except in the largest of restaurants, allowing guests to seat themselves is the least efficient approach.
- 7 March 2011
- Published by SAGE Publications in Journal of Service Research
- Vol. 14 (2) , 201-213
Cherry-Picking Customers by Party Size in Restaurants
This paper examines the phenomenon of cherry-picking customers—serving the high value customers and denying service to low value customers. These actions are fundamental to revenue management, but have received essentially no attention in the context of restaurants. Two common characteristics of restaurants make them a logical environment in which to implement cherry-picking—compared to smaller parties, larger parties typically take longer to dine and spend less per person. A short example provides motivation and the author reports on an international survey of restaurant managers, which indicates that cherry-picking actions are not uncommon. The author also reports on a simulation study to explore the factors affecting the level of excess demand necessary to justify cherry-picking. Important factors are customers’ willingness to wait for tables, the length of the peak-demand window, and the proportionality of space required by different size tables. The author also evaluates two heuristics for predicting the viability of cherry-picking. The author believes that the most important finding for managers is that the viability of cherry-picking is very dependent on restaurant size, being generally impractical for larger restaurants. From an academic perspective, since this is underinvestigated topic, the author identifies nine research extensions of this work, ranging from empirical to analytical investigations.
- 7 March 2011
- Published by SAGE Publications in Journal of Service Research
- Vol. 14 (2) , 201-213
Cherry-Picking Customers by Party Size in Restaurants
This paper examines the phenomenon of cherry-picking customers—serving the high value customers and denying service to low value customers. These actions are fundamental to revenue management, but have received essentially no attention in the context of restaurants. Two common characteristics of restaurants make them a logical environment in which to implement cherry-picking—compared to smaller parties, larger parties typically take longer to dine and spend less per person. A short example provides motivation and the author reports on an international survey of restaurant managers, which indicates that cherry-picking actions are not uncommon. The author also reports on a simulation study to explore the factors affecting the level of excess demand necessary to justify cherry-picking. Important factors are customers’ willingness to wait for tables, the length of the peak-demand window, and the proportionality of space required by different size tables. The author also evaluates two heuristics for predicting the viability of cherry-picking. The author believes that the most important finding for managers is that the viability of cherry-picking is very dependent on restaurant size, being generally impractical for larger restaurants. From an academic perspective, since this is underinvestigated topic, the author identifies nine research extensions of this work, ranging from empirical to analytical investigations.
- 28 February 2011
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 52 (3) , 241-252
Inaccuracy of the “Naïve Table Mix” Calculations
This article extends the analysis of table mixes for restaurants, on the principle that table mix helps drive revenue. The article presents the results of a simulation experiment of more than forty-six thousand restaurant contexts to evaluate the accuracy of “naïve table mix” calculations. An earlier study presented a simple method of calculating a recommended mix of tables in a restaurant, and another found that those calculations provided table mixes yielding within about 1 percent of the revenue provided by the optimal table mix. While the earlier simulation studies assumed that the space required by a table was directly proportional to the number of seats it contained, that is not always the case. Thus, this article presents space-oriented versions of the existing naïve table mix methods. The article also presents the results of a test of two other forms of the naïve table mix calculations, which include party revenue (or contribution). The simplest of the naïve table mix models proved to be the best, on balance. It yielded table mixes that generated more than 98 percent of the revenue achievable by near-optimal table mixes. As a caveat, however, in more than 5 percent of the simulated scenarios, the best naïve table mix model yielded less than 95 percent of the revenue achievable by a near-optimal table mix. These findings suggest that the naïve table mix calculations be used with caution.
- 28 February 2011
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 52 (3) , 241-252
Inaccuracy of the “Naïve Table Mix” Calculations
This article extends the analysis of table mixes for restaurants, on the principle that table mix helps drive revenue. The article presents the results of a simulation experiment of more than forty-six thousand restaurant contexts to evaluate the accuracy of “naïve table mix” calculations. An earlier study presented a simple method of calculating a recommended mix of tables in a restaurant, and another found that those calculations provided table mixes yielding within about 1 percent of the revenue provided by the optimal table mix. While the earlier simulation studies assumed that the space required by a table was directly proportional to the number of seats it contained, that is not always the case. Thus, this article presents space-oriented versions of the existing naïve table mix methods. The article also presents the results of a test of two other forms of the naïve table mix calculations, which include party revenue (or contribution). The simplest of the naïve table mix models proved to be the best, on balance. It yielded table mixes that generated more than 98 percent of the revenue achievable by near-optimal table mixes. As a caveat, however, in more than 5 percent of the simulated scenarios, the best naïve table mix model yielded less than 95 percent of the revenue achievable by a near-optimal table mix. These findings suggest that the naïve table mix calculations be used with caution.
- 11 May 2010
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 51 (3) , 308-322
Restaurant Profitability Management
While the term restaurant revenue management was defined in this journal in 1998, the history of publications in the Cornell Hospitality Quarterly (CHQ) related to managing restaurant profitability spans nearly fifty years. Of the 160 published articles related to restaurant profitability, more than one-quarter have appeared in the CHQ, which is more than three times that of any other journal. This article presents a new, decision-based framework for restaurant profitability, which expands on the earlier revenue-focused framework. The existing CHQ articles are categorized using the framework, and the gaps are used as the basis for identifying a large number of worthwhile, but as yet unanswered research questions related to restaurant profitability.
- 11 May 2010
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 51 (3) , 308-322
Restaurant Profitability Management
While the term restaurant revenue management was defined in this journal in 1998, the history of publications in the Cornell Hospitality Quarterly (CHQ) related to managing restaurant profitability spans nearly fifty years. Of the 160 published articles related to restaurant profitability, more than one-quarter have appeared in the CHQ, which is more than three times that of any other journal. This article presents a new, decision-based framework for restaurant profitability, which expands on the earlier revenue-focused framework. The existing CHQ articles are categorized using the framework, and the gaps are used as the basis for identifying a large number of worthwhile, but as yet unanswered research questions related to restaurant profitability.
- 5 October 2009
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 50 (4) , 520-539
Time- and Capacity-Based Measurement of Restaurant Revenue
In this article, the authors examine how a common metric of revenue performance—RevPASH, or revenue per available seat-hour—is calculated. In a simple example, a real restaurant, and an extensive simulation study, the authors find inaccuracies in the existing approaches to calculating RevPASH. However, the extent of the inaccuracy is much greater when RevPASH is calculated based on check-open times rather than on the time interval from check open to check close. Since accurate RevPASH values are important in guiding managers’ revenue-enhancing decisions, the article’s findings have importance for practice. Furthermore, as RevPASH has been a commonly used metric for academic research, the article’s findings are important for those performing research on restaurant revenue management.
- 5 October 2009
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 50 (4) , 520-539
Time- and Capacity-Based Measurement of Restaurant Revenue
In this article, the authors examine how a common metric of revenue performance—RevPASH, or revenue per available seat-hour—is calculated. In a simple example, a real restaurant, and an extensive simulation study, the authors find inaccuracies in the existing approaches to calculating RevPASH. However, the extent of the inaccuracy is much greater when RevPASH is calculated based on check-open times rather than on the time interval from check open to check close. Since accurate RevPASH values are important in guiding managers’ revenue-enhancing decisions, the article’s findings have importance for practice. Furthermore, as RevPASH has been a commonly used metric for academic research, the article’s findings are important for those performing research on restaurant revenue management.
- 3 October 2008
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 50 (1) , 96-112
(Mythical) Revenue Benefits of Reducing Dining Duration in Restaurants
This article tests and calibrates an often repeated assumption about the revenue benefits of reducing dining duration. This assumption is that a reduction in dining duration yields a proportional increase in revenue, so that, for example, a 20 percent reduction in dining duration would yield a 25 percent increase in revenue. This article's simulation-based study of over twelve hundred restaurant scenarios finds that, on average, the revenue bump experienced by reducing the dining duration is less than one-quarter of the amount predicted by the common assumption. Even in the most favorable circumstances, the revenue bump is less than one-half that predicted by the assumption. Thus, while reducing dining duration might result in a marginal increase in revenues, managers should not count on a substantial revenue bump.
- 3 October 2008
- Published by SAGE Publications in Cornell Hospitality Quarterly
- Vol. 50 (1) , 96-112
(Mythical) Revenue Benefits of Reducing Dining Duration in Restaurants
This article tests and calibrates an often repeated assumption about the revenue benefits of reducing dining duration. This assumption is that a reduction in dining duration yields a proportional increase in revenue, so that, for example, a 20 percent reduction in dining duration would yield a 25 percent increase in revenue. This article's simulation-based study of over twelve hundred restaurant scenarios finds that, on average, the revenue bump experienced by reducing the dining duration is less than one-quarter of the amount predicted by the common assumption. Even in the most favorable circumstances, the revenue bump is less than one-half that predicted by the assumption. Thus, while reducing dining duration might result in a marginal increase in revenues, managers should not count on a substantial revenue bump.