1

Houston, We Have a Problem!: The Use of ChatGPT in Responding

to Customer Complaints

Abstract
This study aims to explore the use of ChatGPT-4 in generating management
responses to customer reviews or complaints posted on Tripadvisor. Customer
reviews and management responses are viewed as information sources for
holidaymakers when they are making their decisions. A negative customer review
about a hotel accommodation experience from TripAdvisor together with the
response made by hotel management to this review, and the ChatGPT-4 generated
management response to the same customer review were evaluated by 40 industry
experts based on six dimensions of a service recovery model and three dimensions
of justice that are frequently used by researchers. The findings suggest several
practical implications mainly that the ChatGPT-4 generated management response
satisfies the requirements of an efficient and effective management response. The
quality of ChatGPT-4 generated management responses tends to be extremely high
and they may be generated within seconds and with little effort. In addition to the
several above practical implications, as ChatGPT-4 can measure the severity of
service failures based on customer complaints the study has important implications
for service failures and recovery literature.

Keywords: Management response, complaints, customer reviews, hospitality,

TripAdvisor, ChatGPT, AI, E-WOM

Paper type: Research paper

1. Introduction

Service failures are the fall points in service delivery processes that may threaten the

survival and growth of service businesses (Coulter, 2009; Weber, 2009; Wang et al.,

2014; Villi & Koc, 2018; Liu & Huang, 2023). Due to the general service

characteristics of inseparability, heterogeneity, and perishability, the hospitality

industry is highly prone to service failures (Mok et al., 2013; Li et al., 2021; Ayyildiz et

al., 2023). The impact of these service failures on tourism and hospitality businesses

tends to be rather significant as they may cause adverse outcomes such as customer

dissatisfaction (Tosun et al., 2022), negative word-of-mouth communication (Seo &

Jang, 2021), customer switching behaviour (Zhang et al., 2021), lower service staff

2
morale and motivation (Xie et al., 2022), and a rise in overall costs (Rasouli et al.,

2022). Service failures may not be completely avoided, and inefficient and ineffective

handling of service failures may be avoided (Koc, 2019).

Customers tend to purchase their holidays, make their online reviews, and

complaints increasingly by using online travel agencies (OTAs), such as TripAdvisor,

Booking.com, Expedia, Airbnb, Hostelworld, Agoda, and Lastminute.com (Koc, 2019;

Rita et al., 2022). These OTAs encourage and enable customers to make reviews for

the holidays they have had and to read the reviews to plan their future holidays (Lan

et al., 2022). Research shows that a significant proportion of customers, more than

55% of them, tend to share their experiences on social media platforms (Zhao et al.,

2020). Perhaps more importantly, about 90% of the customers state that they take

online customer reviews into account before making their final hotel purchase

decisions (Akhtar et al., 2019). The overall quality evaluation of a hospitality

establishment by the customers relies significantly on online reviews of customers

due to the intangible nature of services (Rose & Blodgett 2016; Koc, 2019).

According to Sparks et al. (2016), online reviews serve as a convenient and quick

access tool for prospective customers to evaluate and compare hospitality

businesses beforehand. Online reviews can facilitate online interactions not only

among customers but also between customers and service providers (Guan et al.,

2021). Hospitality businesses tend to respond to positive and negative comments

made by customers through OTAs (Xu, 2020). Research shows that providing a

management response (MR), also called a managerial or organisational response, by

responding to customer complaints and online reviews through the OTAs, is a

commonly resorted service recovery method by hospitality businesses (Gu & Ye,

3
2014; Proserpio & Zervas, 2017; Zhao et al., 2020; Sharma et al., 2022). By doing

so, businesses can make apologies and explanations, as well as offer compensation

to customers who tend to be dissatisfied with any aspect of their hospitality

experiences due to the service failures that occurred. Hence, the feelings of injustice

and unfair treatment experienced after service failures may be alleviated or in some

cases, may be completely removed through the management responses (MRs)

provided for customer complaints (Bradley & Sparks, 2012; Gu & Ye, 2014).

Proserpio & Zervas’ (2017) study showed that the provision of efficient and effective

MRs for negative online reviews not only improves a business's image, but also

increases the number of positive reviews, and decreases the number of negative

reviews. Research also shows that an appropriate management response to a

negative review may improve 87% of the customers’ perception of a hotel (NERVAL,

2015). A review of the literature shows that hospitality businesses are increasingly

becoming aware of responding to online customer reviews including complaints (Levy

et al., 2013; Xie et al., 2014; Olson & Ro, 2020; Khogali & Mekid, 2023).

Given the above wide range of implications, service businesses need to place

significant importance on service failures and their recovery through MRs. According

to Baker (2017), the term service recovery refers to all the actions designed and

implemented to resolve customer problems and change their negative attitudes to

prevent negative outcomes, such as customer switching and the spreading of

negative word-of-mouth communications. Today with the rapid developments in

information technology, particularly in artificial intelligence (AI) (Sony et al., 2021;

Yazici-Ayyildiz et al., 2022; Al-Emran & Griffy-Brown, 2023), not only the way and

which customers purchase and experience services, how they respond to service

4
failures, as well as how service businesses design and manage customer

experiences have changed significantly (Koc, 2019; Yazici-Ayyildiz et al., 2022;

Gligor et al., 2023). Although rather recently developed and released, in November

2022, ChatGPT has attracted phenomenal interest from both practitioners and

researchers. Researchers show that as a newly developed AI application, ChatGPT

may have significant potential in carrying out a wide variety of tasks for a wide variety

of functions ranging from human resource management to marketing in several

industries, including tourism and hospitality (Carvalho & Ivanov, 2023; Erul & Isin,

2023; Kalpokiene & Kalpokas, 2023). Against this background, this study investigates

the potential use of ChatGPT in responding to customers’ online reviews efficiently

and effectively. The efficiency and effectiveness of the responses generated by

ChatGPT have been evaluated by a group of 40 tourism and hospitality experts, 20

tourism, hospitality, and marketing academics with PhDs, and 20 five-star hotel

managers in Turkey. The experts have been asked to evaluate the responses

generated by ChatGPT in comparison with the actual management response made

by the hotel management by taking into account the six dimensions of Davidow’s

(2000), used to assess and evaluate MRs, and the three justice dimensions (Oliver,

2000), customers expect after making a complaint.

2. Literature Review

2.1. Service Failures, Customer Complaints, and Online Management Responses

A service failure refers to an occurrence of an error, mistake, deficiency, or problem

during the delivery of a service resulting in a delay or obstacle in meeting the needs

of a customer (Koc, 2019). A service failure is also related to an instance where there

5
is a discrepancy between the expectations of customers and the actual service

delivered (Mattila, 1999; Ayyildiz et al., 2023). Service failures tend to result in

customer dissatisfaction and hinder the achievement of fundamental business

objectives, such as survival and growth (Wang et al., 2014; Harrison-Walker, 2019).

In the event of a service failure, the service business may engage in service recovery

actions to respond to the customer’s grievance and rectify the situation (Mattila,

1999; Koc, 2019). Responding to customers’ complaints is aimed at resolving

problems, modifying customers' negative attitudes, preventing the spreading of

negative word-of-mouth communications, and retaining customers (Miller et al., 2000;

Koc, 2017). Without the appropriate recovery actions customers are likely to express

dissatisfaction as a result of the service failure they encountered (Chan et al., 2007;

Ladhari et al., 2008). Research shows that there is a strong relationship between the

frequency of service failures experienced by customers and their tendency to make a

complaint (Chang et al., 2012; Ayyildiz et al., 2023). Although businesses develop a

variety of means to manage and respond to complaints, there may still be plenty of

customers who are not enchanted with the way and which their problems and

complaints have been handled (Davidow, 2000; Choi et al., 2021). Van Vaerenbergh

et al. (2019) demonstrated that only 20% of the customers stated that they were

satisfied with the management response they received in relation to their complaints.

This means that, in general, managers not only failed to respond to most customer

complaints but also when they did they did so not in a very inefficient manner.

Online reviews, as a form of electronic word-of-mouth communication (E-WOM),

have significantly transformed the way and which tourism and hospitality customers

share their service experiences with each other (Thakran & Verma, 2013). Hence,

6
managers of today are confronted with the significant task of managing online

reviews effectively, particularly negative ones (Xie et al., 2014; Xie et al., 2017).

Online organisational responses or management responses (MRs) provide

businesses the opportunity to resolve customer complaints as well as to express

appreciation for positive reviews (Lan et al., 2022). Responding to a positive

comment shows that the hotel management listens to and cares for the customers

and appreciates customers’ feedback (Gu & Ye, 2014), which, in turn, increases the

probability of the number of positive reviews in the future (Gu & Ye, 2014). On the

other hand, responding to negative reviews is essential to reduce customer

dissatisfaction and recover services effectively (Zhao et al., 2020; Lan et al., 2022).

A review of the literature in the field shows the importance of MRs to online customer

reviews. For instance, in an experimental study, it was found that hotels that provided

MRs experienced 60% higher online bookings compared with the hotels that did not

(Ye et al., 2008). It was stated above that about 90% of prospective hotel customers

took online customer reviews into account before they decided where to stay (Akhtar

et al., 2019). Further research showed that prospective customers were less likely to

make a reservation if they see that the hotel does not respond to negative reviews

(Evans et al., 2012). In another study by Lee et al. (2018), it was found that a good

majority of the respondents (77%) believed that hotels should respond to online

reviews and those that responded to online reviews were considered as businesses

that prioritised customer satisfaction. Moreover, more than 50% of the customers

indicated that they would make a booking at a hotel that provided MRs for online

reviews (Lee et al., 2018).

7
Apart from the fact of whether to provide MRs to customers' online reviews or not, the

quality of the MRs provided to customers’ online reviews is also important. For

instance, Zhang et al.’s (2019) study showed that while high quality (i.e. appropriate,

efficient, and effective MRs) provided for negative online reviews improved 84% of

customers’ impressions of the hotel, inappropriate MRs reduced customers’

likelihood of choosing the hotel by 64%.

In line with the quality of management responses, Wei et al. (2013) proposed two

types of management responses, generic and specific response. A generic response

has standard content and is free from particular issues raised in the particular review

of the customer, while a specific MR is a personalised response that addresses the

issues stated in the customer's review, i.e. it is customised or tailored specifically to

what the customer stated in the review (Wei et al., 2013). Compared with generic

management responses, specific management responses tend to be more effective,

as they establish more trust with customers due to the high quality of messages they

convey (Wei et al., 2013; Zhang et al., 2019).

When hotel management responses are not personalised, they may not be

considered credible and they are only perceived as promotional and advertising

material (Mauri & Minazzi, 2013; Wang & Chaudhry, 2018). For instance, Min et al.

(2015) found that even the mere paraphrasing of a customer complaint increased

customers' attitudes significantly positively. In addition to generic responses, easy-

going management responses provided for negative reviews had a significantly

negative influence on the image of the hotel (Li et al., 2018; Xu et al., 2020; Lan et

al., 2022)

8
Despite several research findings, businesses still underestimate the importance of

online MRs (Alrawadieh & Dincer, 2019; Levy et al., 2013). For instance, Dincer and

Alrawadieh (2017) found that only about 43% of online reviews were responded to by

the management of hospitality establishments, while Xie et al. (2017) showed that

only 15% of online customer reviews were responded to. In different studies,

Alrawadieh & Dincer (2019) and Istanbulluoglu & Sakman (2022) found 47% and

54% response rates, respectively, to customers' online reviews by the management

of hospitality establishments.

Receiving negative online reviews or complaints may be demoralising for the service

providers in the first instance. However, customer complaints enable hospitality

establishments to identify service quality problems, the shortcomings that may cause

dissatisfaction for the customer. Also, customer complaints may enable the business

to develop and implement service recovery strategies and provide feedback for the

improvement of the service or product (Hwang & Mattila, 2020). Hence, receiving a

certain number of complaints is healthy, as no complaints may mean that customers

are unable to make a complaint, or unwilling, as they may feel that their complaints

may not be heard or taken into consideration, and not responded to (Koc, 2019).

Previous research shows that only 4% of dissatisfied customers voice their

dissatisfaction, while 96% of customers would simply switch without saying anything

(TARP, 2007). A significant majority (91%) of dissatisfied and silent customers switch

to other service providers without the service business noticing it (TARP, 2007;

Swanson & Hsu, 2009; Koc, 2019). However, with the increasing availability of

making online complaints directly to the service business or through certain platforms

such as TripAdvisor, Booking.com, and Expedia, the percentage of customers who

9
make a complaint has increased significantly as they tend to find it easier to make an

online complaint (Koc, 2019).

It must be kept in mind that based on the fact that online reviews and MRs are visible

to potential customers, the MRs benefit not only the individual customer who has

been responded to, but also those prospective customers who may be planning to

make a booking in the future (Xie et al., 2017; Xu et al., 2020). Responding to

customer complaints conveys the message to all prospective customers that the

hotel listens to its customers, cares for them, and is ready to make a recovery of the

service (Gu & Ye, 2014; Kim et al., 2015). On the other hand, reviews without MRs

tend to convey the message that the hotel does not have a good service quality

system and it does not care for its customers (Zhang et al., 2019). Prior research

(Xie et al., 2017; Wang & Chaudhry, 2018; Xu et al., 2020) demonstrated that MRs to

previous online customer reviews encourage customers to provide feedback through

online reviews, both negative and positive.

A further important aspect of MRs is that certain tour operators may tend to use

negative online reviews as a bargaining tool against the hotels. Especially, negative

online reviews with no management responses or with no appropriate responses may

be used by certain tour operators to ask hotel management to reduce its prices and

profit margins when the seasonal contracts are made (Filieri et al., 2015). This means

that failure to respond to customer reviews or responding to them ineffectively may

also have immediate financial implications for the hotel, in addition to the above

mentioned problems.

10
2.2. Service Recovery Dimensions and Justice Theory

As stated above, service failures are almost inevitable, due to the service

characteristics such as intangibility, heterogeneity, and inseparability, (Ayyildiz et al.,

2023). This is particularly the case in hospitality (Collie et al., 2000) where various

sub-services that make up the hospitality service involve frequent and intense

interactions between the service providers and customers (Koc, 2017). As service

failures tend to be more or less inevitable, service businesses need to develop and

implement strategies to deal with and recover service failures (Karatepe, 2006).

Research shows that rather than the service failures themselves, the inappropriate

and inefficient handling of and responding to service failures are most likely to lead to

customer dissatisfaction (del Río-Lanza, 2009). It is also important to note that

effective handling of service failures and complaints may convert dissatisfied

customers into satisfied and loyal ones (Bitner et al., 1990; Agnihotri et al., 2022).

MRs are used to handle service failures, and their effectiveness is extremely

important for a wide variety of reasons explained above (Koc, 2019; Xu et al., 2020;

Ayyildiz et al., 2023). Several researchers developed frameworks or models to

investigate the efficiency and effectiveness of MRs to be used for customer

complaints (Boshoff 1999; Smith et al., 1999; Tax et al., 1998; Hoffman and Chung,

1999; Song et al., 2022; Istanbulluoglu & Sakman, 2022; Kapeš et al., 2022).

However, most of the models in these studies appear to have been based on the six

dimensions developed by Davidow (2000) (Kapeš et al., 2022; Istanbulluoglu &

Sakman, 2022). The six dimensions of effective MR and service recovery in

Davidow’s (2000) model are timeliness, facilitation, redress, apology, credibility, and

11
attentiveness. These dimensions may be briefly explained as follows (Davidow, 2000;

Davidow, 2003):

Timeliness: The perceived promptness of an MR and the speed of the handling of a

customer complaint.

The timeliness dimension is to do with the perceived speed of response when an

organisation responds to or handles a complaint (Davidow, 2003). Customers expect

that businesses will implement strategies to resolve their problems quickly. Hence,

prompt management responses play a crucial role in addressing consumer

complaints and turning dissatisfied customers into satisfied ones (Bilgili et al., 2020;

Lan et al., 2022). According to Hogreve et al. (2017), a significant proportion of

customers, over 60%, perceive delay as the most unfavourable outcome in the

service recovery process. Istanbulluoglu (2017) emphasises that customers attach

significant importance to prompt responses to online complaints. The timely handling

of negative online reviews may have a substantial positive influence in terms of

enhancing customer trust in the business (Zhao et al., 2020). Spark et al. (2016)

found that delayed online management responses may cancel out the potential

benefits such as customer satisfaction (Li et al., 2021) that may accrue from a

management response.

Facilitation: The systems, policies, procedures, and organisational arrangements of

a business that facilitate customer communications and making complaints.

The facilitation dimension relates to the policies, procedures, and structure that a

business has in place to support customers to communicate with the business and

12
make a complaint (Davidow, 2003). The facilitation dimension ensures procedural

justice for customers who have experienced a service failure (Karatepe, 2006; Olson

& Ro, 2020). The fact that a hotel takes the time and provides an efficient and

effective management response, and the fact that this management response may

be visible to potential customers is important. It was explained in the literature that

90% of tourism and hospitality customers took online reviews into account (Akhtar et

al., 2019) and 77% of tourism and hospitality customers thought that the provision of

an MR by a hotel was influential in making their purchase decisions (Lee et al.,

2018). Research shows that previous reviews and the fact that they have been

responded to tend to increase customers' willingness to submit their own reviews

(Xie et al., 2017; Zhang et al., 2019; Xu, 2020). This is because responding to

negative online reviews builds trust and improves customers' impressions of the

tourism and hospitality business (TripAdvisor, 2016; Zhang et al., 2019).

Redress: The actual outcome or the compensation the customer receives from the

business in response to the complaint.

Redress is about the benefits or compensation that a customer receives from the

organisation in response to the complaint (Davidow, 2003) and it ensures

distributional justice (Treviño & Morton, 2017; Aguiar-Quintana et al., 2020; Koc,

2020). Redress or distributional justice comprises refunds, free gifts, coupons, and

discounts offered by the service provider in response to a service failure (Aguiar-

Quintana et al., 2020). The act of offering compensation as a response to a

complaint has been shown to have a favourable influence on both satisfaction of

customers and their repurchase behaviour (Bae et al., 2021; Istanbulluoglu &

Sakman, 2022).

13
Researchers express that redress or distributive justice may not be relevant in the

online context (Olson & Ro, 2020) as it is not possible to physically give something

tangible to the dissatisfied customer. However, when managers respond to negative

online reviews by including compensation offers, e.g. a certain percentage of

discount or a free night’s stay at the hotel, redress or distributional justice may be

established (Treviño & Morton, 2017). In other words, businesses may not provide

the compensation physically in an online platform, but they may communicate their

customers to what they offer to compensate for the service failure.

Apology: The recognition of the customer's distress by the business.

An apology serves as a form of psychological compensation (Davidow, 2000) and

represents an acknowledgment of the distress experienced by the dissatisfied

customer (Wang et al., 2009; Van Vaerenbergh et al., 2019). It is important to note

that an apology is not an admission of guilt, but rather a signal that the business

takes the problem seriously and will address it eagerly (Goodman et al., 1987).

Furthermore, research suggests that in certain cases apologies can be more effective

than the monetary-based recovery efforts (Wei et al., 2020), e.g. in the case of

feminine and collectivistic cultures (Koc, 2021). The act of apologising is associated

with the perception of interactional justice by the customer (Karatepe, 2006). Studies

have demonstrated that apologies for online complaints have a positive impact on the

relationships between the business and the customer (Tripp & Gregorie, 2011; Fan &

Niu, 2016; Istanbulluoglu & Sakman, 2022).

14
Credibility: The willingness of a business to offer an explanation or account for the

problem.

A high level of perceived credibility by the customers tends to result in positive

service recovery experiences leading to a higher level of trust (Istanbulluoglu &

Sakman, 2022). This, in turn, fosters the repurchase behaviour (Weitzl & Hutzinger,

2017; Istanbulluoglu & Sakman, 2022).

Attentiveness: Responsiveness of the service provider in interacting with and

establishing interpersonal communication with the customer.

Attentiveness refers to the communication and interaction between the service

provider and the customer making a complaint (Davidow, 2003). It comprises four

distinct dimensions respect, courtesy, empathy, and willingness to listen (Davidow,

2000; Davidow, 2003; Karatepe, 2006). Hence, the way a management response is

expressed, its tone, and the way and which it takes the customer’s emotions into

consideration, rather than merely dealing with the cause of the problem, are

important (Morris, 1988). Research indicates that attentiveness, which corresponds

to responsiveness in the SERVQUAL service quality model (Parasuraman et al.,

1988), has a significant impact on customer satisfaction and repurchase behaviour,

even more than the redress dimension, i.e. the distributional justice (Aguiar-Quintana

et al., 2020; Koc, 2020). Businesses that are genuinely willing to listen to complaints

and show respect and empathy for their customers are likely to provide fair

interpersonal treatment to complainants (Tax et al., 1998; Bowen et al., 1999). This,

15
in turn, is expected to satisfy the interactional justice needs of the customers (Koc,

2020; Assiouras et al., 2023).

Davidow’s (2000) above explained model and its dimensions are based on the justice

theory (del Río-Lanza et al., 2009; Bacile et al., 2020; Istanbulluoglu & Sakman,

2022). Justice theory is a widely used conceptual framework by researchers to

describe customers' evaluations of the service recovery process, i.e. how the

business responds to a service failure and a customer complaint (Treviño & Morton,

2017; Olson & Ro, 2020). According to the justice theory, customers tend to evaluate

a service business' response to a service failure by taking into account distributional

justice, procedural justice, and interactional justice (Oliver, 2000; Koc, 2019; Olson &

Ro, 2020). Karatepe (2006) matched the dimensions in Davidow’s (2000) model with

the three dimensions of justice (See Figure 1).

16
 Figure 1: Matching the Management Response and Perceived Justice
Dimensions

Distributional or distributive justice is about the compensation offered to the customer

to establish equity after the complaint (Smith et al., 1999), and hence it relates to the

redress dimension in Davidow’s (2000) model (Kapeš et al., 2022). On the other

hand, procedural justice is about “the perceived fairness of policies, procedures, and

criteria used by decision-makers in arriving at the outcome of a dispute or

negotiation'' (Blodgett et al., 1997: 189). According to Davidow (2000), businesses

that have established policies, procedures, and resources tend to have a good

complaint handling system and are able to respond to customer complaints promptly.

Hence, the “timeliness” and “facilitation” dimensions in Davidow’s (2000) model are

considered to match with procedural justice (Martínez-Tur et al., 2006; Kapeš et al.,

2022).

Finally, Interactional justice is defined as the perceived fairness by the customer

regarding the interpersonal treatment received when the service business responds

to the service failure (Tax et al., 1998). Interactional justice corresponds with the

apology, credibility, and attentiveness dimensions in Davidow’s (2000) model (Tax et

al., 1998; Smith et al., 1999; Blodgett et al., 1997). An efficient and effective MR is

expected to take Davidow’s (2000) six dimensions and the three types of justice

(Oliver, 2000) into account (Gu & Ye, 2014; Sparks et al., 2016; Kapeš et al., 2022).

Based on the above, this study explores the potential use of ChatGPT in generating

efficient and effective management responses.

2.3. ChatGPT

17
Artificial intelligence (AI) technology applications are models and systems that allow

and guide users to increase their capacity to understand information and how they

may transform the acquired knowledge into efficient applications (Pichai, 2023:2).

ChatGPT is an example of text-based generative AI, a category of AI algorithms, that

may produce more accurate content based on the training data. The generated

output content may be in a wide variety of forms, such as texts, codes, images,

videos, etc.

ChatGPT is based on the GPT (Generative Pre-Trained Transformer) language

model technology developed and released by OpenAI as a public tool (Kirmani,

2022). ChatGPT is a rather sophisticated Chabot that allows the fulfilment of a wide

range of text- photo-, image-, sound and video-based requests, such as providing

responses questions, ranging from basic to more complex, and carrying out

advanced tasks, such as offering solutions to daily and business life problems (Lund

and Wang, 2023; Tekic & Füller, 2023). After its release in November 2022, ChatGPT

has become popular among both the practitioners and researchers across a wide

variety of sectors and industries, such as manufacturing, medicine, education,

tourism and hospitality, banking, etc. to carry out a wide variety of functional tasks

ranging from human resource management to marketing (Carvalho & Ivanov, 2023;

Ray, 2023). ChatGPT's potential to generate valuable natural language text has

created excitement among a variety of stakeholders dealing with the NLP (Natural

Language Processing) and its subscription has reached over 100 million visitors in

just 3 months after its release (Rudolph et al., 2023). Due to the attractiveness of the

NLP, other players, such as Google Bard AI (LaMDA), Microsoft Bing AI, DeepMind

18
Sparrow, and Amazon’s New Language Model have also entered the market (Pichai,

2023). Chat GPT uses deep learning algorithms to analyse input text prompts and

generate responses based on patterns in the data it has been trained on.

One can start using ChatGPT by typing a question, called a prompt, i.e. a request in

the form of a question, in the text box with all the details wanted to be included in the

response. Then, all the user has to do is press the enter button. ChatGPT, depending

on the breadth and depth of the requested material, can generate content within 5 to

60 seconds. This is possible as ChatGPT is trained on an extensive body of text,

ranging from books, articles, and Internet web pages, allowing it to understand

language minute details in different languages and generate high-quality outcomes.

However, it needs to be kept in mind that ChatGPT is not perfect as it may generate

content that is irrelevant to the request, unsatisfactory or that may contain mistakes

(Euchner, 2023; Lecler, 2023; Ray, 2023). Based on the above, this study

investigates whether ChatGPT may help hotel managers and employees to provide

efficient and effective MRs for customer complaints.

3. Methodology

Based on the above explained objectives, this study explores the effectiveness of

ChatGPT in generating efficient and effective responses to customer complaints.

Although ChatGPT’s GPT-3 language model developed by OpenAI is free of charge

(Balas & Ing, 2023), this study used a fee and subscription-based ChatGPT-4, the

latest version of ChatGPT, launched on March 15, 2023, mainly aimed at more

advanced users and software developers (Balas & Ing, 2023; OpenAI, 2023).

19
First, a pool of 30 online hospitality customer complaints, each of which with a

management response, has been compiled from TripAdvisor. Hotel and customer

details were removed from the complaints and the management responses to ensure

privacy. The complaints and the management responses selected were sent to ten

tourism industry experts (five tourism and hospitality academics with PhDs in tourism

and hospitality, and five hotel managers at five-star hotels) to evaluate and rank them

in order to choose the most suitable customer complaint and management response

to include in the study. The draft version of the survey was also sent to the experts.

Based on the feedback from the experts the customer complaint and the

management response provided by the hotel to be included in the study were

identified and the survey was redesigned.

The reason for choosing TripAdvisor to identify complaints and the management

response to include in the study is due to the fact that TripAdvisor is considered to be

the largest online travel operator and travel and tourism review portal in the world

(Filieri et al., 2021; Rita et al., 2022). The number of tourism and hospitality customer

reviews shared on TripAdvisor reached one billion in 2022 (TripAdvisor, 2023; Olson

& Ro, 2020). Also, TripAdvisor is increasingly found to be a significant and relevant

platform to understand tourism and hospitality related customer reviews, customer

complaints, and service quality issues (Filieri et al., 2021; Mellinas & Martin-Fuentes,

2022). A Google Scholar Search of the word TripAdvisor currently returns over 74000

results. The above explanations show that the use of customer reviews management

reviews from TripAdvisor is relevant and important for this study.

20
The suitable customer complaint and management response identified by ten

industry experts (See Figure 2), based on their ranking scores, were decided to be

included in the study. The customer complaint was then entered in ChatGPT-4 and it

was asked to generate a management response by taking into account the six

dimensions of Davidow (2000) that a management response should comprise, and

the three justice dimensions (Oliver, 2000) customers expect after making a

complaint. The customer complaint in Figure 2 is placed on the top left hand and the

hotel's response is below the customer complaint. The ChatGPT generated

management response for the same customer complaint is on the right-hand side of

Figure 2.

21
Figure 2: Customer Complaint, Management Response by the Hotel, and
Management Response Generated by ChatGPT-4

The standard model configuration was used in ChatGPT for generating alternative

responses without any specific training. Recognising the importance of clarity and

specificity in prompt design and preventing unsatisfactory and irrelevant responses

(Ray, 2023), the prompts were structured to be as explicit and unambiguous as

possible. The model’s output was guided by making reference to Davidow’s (2000)

and Oliver’s (2000) dimensions in the prompts. It is known that ChatGPT can

generate different responses to the same query due to its dynamic and real-time

22
nature. In order to have satisfactory responses and to reduce the probability of

unsatisfactory or irrelevant outputs, iterative refinement was exercised.

Then a survey was prepared by using the factual customer complaint selected, the

management response given to it by hotel management, each of which was from

TripAdvisor, and the management response generated for this particular complaint by

ChatGPT-4 (See Figure 2). The survey mainly asked the participants (40 experts) to

evaluate and compare the two management responses, given by the hotel

management, and the management response generated by the ChatGPT-4 in

relation to the customer complaint, on a five-point Likert scale, 5 being extremely

satisfactory and 1 is not satisfactory at all. The participants were asked to assign a

score from 1 to 5, for the hotel management's response, and the ChatGPT-4

generated management response for the same customer complaint against each of

the six dimensions of Davidow (2000), except for timeliness (please see below), and

Oliver’s (2000) three justice dimensions. There was one exception, timeliness, in the

type of questions relating to Davidow’s (2000) dimensions. As the timeliness

dimension was about providing prompt responses to customer complaints (Davidow,

2000), the participants were asked to specify the time it would take for them to write a

management response for the customer complaint provided in Figure 2.

Convenience sampling was used to determine the expert participants to take part in

the study. The experts who accepted to take part in the study comprised 40 industry

experts, of which 20 were tourism and marketing academics, and the remaining 20

were managers at five-star hotels in Turkey. As stated above the experts were asked

to assign scores for each hotel management response and the management

23
response generated by ChatGPT-4 for the same customer complaint posted on

TripAdvisor by taking into account the six dimensions of Davidow (2000), and the

three justice dimensions (Oliver, 2000) customers expect after making a complaint.

Together with the surveys, the participants were also sent clear explanations of

Davidow’s (2000) six dimensions and Oliver's (2000) three justice dimensions and

instructions as to how to make their evaluations.

4. Findings

As explained above this study explored the extent to which ChatGPT, ChatGPT-4 in

particular, can be instrumental in generating efficient and effective management

responses to customer complaints. Tables 1 and 2 show the evaluations of the

industry experts of the ChatGPT generated management response and the hotel

management’s response against the dimensions provided in Figure 2, on a scale of

1 to 5, i.e. from not satisfactory at all to extremely satisfactory. Table 1 shows the

scores assigned by experts across Davidow’s (2000) six dimensions, except for

timeliness, and Table 2 shows the scores for the three justice dimensions. As the

timeliness dimension is about the speed a business responds to or handles a

complaint (Davidow, 2003: 232) it needed a different approach to measure.

Total Average

Credibility Apology Attentiveness Timeliness Redress Facilitation (Five Dimensions -

Excluding
Timeliness)

Personal
ability to
Hotel's Cha Hotel's Hotel's
Chat Hotel's Hotel's Chat produce the Chat Hotel's Chat Chat Chat
Respon t Respon Respons
GPT Response Response GPT managemen GPT Response GPT GPT GPT
se GPT se e
t response
in minutes

24
Tourism ,
Hospitalit
y and
Marketing
1,70 4,75 2,05 4,85 2,15 4,70 29,50 1,45 4,35 1,80 4,60 1,82 4,67
Academic
s

N: 20
46,37
Five-Star secon
Hotel ds
Managers 1,75 4,70 2,05 4,70 1,80 4,65 44,25 1,45 4,55 1,70 4,35 1,76 4,61

N: 20

Total
1,73 4,73 2,05 4,78 1,98 4,68 37,09 1,45 4,45 1,75 4,48 1,79 4,64
N: 40

Table 1: Evaluation of Management Responses According to Service Recovery

Dimensions

Justice Dimensions

Distributional Justice Interactive Justice Procedural Justice Total Justice

Hotel's Hotel's Chat

Hotel's Response Chat GPT Chat GPT Hotel's Response Chat GPT
Response Response GPT

Tourism ,
Hospitality
and
1,45 3,20 1,35 4,60 1,55 4,50 1,46 4,11
Marketing
Academics
N: 20
Five-Star
Hotel
1,60 4,40 2,00 4,65 1,65 4,60 1,63 4,54
Managers
N: 20

Total
1,53 3,80 1,45 4,63 1,60 4,55 1,50 4,22
N: 40

Table 2: Evaluation of Management Responses According to Three Types of Justice

According to Table 1, both the academic experts and the hotel managers the

ChatGPT-4 generated management response attracted significantly high scores

across all six dimensions (Davidow, 2000) ranging from 4,35 to 4,78, out of 5, all of

which may be considered as almost extremely satisfactory. On the other hand, the

25
experts’ scores based on the assessment of the actual management response

provided by the hotel ranged from 1,45 to 2,15, out of 5, i.e. somewhat

unsatisfactory. This finding corresponds with Van Vaerenbergh et al.’s (2019) finding

which found that only 20% of the management responses were considered to be

satisfactory by the customers. Also, the experts believed that ChatGPT-4 generated

management response did extremely well in terms of the three justice dimensions

(Table 2), though the scores (4,11 the average by academic experts, 4,54 by hotel

managers, and 4,22 overall) were slightly lower than the scores assigned for

Davidow’s (2000) dimensions. Below the findings and their analyses under each of

Davidow’s (2000) dimensions and the three justice dimensions (Oliver, 2000) in

relation to them are presented.

Credibility

Table 1 shows that the experts did not find hotel management’s response to the

customer’s complaint satisfactory as they assigned an average score of 1,73, out of

5. The scores assigned by the academics and the hotel managers were similar,

respectively 1, 70, and 1,75. As explained above the credibility dimension is about

the fact that the business is aware of the problem, it is willing to account for it, taking

certain measures, and doing its best to recover it (Karatepe 2006, Yavas et al.,

2004).

Apology

Table 1 shows that the scores for the hotel’s actual management response are

relatively higher for the apology dimension compared with other dimensions. Yet, the

average of 2,05 out of 5 (See Table 1) cannot be considered satisfactory. The way

the apology expressed in the ChatGPT generated management response was

26
considered to be much better, 4,78 out of 5, i.e. extremely satisfactory. This was

probably due to the fact that the ChatGPT-4 response was considered by the experts

that did not merely pay lip service as several actual management responses do

(Mauri & Minazzi, 2013; Wang & Chaudhry, 2018).

Attentiveness

Table 2 shows that there is a major difference in the evaluation of the hotel’s

management response and the management response generated by ChatGPT-4 in

terms of the attentiveness dimension as well. While the average score assigned to

the hotel management’s response by the experts was 1,98, the average score

assigned to the ChatGPT-4 generated management response by the experts was

4,68, showing a major difference. This means that ChatGPT-4 generated

management response was found to be significantly more effective in terms of the

attentiveness dimension.

Timeliness

It was explained above that a significant proportion of customer complaints, 43% of

according to Dincer and Alrawadieh (2017) and 85% according to Xie et al. (2017)

were not responded to. It was also shown that providing delayed online management

responses may even cancel out the potential benefits such as customer satisfaction

that may accrue from a management response (Spark et al., 2016). The response

rate to customer complaints on TripAdvisor and other platforms is low (Levy et al.,

2013; Dincer & Alrawadieh, 2017; Xie et al. 2017; Alrawadieh & Dincer, 2019;

Istanbulluoglu & Sakman, 2022) because, providing a management response to a

customer complaint is not easy (Koc, 2019). As shown in Table 2, it would take an

average of 37,09 minutes for the experts to produce a genuine and personalised
27
response to a customer complaint. The time it takes to generate the management

response was only 46,37 seconds (See Figure 3).

Figure 3: Time Needed for the ChatGPT-4 to Generate the Management

Response

Redress

The importance of compensation, redress, or distributional justice in resolving

customer complaints was explained above in the literature (Davidow, 2000; Bae et

al., 2021). It was shown that compensation was instrumental in turning dissatisfied

customers into satisfied ones and encouraging repurchases (Bae et al., 2021;

Istanbulluoglu & Sakman, 2022). It was also stated that delivering compensation on

online platforms was problematic (Treviño & Morton, 2017; Olson & Ro, 2020) as it is

28
not possible to physically give something tangible to the dissatisfied customer.

Probably this was why the experts assigned the lowest scores across all dimensions

to the redress dimension, both for the hotel management’s response (1,45) and the

ChatGPT-4 generated management response (4,45). However, the evaluation of

ChatGPT-4 generated management response (4,45 out of 5) is still extremely

favourable. As explained above, the redress dimension in Davidow’s (2000) model

corresponds with distributive or distributional justice in the justice model (Karatepe,

2006; Kapeš et al., 2022). The distributional justice scores for the hotel

management’s response and the management response generated by Chat-GPT-4

are 1,53 and 3,80 out of 5 (Table 2). This means that ChatGPT-4 generated

management response, though it is relatively low compared with the other justice

dimensions due to the fact that it is not possible to offer a tangible benefit online, it

still appears to perform significantly better in terms of satisfying customers’

distributional justice requirements.

Facilitation

According to the findings shown in Table 2, experts believe that ChatGPT-4’s

management response can make a significant contribution in terms of the facilitation

element. While the experts assigned an average of 1,75 to the hotel's management

response to the given complaint, i.e. poor and not satisfactory, they assigned an

average of 4,48 to the ChatGPT-4 generated management response, i.e. rather

satisfactory. Hence, the experts believe that ChatGPT is instrumental in ensuring the

facilitation of customer complaints. It was explained above that the visibility of

effective management responses may increase customers’ engagement with the

tourism and hospitality business and increase the probability of submitting their

29
comments, both positive a negative (Xie et al., 2017; Zhang et al., 2019; Xu, 2020).

As stated above the facilitation dimension, together with the timeliness dimension in

Davidow’s (2000) model relate to procedural justice (Martínez-Tur et al., 2006; Kapeš

et al., 2022), and the procedural justice scores in Table 2 show that the ChatGPT-4

generated management response makes a strong contribution in terms of satisfying

the procedural justice requirements.

5. Data Analysis, Research Contributions, and Discussion

Data Analysis

This study showed that ChatGPT can make a significant positive impact in terms of

generating efficient and effective management responses, both for positive and

negative online comments, in a short period of time, in fact in less than a minute.

Regarding the credibility dimension it was seen in the ChatGPT-4 generated

management response that the complaint is handled effectively and the measures

the business intends to implement in order to prevent the reoccurrence of the failure

in the future (Van Vaerenbergh et al., 2019) are stated. It can also be seen that the

generated message is not generic, i.e. it addresses all the specific issues stated by

the customer, it is comprehensive (Mauri & Minazzi, 2013; Wang & Chaudhry, 2018),

and it does not have an easy-going approach (Li et al., 2018; Xu et al., 2020; Lan et

al., 2022)

As stated above, a high level of perceived credibility by the customers is associated

with positive service recovery experiences that foster trust in the business and

encourage repurchase behaviour among the customers (Weitzl and Hutzinger, 2017;
30
Istanbulluoglu & Sakman, 2022). It is known that the apology dimension in Davidow’s

model, together with credibility and attentiveness are derived from interactional

justice (Karatepe, 2006). The overall scores assigned by the experts for the hotel's

management response and the ChatGPT-4 generated management response from

the perspective of interactional justice are respectively 1,45 and 4,63, out of 5 (Table

2). In other words, the management response generated by ChatGPT-4 is believed to

make a significant contribution in terms of providing a response that satisfies the

interactional justice requirements of customers (Oliver, 2000; Karatepe, 2006; Kapeš

et al., 2022).

Regarding the apology dimension it is known that several researchers (Davidow,

2000; Karatepe, 2006; Olson & Ro, 2020; Honora et al., 2023) emphasised the

importance of making effective apologies to customer complaints, i.e. they are well-

written in the language the complaint was made, are genuine, i.e. not general and

address the individual customer for the specific complaint, and they do not sound as

if they have been written just to pay lip service (Mauri & Minazzi, 2013; Wang &

Chaudhry, 2018; Li et al., 2018; Van Vaerenbergh et al., 2019; Lan et al., 2022). As

mentioned above, the apology dimension, together with attentiveness and credibility

match the interactional justice dimension in the justice model (Karatepe, 2006; Kapeš

et al., 2022), and the experts’ scores show that there is a significant difference

between the hotel’s actual management response and the management response

(1,45 and 4,63) (Table 2). The ChatGPT-4 generated management response, with a

score of 4,63, appears to satisfy the interactional justice requirements of the

customers.

31
The findings regarding the attentiveness dimensions have also important

implications. As explained above, the attentiveness dimension relates to the

responsiveness dimension in the SERVQUAL model (Parasuraman et al., 1988) and

plays a more significant role in customer satisfaction and repurchase decisions than

the redress dimension, i.e. the distributional justice (Aguiar-Quintana et al., 2020;

Koc, 2020). The justice perspective explanations for this dimension are provided

above under the credibility and apology dimensions (see also Table 2), as three of

these dimensions in Davidow’s (2000) model match with interactional justice in the

justice theory model (Oliver, 2000; Karatepe, 2006; Kapeš et al., 2022).

The findings in Table 1 show that the ChatGPT-4 generated response may also

satisfy the requirements for the timeliness dimension. Given the fact that even a five-

star hotel with a few hundred rooms and several thousands of night-stays may

receive significantly more than 100 negative online reviews per month. Then, the time

needed to respond to these negative comments may take approximately 62 hours

(i.e. 37,09 minutes for each complaint x 100 complaints ÷ 60 minutes = 61,8 hours)

(See Table 1). This figure is significantly more than the working hours per person per

week. Yet, the responses provided in this manner may still be unsatisfactory, and

they may be evaluated on average as 1,79 out of 5 (Table 1), as can be seen in the

hotel management response in Figure 2. ChatGPT-4, not only provides a much

speedier response, but it does so in a very efficient manner, in several languages,

and without any typographical errors or grammar mistakes. Hence, by allowing the

generation of timely, efficient, and effective management responses ChatGPT-4, may

not only help managers and employees to post a management response, but may

also encourage and facilitate customers to make more reviews, both positive and

32
negative (Xie et al., 2017; Zhang et al., 2019; Xu, 2020). It would also take a lot of

time and burden off the shoulders of managers which can be spent more productively

at work.

Together with the facilitation dimension, the timeliness dimension of Davidow's

(2000) model match with procedural justice (Martínez-Tur et al., 2006; Kapeš et al.,

2022). The procedural justice scores for the hotel management’s response and the

management response generated by Chat-GPT-4 are 1,60 and 4,55 out of 5 (Table

2). This means that ChatGPT-4 generated management response appears to

perform significantly better in terms of satisfying procedural justice requirements.

Research Contributions

Research shows that hotels that provide management responses experience 60% to

77% higher online bookings (Ye et al., 2008; Lee et al., 2018). Customers appear to

be less likely to make a reservation if they see that a hotel does not respond to

negative online reviews (Evans et al., 2012; Lee et al., 2018; Koc, 2019). Yet, a

significant proportion of customer complaints, 43% according to Dincer and

Alrawadieh (2017) and 85% according to Xie et al. (2017) were not responded to

through providing efficient and effective management responses that had paramount

importance for the customers. This is mainly because writing a management

response to several customer complaints requires a lot of time and effort. It requires

skills in terms of understanding the service recovery dimensions (Davidow, 2000) and

justice expectations (Oliver, 2000). Also, writing management responses requires the

ability to write, a good knowledge of the operations, together with language skills to

write management responses in different languages for customers from different

33
parts of the world that make a complaint. Providing management responses to online

negative comments efficiently and effectively not only improves the image of the

business but also increases the number of positive reviews and decreases the

number of negative reviews (Proserpio & Zervas, 2017). Also, research shows that

rather than the service failures themselves, inappropriate and inefficient handling of

and responding to service failure attend to lead to customer dissatisfaction (del Río-

Lanza, 2009). When handled appropriately, service failures and the ensuing

customer complaints may transform dissatisfied customers into satisfied and loyal

ones (Agnihotri et al., 2022).

The analysis of the results of the study showed that ChatGPT-4 generated

management responses can be highly efficient and effective and make several

practical contributions. Additionally, regarding the redress dimension (See Table 1) it

can be seen that ChatGPT-4 can generate a management response that may include

a realistic compensation offer according to the type and severity of a service problem.

This may be extremely important in the training of especially young managers in

tourism and hospitality where employee and manager turnover are high

(Stamolampros et al., 2019; Khan et al., 2021; Dogru et al., 2023). It may also

support the establishment of empowerment which is highly essential in service

recovery situations (Koc, 2013). Perhaps, more importantly, ChatGPT-4 may

distinguish the severity of customer failure and generate different types of

compensation depending on their level of anger and the severity of the perceived

severity of failure. Miller et al.'s (2000) service recovery framework, a model which is

extensively used by researchers in service recovery research, is based on the

severity of the service failure. However, the framework does not offer any

34
explanations as to how to measure the severity and it may be rather difficult to

assess, as it is subjective. Hence, the ability to have a yardstick to evaluate and

compare customers’ levels of anger may provide a good basis for understanding how

severe a service failure is. Therefore, the study has important theoretical implications

as well as the practical implications shown above. This finding may have important

further general service quality design implications apart from carrying out an effective

service recovery strategy. Figure 4 shows examples of two factual customer

complaints from TripAdvisor and how ChatGPT may detect, evaluate customer

anger, and assign anger scores to each, in two languages, English and Turkish.

Discussion

It was explained above that this study has a number of practical and theoretical

implications. Additionally, further exercises with ChatGPT-4 were carried out to

explore its capabilities as an AI tool to support managers. It was identified that when

ChatGPT-4 is provided with a customer complaint it is capable of identifying problem

areas in the customer complaint and assigning severity scores for each problem area

(See Figure 5). As the hospitality services comprise various sub-service activities it

may be difficult to identify the actual service quality gap in tourism and hospitality

(Koc, 2006; Shin, 2022). This feature of ChatGPT-4 may be used as a training tool

for young managers to better understand customer complaints and how problem

areas stated in the complaint or review may correspond with the gaps in the service

quality system (Koc, 2006). Also, as stated above, ChatGPT-4 may distinguish,

assess, measure, and assign scores in terms of the level of anger and severity in

each area of the problem, something currently missing in the service recovery

35
models, such as the case in Miller et al’s (2000), a model, framework which is widely

used by researchers.

Figure 4: Measuring Customer Anger in a Customer Complaint

36
Figure 5: Rating of the Severity of the Customer Complaint

Contrary to common belief the matter is not only responding to negative online

comments as positive comments also need to be responded to. Positive reviews

need to be responded to in a timely, efficient, and effective manner as a

management response to a favourable online review increases the positive image of

the hospitality business (Gu & Ye, 2014; Aktan et al., 2021). This is because an

efficient and effective management response shows that the business is listening to

and caring for the customers and shows appreciation for them (Gu & Ye, 2014).

Wang and Chaudhry (2018) found that inefficient management responses to positive

online reviews may have a negative impact on customers. This is because most

management responses to positive management reviews tend to be written in the

37
form of promotional materials and elaborate on how efficient and effective the

hospitality business is (Wang and Chaudhry, 2018). These sorts of management

responses to positive online reviews are also viewed as shallow acting by the

customers (Wang and Chaudhry, 2018).

As explained above, customer complaints, especially the ones with no management

responses or with no appropriate responses may be used by tour operators to reduce

the profit margin of the hotels (Filieri et al., 2015). Hence, hotels may establish

speech-operated computer terminals (kiosks) with ChatGPT application in various

areas of the hotels. These speech-operated computer terminals (kiosks) may enable

customers to communicate and express their reviews in voice form. As Sparks et al.

(2016) discovered that management responses using a human voice yielded more

favourable customer evaluations. These speech messages may also be transformed

into written reviews in these speech-operated computer terminals (kiosks), Doing so

may reduce the number of customer complaints to be made later by the customers in

online platforms such as TripAdvisor.

Moreover, these speech-operated computer terminals (kiosks) with ChatGPT may

facilitate customers to make direct complaints without confronting anyone as people

from certain cultures may not feel comfortable to make face to face complaints (Koc,

2020). Also, it is known that writing and posting customer complaints may be seen as

burdensome by people (Bourdin & Fayol, 1994; Striepe, 2021). Koc (2020) showed

that people from collectivistic cultures may be harmony oriented and refrain from

making direct and face-to-face complaints as they prefer to avoid being

confrontational.

38
Regarding the cross-cultural issues, when ChatGPT-4 was asked to generate

different management responses to customers from different cultures with different

cultural characteristics, such as the ones offered by Hall (1989) and Hofstede et al.

(2010), it was not capable of doing so. In the future, ChatGPT may provide different

management responses for people from different cultures who may have different

expectations regarding the service recovery and justice dimensions (Koc, 2020).

6. Conclusion and Further Directions

Managers are confronted with the significant task of providing effective, efficient, and

timely management responses to negative or positive online customer reviews.

Online management responses provide businesses an opportunity to resolve

customer complaints quickly, express appreciation for positive reviews, and establish

trust and positive images for the hospitality establishment. Responding to customer

reviews shows that the hotel management listens to and cares for the customers and

appreciates customers' feedback which, in turn, increases the number of positive

reviews in the future. Also, responding to reviews is essential to reduce customer

dissatisfaction and enable successful service recoveries.

This study investigated the ability of ChatGPT-4 to generate management responses

for factual tourism and hospitality customer complaints posted on TripAdvisor. Based

on the evaluations by the industry experts it can be seen that ChatGPT-4 generated

responses satisfy the requirements in Davidow’s (2000) service recovery dimensions

and customers' three types of justice expectations (Oliver, 2000) extremely well. As

ChatGPT-4 may generate efficient and effective responses in a very short time with

39
little effort all customer reviews may be responded to in a timely manner. The study

showed that ChatGPT-4 may also be used as a significant tool to train, especially

young managers, in an industry where staff turnover is high (Stamolampros et al.,

2019; Khan et al., 2021; Dogru et al., 2023).

As in any study, this study has several limitations to consider as well. Firstly, it

focused on customer complaints and management responses to European five-star

hotels. Future, studies may have a much wider perspective in terms of countries and

hotels and may take a cross-cultural perspective. Also, this study resorted to the

views of 40 industry experts, 20 academics, and 20 five-star hotel managers. Future

studies may use a larger group of participants, and may also investigate the

customers’ evaluations.

This study concentrated on the hospitality industry, particularly the hotels. Future

studies may explore the use of ChatGPT in generating management responses in

other industries both in the service sector and in manufacturing. Also, the

instrumentality of ChatGPT may be explored in different fields other than customer

complaints and management responses. Several problematic and tedious tasks

across a variety of other functions such as human resource management require

significant time and effort to carry out. These sorts of tasks that require timeliness

may be determined and the potential use of ChatGPT in handling these tasks may be

investigated. Also, as service recovery attempts may be evaluated differently by

people with different cultural characteristics, future research may ask customers from

different cultures to assign a weight to the dimensions in Davidow’s (2000)

management response model and the three justice dimensions (Oliver, 2000).

40
Lastly, it needs to be noted that from time to time ChatGPT may generate varying

responses for the same prompt due to its real-time and dynamic nature. Although not

experienced in this particular study during its various phases and iterations, this

variability might influence the consistency of AI responses generated for a study.

Hence, the researchers and practitioners need to bear in mind the variable nature of

responses that may be generated by ChatGPT.

References

Agnihotri, D., Kulshreshtha, K., Tripathi, V., & Chaturvedi, P. (2022). Actions speak louder than words: An impact
of service recovery antecedents on customer delight in quick-service restaurants. Asia-Pacific Journal of
Business Administration, 14(4), 421-444.
Aguiar-Quintana, T., Araujo-Cabrera, Y., & Park, S. (2020). The sequential relationships of hotel employees’
perceived justice, commitment, and organizational citizenship behaviour in a high unemployment context.
Tourism Management Perspectives, 35, 100676.
Akhtar, N., Sun, J., Akhtar, M. N., & Chen, J. (2019). How attitude ambivalence from conflicting online hotel
reviews affects consumers’ behavioural responses: The moderating role of dialecticism. Journal of
Hospitality and Tourism Management, 41, 28-40.
Aktan, M., Zaman, U., & Nawaz, S. (2021). Examining destinations’ personality and brand equity through the lens
of expats: Moderating role of expat’s cultural intelligence. Asia Pacific Journal of Tourism Research, 26(8),
849-865.
Al-Emran, M., & Griffy-Brown, C. (2023). The role of technology adoption in sustainable development: Overview,
opportunities, challenges, and future research agendas. Technology in Society, 102240.
Alrawadieh, Z., & Dincer, M. Z. (2019). Reputation management in cyberspace: evidence from Jordan’s luxury
hotel market. Journal of Hospitality and Tourism Technology, 10(1), 107-120.
Assiouras, I., Skourtis, G., Giannopoulos, A., Buhalis, D., & Karaosmanoglu, E. (2023). Testing the relationship
between value co-creation, perceived justice and guests’ enjoyment. Current Issues in Tourism, 26(4), 587-
602.
Ayyildiz, T., Ayyildiz, A. Y., & Koc, E. (2023). Illusion of control in service failure situations: customer
satisfaction/dissatisfaction, complaints, and behavioural intentions. Current Psychology, 1-16.
https://doi.org/10.1007/s12144-023-04292-y
Bacile, T. J., Wolter, J. S., Allen, A. M., & Xu, P. (2018). The effects of online incivility and consumer-to-consumer
interactional justice on complainants, observers, and service providers during social media service
recovery. Journal of Interactive Marketing, 44(1), 60-81.
Bae, G., Lee, S., & Kim, D. Y. (2021). Interactions between service recovery efforts and customer characteristics:
Apology, compensation, and empowerment. Journal of Quality Assurance in Hospitality & Tourism, 22(2),
218-244.
Baker, M. A. (2017). Service failures and recovery: theories and models. In Service Failures and Recovery in
Tourism and Hospitality: A Practical Manual, 24-41. Wallingford UK: Cabi.
Balas, M., & Ing, E. B. (2023). Conversational AI models for ophthalmic diagnosis: Comparison of ChatGPT and
the isabel pro differential diagnosis generator. JFO Open Ophthalmology, 1, 100005.
https://doi.org/10.1016/j.jfop.2023.100005

41
Bilgili, B., Ozkul, E., & Koc, E. (2020). The influence of colour of lighting on customers’ waiting time
perceptions. Total Quality Management & Business Excellence, 31(9-10), 1098-1111.
Bitner, M. J., Booms, B. H., & Tetreault, M. S. (1990). The service encounter: Diagnosing favorable and
unfavorable incidents. Journal of Marketing, 54, 71–84.
Blodgett, J. G., Hill, D. J., & Tax, S. S. (1997). The effects of distributive, procedural, and interactional justice on
postcomplaint behavior. Journal of retailing, 73(2), 185-210.
Boshoff, C. (1999). RECOVSAT: an instrument to measure satisfaction with transaction-specific service recovery.
Journal of Service Research 1 (3), 236–249.
Bourdin, B., & Fayol, M. (1994). Is written language production more difficult than oral language production? A
working memory approach. International journal of psychology, 29(5), 591-620.
Bowen, D. E., Gilliland, S. W., & Folger, R. (1999). HRM and service fairness: How being fair with employees
spills over to customers. Organizational Dynamics, 27(3), 7-23.
Bradley, G., Sparks, B., 2012. Explanations: if, when, and how they aid service recovery. Journal of Services
Marketing, 26(1), 41-51.
Carvalho, I. and Ivanov, S. (2023). ChatGPT for tourism: applications, benefits and risks. Tourism Review, Vol.
ahead-of-print No. ahead-of-print. https://doi.org/10.1108/TR-02-2023-0088
Chan H, Wan LC, Sin LYM. (2007). Hospitality service failures: who will be more dissatisfied? International
Journal of Hospitality Management 26(3), 531–545.
Chang, J., Khan, M. A., & Tsai, C. T. (2012). Dining occasions, service failures and customer complaint
behaviours: An empirical assessment. International Journal of Tourism Research, 14(6), 601-615.
Choi, S., Mattila, A. S., & Bolton, L. E. (2021). To err is human (-oid): how do consumers react to robot service
failure and recovery?. Journal of Service Research, 24(3), 354-371.
Collie, T.A., Sparks, B., Bradley, G. (2000). Investing in interactional justice: a study of the fair process effect
within a hospitality failure context. Journal of Hospitality & Tourism Research, 24(4), 448–472.
Davidow, M. (2000). The bottom line impact of organizational responses to customer complaints. Journal of
Hospitality & Tourism Research, 24(4), 473-490.
Davidow, M. (2003). Organizational responses to customer complaints: What works and what doesn’t. Journal of
Service Research, 5(3), 225-250.
del Río-Lanza, A. B., Vázquez-Casielles, R., & Díaz-Martín, A. M. (2009). Satisfaction with service recovery:
Perceived justice and emotional responses. Journal of Business Research, 62(8), 775-781.
Dinçer, M. Z., & Alrawadieh, Z. (2017). Negative word of mouse in the hotel industry: A content analysis of online
reviews on luxury hotels in Jordan. Journal of Hospitality Marketing & Management, 26(8), 785-804.
Dogru, T., McGinley, S., Sharma, A., Isık, C., & Hanks, L. (2023). Employee turnover dynamics in the hospitality
industry vs. the overall economy. Tourism Management, 99, 104783.
Erul, E., & Isin, A. (2023). ChatGPT ile Sohbetler: Turizmde ChatGPT’nin Önemi (Chats with ChatGPT:
Importance of ChatGPT in Tourism). Journal of Tourism & Gastronomy Studies, 11(1), 780-793.
Euchner, J. (2023). Almost Human. Research-Technology Management, 66(2), 10-11.
Evans, D., Oviatt, J., Slaymaker, J. and Tapado, C. (2012). An experimental study of how Restaurant- Owners’
responses to negative reviews affect readers’ intention to visit. The Four Peaks Review, 2, 1-12.
Fan, Y., & Niu, R. H. (2016). To tweet or not to tweet? Exploring the effectiveness of service recovery strategies
using social media. International Journal of Operations & Production Management, 36(9), 1014-1036.
Filieri, R., Acikgoz, F., Ndou, V., & Dwivedi, Y. (2021). Is TripAdvisor still relevant? The influence of review
credibility, review usefulness, and ease of use on consumers’ continuance intention. International Journal of
Contemporary Hospitality Management, 33(1), 199–223.
Filieri, R., Alguezaui, S., & McLeay, F. (2015). Why do travelers trust TripAdvisor? Antecedents of trust towards
consumer-generated media and its influence on recommendation adoption and word of mouth. Tourism
Management, 51, 174-185.
Gligor, D., Bozkurt, S., & Welch, E. (2023). Building theoretical sand castles: the case of customer brand
engagement. Journal of Marketing Analytics, 1-20. https://doi.org/10.1057/s41270-023-00227-5
Goodman, J. A., Malech, A., & Boyd, S. (1987). Danger, angry customer. ABA Banking Journal, 79(1), 63-66.
Gu, B. and Ye, Q. (2014). First step in social media: measuring the influence of online management responses on
customer satisfaction. Production and Operations Management, 23(4), 570–582.

42
Guan, X., Xie, L., Shen, W. G., & Huan, T. C. (2021). Are you a tech-savvy person? Exploring factors influencing
customers using self-service technology. Technology in Society, 65, 101564.
Hall, E. T. (1989). Beyond culture, New York: Anchor Press.
Harrison-Walker, L. J. (2019). The critical role of customer forgiveness in successful service recovery. Journal of
Business Research, 95, 376–391.
Hoffman, K.D., & Chung, B.G. (1999). Hospitality recovery strategies, customer preference Versus firm use.
Journal of Hospitality & Tourism Research 23(1), 71–84.
Hofstede, G., (2010). Cultures and Organizations: Software of the Mind. McGraw Hill, 3rd edition.
Hogreve, J., Bilstein, N., & Mandl, L. (2017). Unveiling the recovery time zone of tolerance: When time matters in
service recovery. Journal of the Academy of Marketing Science, 45, 866-883.
Honora, A., Chih, W. H., & Wang, K. Y. (2022). Managing social media recovery: The important role of service
recovery transparency in retaining customers. Journal of Retailing and Consumer Services, 64, 102814.
Hwang, Y., & Mattila, A. S. (2020). The impact of customer compassion on face-to-face and online
complaints. Journal of Hospitality Marketing & Management, 29(7), 848-868.
Istanbulluoglu, D. (2017). Complaint handling on social media: The impact of multiple response times on
consumer satisfaction. Computers in Human Behavior, 74, 72-82.
Istanbulluoglu, D., & Sakman, E. (2022). Successful complaint handling on social media predicts increased
repurchase intention: The roles of trust in company and propensity to trust. European Management Journal.
https://doi.org/10.1016/j.emj.2022.06.004
Kalpokiene, J., & Kalpokas, I. (2023). Creative encounters of a posthuman kind–anthropocentric law, artificial
intelligence, and art. Technology in Society, 72, 102197.
Kapeš, J., Keča, K., Fugošić, N., & Čuić Tanković, A. (2022). Management response strategies to a negative
online review: Influence on potential guests’ trust. Tourism and hospitality management, 28(1), 1-27.
Karatepe, O. M. (2006). Customer complaints and organizational responses: the effects of complainants’
perceptions of justice on satisfaction and loyalty. International Journal of Hospitality Management, 25(1), 69-
90.
Khan, A. N., Khan, N. A., & Bodla, A. A. (2021). The after-shock effects of high-performers turnover in hotel
industry: a multi-level study. International Journal of Contemporary Hospitality Management, 33(10), 3277-
3295.
Khogali, H. O., & Mekid, S. (2023). The blended future of automation and AI: Examining some long-term societal
and ethical impact features. Technology in Society, 73, 102232.
Kim, W.G., Lim, H. and Brymer, R.A. (2015). The effectiveness of managing social media on hotel performance.
International Journal of Hospitality Management, 44, 165-171.
Kirmani, A. R. (2022). Artificial intelligence-enabled science poetry. ACS Energy Letters, 8, 574-576.
Koc, E. (2006). Total quality management and business excellence in services: The implications of all-inclusive
pricing system on internal and external customer satisfaction in the Turkish tourism market. Total Quality
Management & Business Excellence, 17(7), 857-877.
Koc, E. (2013). Power distance and its implications for upward communication and empowerment: Crisis
management and recovery in hospitality services. The International Journal of Human Resource
Management, 24(19), 3681-3696.
Koc, E. (2017). Service failures and recovery in tourism and hospitality: A practical manual. Wallingford, Oxford:
CABI.
Koc, E. (2019). Service failures and recovery in hospitality and tourism: A review of literature and
recommendations for future research. Journal of Hospitality Marketing & Management, 28(5), 513-537.
Koc, E. (2020).Cross-cultural Aspects of Tourism and Hospitality: Service Marketing and Management
Perspectives, Routledge, London, doi: 10.4324/9781003018193.
Koc, E. (2021). Intercultural competence in tourism and hospitality: Self-efficacy beliefs and the Dunning Kruger
Effect. International Journal of Intercultural Relations, 82, 175-184.
Ladhari R., Brun I., Morales M. (2008). Determinants of dining satisfaction and post-dining behavioral intentions.
International Journal of Hospitality Management 27(4), 563–573.
Lan, T., Feng, X., & Zeng, Z. (2022). Effect of B&B host responses to online reviews on subsequent reviews: The
moderating effects of class level. Tourism Review, 77(4), 1097-1115.

43
Lecler, A., Duron, L., & Soyer, P. (2023). Revolutionizing radiology with GPT-based models: Current applications,
future possibilities and limitations of ChatGPT. Diagnostic and Interventional Imaging, 104(6), 269-274.
Lee, Y., J. Xie, K., Besharat, A., and Tan, Y. (2018). Should managerial responses to online word-of-Mouth be
prescriptive?. Impacts on firm performance. http://dx.doi.org/10.2139/ssrn.2919553
Levy, S. E., Duan, W., & Boo, S. (2013). An analysis of one-star online reviews and responses in the Washington,
DC, lodging market. Cornell Hospitality Quarterly, 54(1), 49–63.
Li, C., Cui, G., & Peng, L. (2018). Tailoring management response to negative reviews: The effectiveness of
accommodative versus defensive responses. Computers in Human Behavior, 84, 272-284.
Li, F., Lu, H., Hou, M., Cui, K., & Darbandi, M. (2021). Customer satisfaction with bank services: The
role of cloud services, security, e-learning and service quality. Technology in Society, 64, 101487.
Li, Z., Hua, C., Fu, X., & Liu, X. (2021). Beyond complainers: Reclassifying customers of travel agency regarding
post-failure behavior and loyalty performance. Asia Pacific Journal of Tourism Research, 26(3), 329–346.
Liu, D., Lv, Y., & Huang, W. (2023). How do consumers react to chatbots' humorous emojis in service
failures. Technology in Society, 73, 102244.
Lund, B. D., & Wang, T. (2023). Chatting about ChatGPT: How may AI and GPT impact academia and libraries?.
Library Hi Tech News, 40(3), 26-29.
Martínez‐Tur, V., Peiró, J. M., Ramos, J., & Moliner, C. (2006). Justice perceptions as predictors of customer
satisfaction: The impact of distributive, procedural, and interactional justice. Journal of Applied Social
Psychology, 36(1), 100-119.
Mattila, A. S. (1999). An examination of factors affecting service recovery in a restaurant setting. Journal of
Hospitality & Tourism Research, 23(3), 284–298.
Mauri, A. G., & Minazzi, R. (2013). Web reviews influence on expectations and purchasing intentions of hotel
potential customers. International Journal of Hospitality Management, 34, 99-107.
Mellinas, J. P., & Martin-Fuentes, E. (2022). Tourism Online Reviews: Databases and Samples. In Advanced
Research Methods in Hospitality and Tourism (pp. 127-143). Emerald Publishing Limited.
Miller, J., Craighead, C., & Karwan, K. (2000). Service recovery: A framework and empirical investigation. Journal
of Operations Management, 18(4), 387–400.
Min, H., Lim, Y., & Magnini, V. P. (2015). Factors affecting customer satisfaction in responses to negative online
hotel reviews: The impact of empathy, paraphrasing, and speed. Cornell Hospitality Quarterly, 56(2), 223-
231.
Mok, C., Sparks, B., & Kadampully, J. (2013). Service quality management in hospitality, tourism, and leisure.
Routledge.
Morris, S. V. (1988). How many lost customers have you won back today?: An aggressive approach to complaint
handling in the hotel industry. The Journal of Consumer Satisfaction, Dissatisfaction and Complaining
Behavior, 1, 86-92.
Nerval Corp. (2015). INFOGRAPHIC: How online reviews are impacting the hotel industry. Travel Pulse.
Retrieved from: https://www.travelpulse.com/News/Hotels-and-Resorts/INFOGRAPHIC-How-Online-
Reviews-Are-Impacting-the-Hotel-Industry
Oliver, R. L. (2000). Customer satisfaction with service. In T. A. Swartz & D. Iacobucci (Eds), Handbook of
services marketing and management (pp. 247–254). Thousand Oaks, CA: Sage.
doi:10.4135/9781452231327
Olson, E. D., & Ro, H. (2020). Company response to negative online reviews: The effects of procedural justice,
interactional justice, and social presence. Cornell Hospitality Quarterly, 61(3), 312-331.
OpenAI. (2023). GPT-4 Technical Report. Retrieved from: https://arxiv.org/abs/2303.08774
https://doi.org/10.48550/arXiv.2303.08774
Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1988). SERVQUAL: a multiple-item scale for measuring
consumer perceptions of service quality. Journal of Retailing, 64(1), 12-40.
Pichai, S. (2023). An important next step on our AI journey. Retrieved from:
https://blog.google/technology/ai/bard-google-ai-search-updates/.
Proserpio, D., Zervas, G. (2017). Online reputation management: estimating the impact of management
responses on consumer reviews Marketing Science, 36(5), 645-665.
Rasouli, N., Rasoolimanesh, S. M., Rahmani, A. K., Momayez, A., & Torabi, M. A. (2022). Effects of customer
forgiveness on brand betrayal and brand hate in restaurant service failures: does apology letter matter?.
Journal of Hospitality Marketing & Management, 31(6), 662–687.

44
Ray, P. P. (2023). ChatGPT: A comprehensive review on background, applications, key challenges, bias, ethics,
limitations and future scope. Internet of Things and Cyber-Physical Systems, 3, 121-154.
Rita, P., Ramos, R., Borges-Tiago, M. T., & Rodrigues, D. (2022). Impact of the rating system on sentiment and
tone of voice: A Booking. com and TripAdvisor comparison study. International Journal of Hospitality
Management, 104, 103245.
Rose, M., & Blodgett, J. G. (2016). Should hotels respond to negative online reviews? Cornell Hospitality
Quarterly, 57(4), 396–410.
Rudolph, J., Tan, S., & Tan, S. (2023). ChatGPT: Bullshit spewer or the end of traditional assessments in higher
education? Journal of Applied Learning and Teaching, 6(1), 1-22.
Seo, S., & Jang, S. S. (2021). A negative or positive signal? The impact of food recalls on negative word-of-mouth
(N-WOM). Journal of Hospitality and Tourism Management, 47, 150–158.
Sharma, M., Antony, R., Sehrawat, R., Cruz, A. C., & Daim, T. U. (2022). Exploring post-adoption behaviors of e-
service users: Evidence from the hospitality sector/online travel services. Technology in Society, 68, 101781.
Shin, H. (2022). A critical review of robot research and future research opportunities: adopting a service
ecosystem perspective. International Journal of Contemporary Hospitality Management, 34(6), 2337-2358.
Smith, A.K., Bolton, R.N., Wagner, J. (1999). A model of customer satisfaction with service encounters involving
failure and recovery. Journal of Marketing Research, 36, 356–372.
Song, M., Du, J., Xing, X., & Mou, J. (2022). Should the chatbot “save itself” or “be helped by others”? The
influence of service recovery types on consumer perceptions of recovery satisfaction. Electronic Commerce
Research and Applications, 55, 101199.
Sony, M., Antony, J., Mc Dermott, O., & Garza-Reyes, J. A. (2021). An empirical examination of benefits,
challenges, and critical success factors of industry 4.0 in manufacturing and service sector. Technology in
Society, 67, 101754.
Sparks, B. A., So, K. K. F., & Bradley, G. L. (2016). Responding to negative online reviews: The effects of hotel
responses on customer inferences of trust and concern. Tourism Management, 53, 74–85.
Stamolampros, P., Korfiatis, N., Chalvatzis, K., & Buhalis, D. (2019). Job satisfaction and employee turnover
determinants in high contact services: Insights from Employees’ Online reviews. Tourism Management, 75,
130-147.
Striepe, M. (2021). Combining concept mapping with semi-structured interviews: adding another dimension to the
research process. International Journal of Research & Method in Education, 44(5), 519-532.
Swanson, S. R., & Hsu, M. K. (2009). Critical incidents in tourism: Failure, recovery, customer switching, and
word of mouth behaviours. Journal of Travel & Tourism Marketing, 26(2), 180– 194.
doi:10.1080/10548400902864800
TARP. (2007). Consumer complaint handling in America: An update study. Washington, DC: White House Office
of Consumer Affairs.
Tax, S.S., Brown, S.W., & Chandrashekaran, M. (1998). Customer evaluations of service complaint experiences:
implications for relationship marketing. Journal of Marketing 62, 60–76.
Tekic, Z., & Füller, J. (2023). Managing innovation in the era of AI. Technology in Society, 73, 102254.
Thakran, K., & Verma, R. (2013). The emergence of hybrid online distribution channels in travel, tourism and
hospitality. Cornell Hospitality Quarterly, 54, 240-247.
Tosun, P., Sezgin, S., & Uray, N. (2022). Consumer complaining behavior in hospitality management. Journal of
Hospitality Marketing & Management, 31(2), 247–264.
Treviño, T., & Morton, F. (2017). Online Service Recovery: Exploring the Effects of Justice Theory on Managerial
Responses to Negative Online Reviews. Multidisciplinary Business Review, 10(1), 62-69.
TripAdvisor (2016). TripAdvisor network effect and the benefits of total engagement. Retrieved from:
https://www.tripadvisor.com/TripAdvisorInsights/w828#%20sthash.31SwlWf1.vHe8AMv0.dpuf.
Tripadvisor (2023). Travelers Push Tripadvisor Past 1 Billion Reviews & Opinions! Retrieved from:
https://ir.tripadvisor.com/news-releases/news-release-details/travelers-push-tripadvisor-past-1-billion-
reviews-opinions
Tripp, T. M., & Grégoire, Y. (2011). When unhappy customers strike back on the Internet. MIT Sloan Management
Review, 52, 37–44.
Van Vaerenbergh, Y., Varga, D., De Keyser, A., & Orsingher, C. (2019). The service recovery journey:
Conceptualization, integration, and directions for future research. Journal of Service Research, 22(2), 103-
119.

45
Villi, B., & Koc, E. (2018). Employee attractiveness and customers’ service failure perceptions. Journal of
Hospitality Marketing & Management, 27(1), 41-60.
Wang, C. Y., Mattila, A. S., & Bartlett, A. (2009). An examination of explanation typology on perceived
informational fairness in the context of air travel. Journal of Travel & Tourism Marketing, 26(8), 795-805.
Wang, K. Y., Hsu, L. C., & Chih, W. H. (2014). Retaining customers after service failure recoveries: A contingency
model. Managing Service Quality, 24(4), 318–388.
Wang, Y., & Chaudhry, A. (2018). When and how managers' responses to online reviews affect subsequent
reviews. Journal of Marketing Research, 55(2), 163-177.
Weber, K. (2009). Service failure and recovery in an all-suite hotel/serviced apartment context: A case study.
Journal of Travel & Tourism Marketing, 26(2), 195–199.
Wei, C., Liu, M. W., & Keh, H. T. (2020). The road to consumer forgiveness is paved with money or apology? The
roles of empathy and power in service recovery. Journal of Business Research, 118, 321-334.
Wei, W., Miao, L., & Huang, Z. J. (2013). Customer engagement behaviors and hotel responses. International
Journal of Hospitality Management, 33, 316-330.
Weitzl, W., & Hutzinger, C. (2017). The effects of marketer-and advocate-initiated online service recovery
responses on silent bystanders. Journal of Business Research, 80, 164-175.
Xie, C., Zhang, J., Chen, Y., & Morrison, A. M. (2022). Hotel employee perceived crisis shocks: Conceptual and
scale development. Journal of Hospitality and Tourism Management, 51, 361–374.
Xie, K. L., Zhang, Z., & Zhang, Z. (2014). The business value of online consumer reviews and management
response to hotel performance. International Journal of Hospitality Management, 43, 1–12.
Xie, K., Kwok, L., & Wang, W. (2017). Monetizing managerial responses on TripAdvisor: Performance
implications across hotel classes. Cornell Hospitality Quarterly, 58(3), 240-252.
Xu, X. (2020). Examining consumer emotion and behavior in online reviews of hotels when expecting managerial
response. International Journal of Hospitality Management, 89, 102559.
Xu, Y., Zhang, Z., Law, R., & Zhang, Z. (2020). Effects of online reviews and managerial responses from a review
manipulation perspective. Current Issues in Tourism, 23(17), 2207-2222.
Yavas, U., Karatepe, O. M., Babakus, E., & Avci, T. (2004). Customer complaints and organizational responses:
A study of hotel guests in Northern Cyprus. Journal of Hospitality & Leisure Marketing, 11(2-3), 31-46.
Yazici-Ayyildiz, A.., Baykal, M., & Koc, E. (2022). Attitudes of hotel customers towards the use of service robots in
hospitality service encounters. Technology in Society, 70, 101995.
Ye, Q., Gu, B., Chen, W. and Law, R. (2008). Measuring the value of managerial responses to online Reviews - A
natural experiment of two online travel agencies. in Proceedings of the International Conference on
Information Systems, Paris, France, pp. 1-8.
Zhang, C., Park, J., Bonn, M. A., & Cho, M. (2021). Understanding customer responses to service failures during
the COVID-19 pandemic for sustained restaurant businesses: Focusing on guanxi. Sustainability, 13(6),
3581.
Zhang, Z., Li, H., Meng, F. and Li, Y. (2019). The effect of management response similarity on online hotel
booking. International Journal of Contemporary Hospitality Management, (31)7, 2739-2758.
Zhao, Y., Wen, L., Feng, X., Li, R., & Lin, X. (2020). How managerial responses to online reviews affect customer
satisfaction: An empirical study based on additional reviews. Journal of Retailing and Consumer
Services, 57, 102205.

46