Worth the wait? How restaurant waiting time influences customer behavior and revenue

• Author: Jelle De Vries,Debjit Roy,René De Koster
• Date: 01 November 2018
• Published date: 01 November 2018
• DOI: http://doi.org/10.1016/j.jom.2018.05.001

Contents lists available at ScienceDirect

Journal of Operations Management

journal homepage: www.elsevier.com/locate/jom

Worth the wait? How restaurant waiting time influences customer behavior

and revenue

Jelle De Vries

a,b,∗

, Debjit Roy

b,c

, René De Koster

b

a

VU University, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands

b

Rotterdam School of Management, Erasmus University, Burgemeester Oudlaan 50, 3062 PA Rotterdam, The Netherlands

c

Indian Institute of Management, Vastrapur, Ahmedabad, Gujarat 380015, India

ARTICLE INFO

Keywords:

Waiting time

Customer behavior

Transaction data

Revenue

Simulation

ABSTRACT

In many service industries, customers have to wait for service. When customers have a choice, this waiting may

influence their service experience, sojourn time, and ultimately spending, reneging, and return behavior. Not

much is known however, about the system-wide impact of waiting on customer behavior and resulting revenue.

In this paper, we empirically investigate this by analyzing data obtained from 94,404 customers visiting a

popular Indian restaurant during a 12 month period. The results show that a longer waiting time relates to

reneging behavior, a longer time until a customer returns, and a shorter dining duration. To find out the impact

of the consequences of waiting time, we use the empirical findings and data collected in a simulation experiment.

This experiment shows that, without waiting, the total revenue generated by the restaurant would increase by

nearly 15% compared to the current situation. Stimulating customers to reserve could enable restaurants to reap

part of this benefit. Furthermore, the results of simulation experiments suggest that, within the boundaries of the

current capacity, revenue could be increased by a maximum of 7.5% if more flexible rules were used to allocate

customers to tables. Alternatively, by increasing the existing seating capacity by 20%, revenue could be boosted

by 7.7% without the need to attract additional customers. Our findings extend the knowledge on the con-

sequences of customer waiting, and enable service providers to better understand the financial and operational

impact of waiting-related decisions in service settings.

1. Introduction

In the U.S. approximately 37 billion hours are spent on waiting in

physical lines annually (Stone, 2012), which adds up to a wait between

two and three years in the lifespan of an average American (Cox, 2005).

This waiting takes place at a variety of service settings, such as res-

taurants, banks, amusement parks, retail stores, and healthcare facil-

ities. Waiting for treatment in a healthcare facility might be unavoid-

able because of the lack of alternative options. In other service settings,

however, customers are apparently consciously choosing to spend

substantial amounts of time in line before they are served. Even though

companies do not directly experience the costs of the discomfort in-

curred by their customers because of waiting, it is not clear to what

extent these costs could have a direct and delayed impact on profit-

ability through customer decisions and actions. In this paper, we em-

pirically investigate several of the implicit consequences of letting

customers wait, and we estimate the impact of these consequences in

various scenarios using simulation.

The importance of waiting in service practice is to a large extent

reflected in the attention academia has devoted to the topic from dif-

ferent perspectives. From an operations perspective, waiting is com-

monly modeled as a cost function in which the wait results from a

mismatch between demand and capacity that could be fixed by

tweaking operational parameters (Osuna, 1985). Actual and perceived

waiting can then be influenced by capacity, layout, and service and

processing policy decisions (Luo et al., 2004;Nie, 2000). A large

number of studies focus on the behavioral consequences of waiting by

showing that long queues can impact aspects such as service evalua-

tions and customer satisfaction (Davis and Maggard, 1990;Houston

et al., 1998;Taylor, 1994), the perceived value of products and services

(Debo et al., 2012;Koo and Fishbach, 2010;Kremer and Debo, 2015),

and customer loyalty (Bielen and Demoulin, 2007;Dube et al., 1994).

At the same time, empirical research and data collection in this domain

is challenging. Whereas virtual queueing settings such as call centers

are characterized by hi-tech environments in which data is abundantly

available (Koole and Mandelbaum, 2002), studies involving physical

queues primarily make use of survey data and self-reports (Munichor

and Rafaeli, 2007;Rafaeli et al., 2002). These (repeat) purchase

https://doi.org/10.1016/j.jom.2018.05.001

Received 5 April 2017; Received in revised form 1 May 2018; Accepted 8 May 2018

∗

Corresponding author.

E-mail address: j3.de.vries@vu.nl (J. De Vries).

Journal of Operations Management 63 (2018) 59–78

Available online 21 May 2018

0272-6963/ © 2018 Elsevier B.V. All rights reserved.

T

intentions do not necessarily lead to actual behavior and corresponding

capacity usage (Chandon et al., 2005). In the current study we cir-

cumvent this limitation by using data on actual customer behavior.

Furthermore, the majority of studies on waiting time and its con-

sequences make several assumptions that might not hold across prac-

tical settings. For example, the arrival rate of customers is commonly

treated as an exogenous, fixed parameter (e.g. Hwang et al., 2010;Roy

et al., 2016). Whereas this might be reasonably accurate if the model

describes only a short period of time, this assumption does not hold in

the long-term. In reality, a customer who faces an excessive waiting

time during a visit to a service provider may renege or never return

after leaving unsatisfied after being served. Very few studies in-

corporate relations between the waiting time and arrival rate at a later

point in time (Ittig, 1994;Umesh et al., 1989), which can substantially

influence revenue and profit(Ittig, 2002).

Additionally, only few studies acknowledge that the experienced

waiting time or the tolerance for waiting might impact customers' service

requirement and duration (Wu et al., 2018). Also the servicestaffhas some

discretion in determining completion time (Hopp et al., 2007). They might

increase their effort and decrease the service duration in response to a

slightly higher workload, but they also might become demotivated and

unproductive in response to excessive workloads (Tan and Netessine,

2014). This implies that observed customer behavior is simultaneously

influencing and influenced by waiting time. Consequently, a service op-

eration should not focus on minimizing the waiting time, but on max-

imizing revenue through minimizing the costs associated with waiting

(Gavirneni and Kulkarni, 2016). To truly understand the operational im-

plications of these dynamics, we combine an empirical model to in-

vestigate the isolated consequences of waiting with a simulation model

incorporating the combined effects of waiting time on reneging, customer

returns, and revenue. This combination between empirical analyses and

simulation enables us to experimentally investigate waiting time in the

context of specific restaurant policies, and provides results that are better

generalizable and meaningful to practice.

More specifically, in this study we aim to address the following

research questions:

RQ1. What are the isolated effects of waiting time on customer

behavior (in terms of reneging, dining duration, and returning)?

RQ2. What are the dynamic consequences of the combined em-

pirically identified effects of waiting time?

RQ3. How do specific proposed operational strategies, such as in-

creasing capacity, flexible seat allocation, and encouraging reservations

impact waiting time and its consequences?

To achieve this, we employ operational field data obtained from

94,404 groups of customers of a restaurant operation. To answer RQ1

we use isolated empirical models to identify the effect of waiting time

on customer behavior. To address RQ2, we embed the empirical models

in a simulation framework to capture the complex interactions and

dynamics of the investigated constructs. This simulation model enables

us to demonstrate the (longer-term) impact of waiting time on customer

returns and revenue by incorporating the endogenous effect of waiting

time on reneging, dining duration, and future arrivals. Subsequently, to

address RQ3, we leverage the integrated simulation model to evaluate

the effect on revenue of various operational policies that restaurants

could deploy. As a consequence, this study should not only help to

improve understanding of the waiting process, but could also lead to

new insights regarding company policies in order to maximize revenue.

The study is therefore divided in two parts. We first develop hy-

potheses, explain the method, and test the hypotheses in Sections 2, 3,

and 4, respectively. We show the impact of waiting time on return

behavior, reneging, and dining duration. We then investigate the im-

pact of several hypothetical operational scenarios on waiting time, re-

turn behavior, reneging, and revenue through simulation. This simu-

lation model and the results of the simulation are explained in Section

5. Section 6draws conclusions and discusses implications for Opera-

tions Management theory and practice.

2. Hypothesis development

2.1. Waiting time

Waiting is in many cases one of the first interactions between ser-

vice providers and customers. Because of this, adequately managing the

waiting time is a vital issue (Davis and Heineke, 1998). Waiting time

can be considered in a subjective way as the waiting time perceived by

the customer, or in an objective way as the actual waiting time. Even

though the actual waiting time might differ from the waiting time

perceived by the customer, actual waiting time is still the most im-

portant predictor of perceived waiting time (Dabholkar, 1990;

Thompson et al., 1996). This study therefore focuses on the impact of

actual waiting time on three outcomes: customer loyalty, reneging, and

dining duration.

2.2. Customer loyalty

For companies operating in competitive markets, obtaining a base of

loyal customers is essential for survival (Srivastava et al., 1998). Cus-

tomer loyalty, which can be defined in terms of repurchase behavior

(Estelami, 2000), repurchase intention (De Ruyter and Bloemer, 1999),

or long-term commitment to repurchase (Ellinger et al., 1999), can

increase profits through reducing the costs associated with acquiring

new customers, through generating a base of customers that is less

price-sensitive, and through lower operational costs due to the famil-

iarity of customers with the procedures and systems of the company

(Hallowell, 1996).

Customer loyalty is especially important in industries with low

switching costs for consumers, as consumers can freely decide to move

their business to competitors (Shapiro and Varian, 2013). In service

contexts such as restaurants, a dissatisfied customer will face virtually

no barriers to dine somewhere else next time. One of the most im-

portant drivers of customer loyalty is service quality (Devaraj et al.,

2001;Stank et al., 1999). The literature on service quality highlights

two critical components: relational elements and operational elements.

Relational elements refer to activities focused on understanding the

needs and expectations of customers. The importance of relational

elements of service quality in determining customer loyalty have been

demonstrated frequently, mainly in the marketing literature (e.g. Bell

et al., 2005;Crosby et al., 1990;Payne and Frow, 2005). Operational

elements, referring to all activities service providers perform to achieve

consistent high level of productivity, quality, and efficiency (Stank

et al., 1999), are essential determinants of service quality as well

(Harvey, 1998). Waiting time is such an important operational element

of service quality. An increased queue length can attract customers on

the short term by signaling quality (Debo et al., 2012;Kremer and

Debo, 2015;Veeraraghavan and Debo, 2009), but this effect is only

expected to apply in case of quality uncertainty and in case alternative

options are available. In deciding whether or not to come back to a

restaurant, customers take the actual experienced quality into con-

sideration. A longer wait during a past visit is therefore not expected to

make customers more likely to come back soon in the future. Conse-

quently, we expect that waiting time will have a negative impact on

customer loyalty, as defined by the time until a customer returns:

H1. A longer waiting time will be associated with longer time until a

customer returns

2.3. Reneging

In addition to the longer-term effect of waiting on customer loyalty,

waiting time can also have more direct implications. When customers

enter a queue, they might observe or be informed about information on

the expected length of delay. Subsequently, customers can make a de-

cision between entering the queue or leaving before even joining the

J. De Vries et al. Journal of Operations Management 63 (2018) 59–78

60