The effect of a hotel's star-rating-based expectations of safety from the pandemic on during-stay experiences

Abstract

As a result of the COVID-19 pandemic, safety is one of the top priorities for travellers when choosing a hotel. This work examines the effect of customers’ pre-stay expectations of a hotel about its safety-focused services, shaped through its official star-rating, on the during-stay confirmation of those expectations, satisfaction, and revisit intentions. A cross-sectional research design is used spanning temporally from the pre-stay to the during-stay phases. The pre-stay phase was the peak COVID-19 period in India (June–July 2021) to stimulate the safety concerns in the travellers planning their travel, while the during-stay phase was when the planned travel was undertaken with the traveller staying at the planned hotel (October 2021–January 2022). Data were collected from 452 customers and the results supported the proposed model. Further, the star-rating, as a signal for safety-focused services, was found to have a serial effect on revisit intentions, through the pre-stay expectations of safety services, and the during-stay confirmation of expectations and satisfaction.

1. Introduction

The COVID-19 pandemic had a devastating impact on several sectors, including the hospitality and tourism industry (Mehta et al., 2021). More than 60% of hotels globally have been facing challenges to survive since the onset of the pandemic (Schoening and Shapiro, 2020). Pre-pandemic, the global hospitality industry was forecasted to grow to USD 211.54 billion by 2026, with a growth rate of 4.6%1 ; however, the effects of the pandemic have resulted in a readjustment of these projections. For example, the STR and Tourism Economics estimate that by 2023, the demand for hotel rooms will fall from 57.4% to 51.2%, reflecting the continued anxiety of travellers even if the effects of the pandemic are significantly reduced (Airoldi, 2020).

The criticality of studying a traveller's decision-making process during a pandemic is well-acknowledged (e.g., Zenker and Kock, 2020). Yet, there are limited insights into an individual's motivation to choose a specific hospitality venue during a travel overlapping with the pandemic (Aebli et al., 2022). It is well-known that different needs drive motivations and user decisions, consistent with the importance of those needs in a context (Gnoth, 1997; Herzberg et al., 2007). Hence, it is argued that travel-related decisions have been refocused towards safety needs evoked during a global pandemic. Interestingly, most research in the context of COVID-19 discusses customer resistance to travel, with safety from infection as a primary demotivator (e.g., Chua et al., 2020). However, the bases on which individuals who undertake travel during a pandemic, despite all risks, choose their stay venue, and experience its services remains under-researched (Aebli et al., 2022).

Studies published during the pandemic suggest that projecting an image of safety by hotels is critical in uncertain times (e.g., Hoque et al., 2020; Kim et al., 2021; Rivera, 2020). Indeed, guests are concerned about their safety needs and will likely remain so long after the pandemic (Agag et al., 2022; Li et al., 2020; Villa et al., 2020). This implies that a property's image of effective COVID-19/pandemic mitigation through hygiene-focused services (e.g., sanitisation of the rooms and hotel amenities), medical services (e.g., doctor-on-call, emergency hospitalization), and deployment of modern technologies (e.g., contactless check-in) is helpful for travellers to plan their stay at the hotel. Further, the corroboration of expectations during the stay has important implications for the hotel-customer relationship, including the customer's intention to revisit the hotel (Atadil and Lu, 2021; Jiang and Wen, 2020; Rivera, 2020; Vij et al., 2021).

The expectations of potential customers concerning the service quality of a hotel are often based on its official star-rating in which hotels are given a rating from 1 to 5 (with 1 being the lowest and 5 the highest) (Cser and Ohuchi, 2008; Rhee and Yang, 2015). In recent pandemic-focused research, Nunkoo et al. (2020) argue for the role of hotel star-rating in shaping customers’ expectations about the safety of a hotel as well as their stay experiences. However, the study data was captured at one point in time. Not only are expectations and confirmation two different theoretical entities, but there may also be a temporal gap when the perceptions about the safety of a hotel are formed and when the customer experiences those attributes during the stay (e.g., Gupta et al., 2020; Venkatesh et al., 2011). Thus, a research question that has not yet been addressed in the extant literature is whether hotel star-rating, as a signal of safety-related services, create expectations of safety at the hotel, and how such expectations, if met during actual stay experiences, affect the customer–hotel relationship.

Based on principles of the signalling theory and the expectation–confirmation model (ECM), as well as the tenets of Herzberg's two-factor motivation theory to underpin the need for safety during travel, this study explores the influence of pre-stay expectations of safety based on hotel star-ratings on the during-stay confirmation, satisfaction and revisit intention. The model is evaluated using covariance-based structural equation modelling (CB-SEM) with data collected from 452 respondents in India. Theoretically, the study offers a unique integration of the signalling theory and the ECM to propose a temporal model spanning from the pre-stay phase to the during-stay phase of a hotel stay. For practitioners, this study examines the customer journey in hospitality service by exploring how hotel star-rating–based safety perceptions can serve as an instrument for maintaining customer loyalty.

2. Literature review

2.1. Hotel star-rating and service quality

A hotel star-rating system is provided to each hotel's individual property by the competent authority based on the types of services offered and the overall service quality, with higher ratings reflecting higher service standards (Dioko et al., 2013; Nunkoo et al., 2020). Some hospitality brands consciously secure different ratings for their properties across geographies, sometimes under a different brand name, to target specific customer segments (Claver et al., 2006). This is because star-ratings not only indicate the facilities provided at the hotel, but also the prices for the same (Nilashi et al., 2022). Potential guests, with different needs, utilise various criteria to make stay-related decisions and the official star-rating may serve as a credible and trustworthy signal of the hotel's services to make that decision easier (Masiero et al., 2015).

Some previous studies have shown that hotel star-ratings do create service expectations and drive experiences for hotel guests (Abrate et al., 2011; Kim et al., 2019). For example, Schuckert et al. (2015) found that customers of hotels with higher star-ratings were, in general, less satisfied compared to visitors of lower-rating hotels. This is because while customers consider reviews and electronic word-of-mouth when shaping their expectations of low-star-rating hotels, such expectations are very high for high-star-rating ones, irrespective of co-customer feedback. It is apparent that customers have different service quality expectations for hotels with different star-ratings, and such expectations can strongly influence satisfaction based on the services obtained during their stay (Lie et al., 2019). However, the existing literature provides conflicting evidence about the polarity of the relationship between the pre-stay expectations, driven by the hotel's star-rating, and the during-stay customer experiences. For example, Bulchand et al. (2011) report a consistent positive relationship, Qu et al. (2000) report the same but suggest that the relationship is stronger for high star-rating hotels, while Torres et al. (2014) argue for a weak relationship. Hence, the role of hotel star-ratings in creating service quality expectations with concomitant effects on service experiences requires greater clarity.

2.2. Travel anxiety during a pandemic

Multiple works concur that stressful times are generally followed by coping behaviours by individuals (e.g., Ruvio et al., 2014). In other words, following the period of restrictions when certain activities important to an individual could not be undertaken, overconsumption of those activities may be adopted as a coping behaviour. This process helps people to release the built-up stress and is referred to as compensatory consumption (Kim and Gal, 2014). The COVID-19 pandemic was one such period where people, during the peak infection months, were constrained to their homes with restricted travel. Consequently, once the effects of a specific wave of the pandemic weakened and the travel restrictions were relaxed, the latent desire to travel led to compensatory or ‘revenge’ travel, with people travelling for longer durations and spending more money (Kim et al., 2021).

However, travellers have mixed motivations during such travel (Lee et al., 2012). While one motivation is to escape the daily routine and overconsume the travel experiences, the other motivation is to stay safe and under-consume the travel occasions, since such consumption is considered a risky endeavour with implications on one's health (Kim et al., 2021). The contradictory motivations find their underpinning in Herzberg's two-factor motivation theory, with individuals facing two unique needs: hygienic needs (e.g., perceived safety) and psychological needs (e.g., achievement). The absence of the former is a demotivator/dissatisfier while the presence of the latter is a motivator/satisfier (Herzberg et al., 2007). In the context of travel safety, using this theory, it is posited that travellers seek to maximize the benefits (fruitful travel) while minimizing the costs (potential risks). Hence, increased safety and health at the destination during a pandemic may not necessarily motivate people to travel, an outcome of the socio-psychological needs; however, the absence of those safety-enabling attributes will lead to an unpleasant experience (Aebli et al., 2022).

This phenomenon leads to travel anxiety. We argue that while individuals are keen to travel, driven by psychological needs, after the reduced effects of the pandemic, the anxiety during the travel, due to safety needs, will induce them to choose hotel stays which keep them safe from the infection. This argument is supported by Wen et al. (2005) and Kim et al. (2021) who confirm that pandemics such as severe acute respiratory syndrome (SARS) and COVID-19 affect travellers’ disposition towards the hotel with enhanced attention towards the hygiene at a given location or a property. For hotels, this means that the absence of hygiene factors may be strong demotivator for travellers to stay at the specific property (Herzberg et al., 2007).

2.3. Hotel safety

Before the outbreak of COVID-19 pandemic, several studies examined the importance of safety for hotel customers (e.g., Alnawas and Hemsley-Brown, 2019; Kim et al., 2022, Kim et al., 2022; Nagaj and Žuromskaitė, 2020). Shin and Kang (2020) highlight that there has been a prominent rise in customer safety concerns since the 2008 terror attacks in Mumbai, India, when two prominent hotels were attacked.2 Besides physical safety, the concept of safety also applies to the health of an individual due to any major viral infections prevalent at the time of stay (Villa-Clarke, 2020). Following the SARS and the H1N1 influenza outbreaks, Henderson and Ng (2004) and Lee et al. (2012), respectively, assert that non-pharmaceutical intervention by hotels to mitigate infection risk is a key dimension to the service quality of a hotel.

Thus, in the current context of health safety, we argue that when customers focus on the benefits of risk-mitigating measures taken by a hotel, they prefer a hotel with a maximum of such measures in place. Thus, to reduce the inherent uncertainty about travel and health risk, hospitality brands need to focus on improving safety perceptions and mitigating customer anxiety (Durna et al., 2015; Li and Huang, 2022; Reisinger and Mavondo, 2005; Siddiqi et al., 2022). Such best practices can range from ensuring food hygiene; offering free masks and online medical consultations; monitoring customer and employee health; closing public facilities such as laundromats, gymnasiums, and bars; inviting customers to e-observe the deep cleaning process of a hotel; leveraging technologies such as self-service check-in, cleaning with ultraviolet technology-enabled equipment, cleaning robots, voice control for room service, and facial recognition for unlocking rooms; taking temperature checks as part of the check-in process; limiting elevator rides to one customer per car; and using thermal cameras for screening guests (Elshaer and Marzouk, 2022; Hao et al., 2020). These services at a hotel are embedded into its star-rating since higher ratings allow the hotels to not only deploy the best safety technologies but also charge the customers for the same. Thus, star-ratings and the implied safety features are expected to shape the perceptions of safety at the hotel.

3. Proposed model

3.1. Integrating signalling theory and ECM

Signalling theory suggests that signals of a brand/organization are the observable attributes that are/can be deployed to communicate unique values to the customers (Spence, 1973). Such signals are used by organizations for communicating the quality of their products/services (Sekar and Santhanam, 2022). Signals are used by firms because processing the information about ‘intricate’ details of a product/service is a complex task for an individual and may lead to information asymmetry (Spence, 2002). This asymmetry can be reduced by developing comprehensive trustworthy signals that provide unambiguous quality-related information to the receiver. The signalling theory has primary application in the area of brand communication, where effective messages, as signals, can enable customers to assess the brand's product/service quality (Filieri et al., 2021; Wang et al., 2021).

In the context of this study, the hotel's star-rating classification is considered to be an important signal because a higher classification indicates better service attributes (Abrate et al., 2011), high-quality hospitality (Ariffin and Maghzi, 2012), and generally, at a greater cost (Martin-Fuentes, 2016). The star classification also signals a hotel's reputation (Abrate and Viglia, 2016), the reliability and integrity of its employees, and low customer risk (San-Martín et al., 2016). Hence, the hotel's star-rating can be considered a proxy for its prestige, reputation, and price, as well as the quality of customers' service consumption experiences.

In the service consumption domain, Bhattacherjee (2001) proposed the ECM which aims to measure a customer's confirmation of their overall expectations, with subsequent influence on continuance intentions. The concept of confirmation of expectations as a mode for predicting behavioural intentions has gained prominence in the post-adoption literature (Oh et al., 2022). According to ECM, the desired levels of expectations from a service are based on the visible attributes of an organization that evoke those expectations (Kim and Lee, 2020). When the actual consumption experience meets/exceeds primary expectations, users confirm the expectations, are satisfied and continue their engagement with the service provider. Most applications of this model assume that satisfaction is the immediate cause of behavioural intentions; however, Churchill and Surprenant (1982), and later Hsu et al. (2006), argue that the perceived performance, shaped through signals, also needs to be considered as an antecedent to the expectations and satisfaction (Filieri et al., 2020; Jung et al., 2020).

Integrating the signalling theory with the ECM, this work builds the premise that the star-rating of the hotel serves as a signal to the overall service quality that ensures the safety of its customers. Further, a higher star-rating reduces the uncertainty regarding safety-oriented hygienic services and reassures the potential patrons of their protection against any infections. The ECM helps us posit that customers' satisfaction with such safety-focused services, during their stay at the hotel, is a function of their expectations from such services, formed through the official star-rating of the hotel, and confirmation of those expectations (Bhattacherjee, 200). Such expectations are formed before the customer visits the hotel, referred to as pre-stay expectations, and enable his/her evaluation of the services during the stay to confirm their predispositions and ultimately, determine satisfaction with the stay.

Based on this discussion, this work proposes that the ability of a hotel's star-rating to signal its safety, through various service features and technology, to potential visitors is referred to as star-rating-as-safety-signal (SRSS; Sekar and Santhanam, 2022). The SRSS will help customers form pre-stay safety-related dispositions for the hotel's services (Chi et al., 2022), measured in this work by the variable expectation of safety services (ESS). On arrival at the designated hotel, the visitors will validate the expectations through their during-stay experiences (Bonfanti et al., 2021), measured as confirmation of safety services (CSS). In alignment with the ECM, customer satisfaction with the stay is proposed as a during-stay variable, while the customers' intention to revisit the hotel is proposed as revisit intention (RI) and a proxy to continuance intention (Oh et al., 2022). The study's conceptual model is presented in Fig. 1 , followed by arguments for individual hypotheses.

Fig. 1.

Conceptual model.

3.2. Star-rating-as-safety-signal (pre-stay) and expectation of safety services (pre-stay)

Travellers generally select appropriate service suppliers with pre-contractual information asymmetry potentially leading to adverse outcomes for them (Belver-Delgado et al., 2020). This is especially true if the information search occurs online (San-Martín and Jimenez, 2017) and the product/service quality is not directly observable. So, businesses, in this case, hotels, deploy signals to help reduce the uncertainty in guest perceptions (Chen et al., 2010; Melo et al., 2017). Star-ratings represent one such signal. However, the relation between the absolute hotel star-rating and customer perceptions of the hotel is unclear. Rhee and Yang (2015) suggest that customers of 4-star hotels are more inclined towards value, whereas customers of hotels with lower star-ratings are concerned about room attributes. This implies that customer expectations vary with star-ratings. López and Serrano (2004), on the contrary, find that there is no significant association between the star-rating of a hotel and expectations of its services by customers. Recently, Rajaguru and Hassanli (2018) argue that hotel star-ratings are more impactful than has been explained by researchers and that, beyond co-customer reviews, star-ratings can significantly affect customer expectations before arrival at the property. Hence, based on previous research and in the current context of customer safety during the pandemic, we posit that star-ratings will be positively associated with the superior expected safety measures taken by hotels to protect customers from infection. We thus hypothesise:

H1

SRSS of a hotel has a positive influence on ESS.

3.3. Expectation of safety services (pre-stay) and confirmation of safety services/satisfaction (during-stay)

Ajzen and Fishbein (2000) argue that customer assessment leads to an immediate and inescapable formation of beliefs about an entity. Such beliefs reflect expectations and are a prime determinant of an individual's consumption-based evaluation of the attributes. Following this argument, the actual evaluation of the safety of a hotel shaped during product/service consumption is driven by the individual's expectations regarding the benefits of consumption (Kim et al., 2021). These evaluations of products/services during consumption, as well as post-consumption reflections, curate satisfaction with the services (Bolton and Drew, 1991; Olshavsky, 1985). Hence, a primary determinant of satisfaction is the belief expressions that shape primary expectations towards the service (Hitzeroth and Megerle, 2013). Expectations represent the pre-consumption cognitive state of users, while during-consumption satisfaction indicates the cognitive and emotional evaluation of the services, determining the confirmation of those expectations (Bhattacherjee, 2001). Overall, we argue that each individual carries specific expectations about safety at a hotel, which enables them to shape their perceptions of safety, which are confirmed while they are staying at the hotel, as well as determine their satisfaction with the stay (Slattery et al., 2012). Hence, we hypothesise:

H2a

ESS has a positive influence on CSS.

H2b

ESS has a positive influence on satisfaction.

3.4. Star-rating-as-safety-signal (pre-stay) and satisfaction (during-stay)

Existing frameworks fail to differentiate between the expected and the actual experiences of hotel customers (Alcántara-Alcover et al., 2013). Unlike safety expectations, which in this case is the expected experience with the safety attributes of a hotel, satisfaction is the extent to which a client feels that the facilities, amenities, staff behaviour, and deployed technologies offer real-time protection to customers during their hotel stay (Nilashi et al., 2022). Kim et al. (2016) and Gupta et al. (2019) also explain that hotel attributes such as hygiene, location, well-trained staff, hotel design, and additional amenities are key enablers for the satisfaction of customers staying at a property.

However, the star-rating of a hotel may shape the evaluation of these experiences during a stay (Kim et al., 2016). For example, during a stay, a guest may feel that the hotel attributes and/or services are not up to standard for a 4- or 5-star hotel; similarly, a guest may judge the quality of a 1- or 2-star hotel to be higher than expected. In the domain of hotel safety, there is limited literature examining the relationship between hotel star-rating–based safety perceptions and evaluation of safety during the stay. We argue that perceptions of safety, which have been heightened due to the pandemic, may be driven both by the star-rating of a hotel and by the resultant practices at the hotel. Together, these drive guests’ satisfaction with the services ensuring customer safety, and we thus hypothesise:

H3

SRSS of a hotel has a positive influence on CSS.

3.5. Confirmation of safety service (during-stay) and satisfaction (during-stay)

For hotels to survive in the competitive tourism market in the current climate, it is essential for them to both create a safer environment and develop their image by fulfilling the customers' safety needs (Pal et al., 2019). However, little research has been conducted on what shapes customer satisfaction towards the safety facilities of a hotel (e.g., Tasci and Sönmez, 2019). A traveller's satisfaction with the stay is a mental reinforcement of the traveller's knowledge of safety, enabled through the evaluation of the safety measures to confirm their expectations of the same (Chew and Jahari, 2014). In the wake of COVID-19, travelling to new places is subject to unpredictable health-related challenges. Thus, travellers carefully evaluate safety measures at hotels in line with their expectations, and this, in turn, shapes their satisfaction towards these properties (Godovykh et al., 2021). Hence, we hypothesise:

H4

CSS has a positive influence on satisfaction.

3.6. Satisfaction (during-stay) and revisit intention

Various studies have measured the impact of customer satisfaction on behavioural intentions, such as repeating the visit to the destination or the leisure activities therein (Lam and Hsu, 2006). This is a post hoc assessment of products/services and is a cognitive projection of the outcomes of performing the same behaviour in the future (Ajzen, 1991). In other words, when the experiential outcomes are positive, individuals are more likely to hold a favourable disposition towards the service provider, which motivates them to repeat the behaviour (Kwon and Ahn, 2020). Negative experiences, on the other hand, lead to adverse beliefs, which cause individuals to avoid similar behaviours in the future (Jalilvand et al., 2012). We expect that, where careful evaluations of hotel safety services by guests lead to satisfactory experiences, these should encourage them to visit the same hotel again (or, in the case of a multi-property brand, to visit another property by the same hotel brand). Hence, we hypothesise:

H5

Satisfaction (during-stay) has a positive influence on RI.

3.7. Star-rating-as-safety-signal and revisit intention: serial mediation

The mediating role of ECM constructs, like confirmation and satisfaction, in determining continuance intention is evident in prior research. Oh (1999) argues that confirmation of expectations mediates the path between the pre-consumption perceived service quality and the satisfaction derived from the consumption. Similarly, satisfaction has also been discussed as a mediator to confirmation and behaviour. For example, Westbrook (1987) discusses satisfaction as a common link between prior product expectations, post-consumption cognitive evaluation (confirmation), and repurchase intention. Similarly, in the context of service recovery, Wirtz and Mattila (2004) and Zhu et al. (2021) suggest that post-consumption satisfaction serially mediates the relationship between the perceived and the validated service attributes and the subsequent behaviours. Based on these arguments, we argue that travellers try to make predictions about the service quality related to safety at a hotel based on signals like the hotel's star-rating. After experiencing the service at the hotel, such customers confirm those expectations, generate customer satisfaction with the stay, and develop positive intent to revisit the hotel. Thus, we propose:

H6

ESS, CSS, and satisfaction serially mediate the relation between the hotel's SRSS and RI.

4. Methodology

4.1. Construct measurement

To operationalise the constructs in this study, the questionnaire items were generated by aggregating items from different sources in the existing literature. The aggregated items were shortlisted and refined based on the discussion with three subject-matter experts: two academicians with significant publications in hospitality literature and one practitioner who was a top manager at a prominent hospitality chain. To measure the items of SRSS, we chose items from Zemke et al. (2015) and Rajaguru and Hassanli (2018). The items to measure ESS were adapted from Loizos and Lycourgos (2005), while items for measuring CSS were taken from Bigné et al. (2005). Finally, items to measure satisfaction and RI were adapted from Lam and Hsu (2006). This process led to a total of 31 items. Once the items were aggregated and developed, the context of the study was provided to the three experts, and they were asked to review the items related to each construct based on its operational definition. The experts were asked to select the relevant items for each construct, at first individually and then as a group to arrive at a consensus. This process led to a reduced list of 24 items. The panel also helped the authors to phrase the items more clearly.

All items were measured on a 7-point Likert scale (1: strongly disagree to 7: strongly agree) (Dawes, 2008). Finally, before the main data collection, a pilot test of the questionnaire was carried out with 25 respondents who had travelled during the COVID-19 period in the months of December (2020) and January (2021) and stayed in a hotel, to check for the clarity of the questionnaire. The pilot respondents gave some suggestions for three items of the questionnaire (SRSS3, SAT4, and ESS4) which were suitably modified. The final list of items is presented in Table 2.

Table 2.

Psychometric properties (Dataset 1).

Scale items/(Code)	Mean (Std Dev)	Factor loadings	Reliability		Convergent Validity	Discriminant Validity
			α	CR	AVE	MSC
Star-rating as safety signal (pre-stay) (SRSS)			.838	.859	.549	.534
A higher star-rating reflects the hotel's caring nature towards guests to keep them safe (SRSS1)	5.67 (1.22)	.707
A higher star-rating implies error-free services to keep the guests safe (SRSS2)	5.77 (1.08)	.693
A higher star-rating entails improving and improvising services to provide a safe and satisfying service to guests (SRSS3)	5.81 (1.18)	.746
A hotel with high star-rating upgrades the resources to provide enhanced medical facilities to the guests (SRSS4)	5.49 (1.21)	.791
A hotel with a high star-rating provides high-quality location-based services, like safe pick/drop, hospitalization, and safe regional travel (SRSS5)	5.23 (1.19)	.763
Expectation of safety services (pre-stay) (ESS)			.883	.876	.585	.512 .
The hotel is expected to follow all hygiene protocols (ESS1)	5.11 (1.05)	.796
I believe the hotel would be ready for any medical emergencies of guest (ESS2)	5.59 (1.24)	.768
I believe the hotel would provide more personal protective equipment which is easily accessible (ESS3)	5.71 (1.18)	.769
I believe the hotels would provide self-service technology to offer services in a contactless way (e.g., for check-in and out, kiosks, in-room services) (ESS4)	5.86 (1.19)	.721
I expect the hotel to be certified for preventing and controlling infectious diseases. (ESS5)	5.68 (1.11)	.768
Confirmation of safety services (during-stay) (CSS)			.791	.788	.554	.481
The hotel provided safety services which were much better than what I expected (CSS1)	5.48 (1.12)	.777
My stay experience at the hotel was much better than what I expected (CSS2)	5.45 (1.11)	.712
Overall the safety services of the hotel mostly met my expectations (CSS3)	5.53 (1.14)	.743
Satisfaction (during-stay) (SAT)			.884	.843	.519	.498
The hotel provided satisfactory service for the guest rooms for new arrivals (SAT1)	5.41 (1.15)	.702
The hotel employees' health condition was always satisfactorily monitored (SAT2)	5.47 (1.14)	.703
The personal hygiene of front-line employees was satisfactory (SAT3)	5.65 (1.19)	.724
The hotel performance in providing safety-related information to guests was satisfactory (SAT4)	5.76 (1.29)	.743
The employees respiratory etiquettes were satisfactory (SAT5)	5.71 (1.23)	.728
Revisit Intention (RI)			.834	.865	.562	.534
I would stay at this hotel soon (RI1)	5.34 (1.09)	.771
I plan to stay at this hotel on regular basis. (RI2)	5.21 (1.14)	.725
I intend to book this hotel for my long term health benefits. (RI3)	5.32 (1.21)	.752
I intend to book this hotel because they are more concerned about safety. (RI4)	5.66 (1.25)	.721
I intend to stay at such hotels as I am concerned about my health. (RI5)	5.21 (1.14)	.779
CR: Composite reliability; AVE: Average variance extracted; MSC: Maximum squared correlation

4.2. Sampling and data collection

This study targeted respondents who have planned and executed their travel, domestic or international, during the COVID-19 pandemic in India. India was chosen as it was one of the countries most affected by the pandemic, with the second-largest caseload in the world at the end of 2021.3 It also has a very dynamic hospitality industry, which was severely affected by the pandemic (Gupta and Sahu, 2021). The tenet of the study is based on the fundamentals of travel anxiety, where travellers desire to travel but are also cautious about their stay safety at the hotel. While travellers around the world faced travel anxiety during the pandemic, it was stronger in India due to the high caseload with travellers overtly focused on the quality of the hotel stayed in.4 Close to 92% of Indians, as per the Economic Times survey, sought for visible cues of safety at a destination during their travel.5

Further, the questionnaire was divided into two parts: the first part covered the pre-stay phase, and the second part covered the during-stay phase. In the first section, the respondents were asked to reflect on their perceptions of the star-rating of a hotel and what it implies for safety when planning their travel. In the second section, the same respondents, after their travel, were asked to recall their actual stay experiences. Each section was completed at two different points in time: the first section before the travel and the second section after the travel. The choice of a COVID-19 intensive period, while the travel was being planned, was taken as extant research indicates that the information cues for infectious diseases, like the number of infections, recoveries, and deaths, enhance the general proclivity of the customer to seek safety (Kim and Lee, 2020; Kim et al., 2021). Based on such research, it can be argued that individuals planning travel at such times would be facing travel anxiety and increased sensitivity to personal safety, which affects the type of hotel chosen for the stay.

To diminish the probability of common method bias (CMB) in the data collected, the a-priori method of randomising the independent variables in the pre-stay questionnaire and the dependent variables in the during-visit questionnaire was used, as suggested by Chang et al. (2020). The questionnaire was distributed through a prominent global travel agency, which had agreed to support the project. The agency is the largest in the country and serves the travel needs of over 15000 unique clients every year.6 In the absence of a population repository, the database of this firm – comprising over 100000 unique customers – was considered the population of all travellers. At the time of contact with the agency (April 2021), 8236 people were in active contact with the agency and were planning their travel once the second Delta-variant-led COVID-19 wave receded in India. The second wave officially lasted from April 2021 to October 2021, with its peak in May 2021 (Yang and Shaman, 2022). Since the objective of the study was to capture the safety expectations of a hotel when the customers' sensitivity to travel safety is highest, at the authors’ request, the agency disseminated the study questionnaire through an online survey platform to these 8236 people in April of 2021 seeking their interest to participate. Of these, a positive response was secured from 1134 people by the middle of May 2021. These people agreed to participate in the study and answer the questionnaire in both phases (before the travel and following the travel).

The pre-stay data were collected in June–July of 2021, when the second COVID-19 wave was at its peak in India, while the during-stay data was collected from the same people, after they had completed their travel, in the months of October 2021 to January 2022. This was a period when the second wave had subsided, the travel restrictions were eased, and people started to travel (Yang and Shaman, 2022). All the 1134 people promptly returned the pre-stay questionnaire which was complete in all respects. However, only 610 respondents undertook the planned travel in the period mentioned above and stayed in the same hotel. These people were contacted by the agency, with the during-stay questionnaire, immediately after they returned from their travel. Of those, we obtained 452 complete and useable questionnaires across both phases.

To ensure the statistical robustness of the measurement model, one-half of the data collected from June to July 2021 (N = 226) and their corresponding responses from October 2021 to January 2022, referred to as Dataset 1, were used to evaluate the psychometric properties of the measures, while the other half of the data, referred to as Dataset 2, was used to evaluate the structural model (Bagozzi and Heatherton, 1994). The sample profile is given in Table 1 .

Table 1.

Sample profile.

Variable	Category	N	Percentage	Categorization for MGA*
Gender
	Male	238	52.7	Male
	Female	214	47.3	Female
Age
	18–25 years	78	17.2	Low Age
	26–35 years	140	30.9	Low Age
	36–50 years	178	39.4	High Age
	51+	56	12.4	High Age
Marital status
	Married	184	59.3	Married
	Single	268	40.7	Single
Education
	Under Graduate	38	8.4	Low Education
	Graduate	218	48.2	Low Education
	Post Graduate	160	35.4	High Education
	Doctorate	36	7.9	High Education
Income (Monthly)
	5K-20K	30	6.6	Low Income
	21K-40K	172	38	Low Income
	41K-70K	198	43.8	High Income
	71K-1L	38	8.4	High Income
	>1L	14	3.1	High Income
Travel Period
	Oct–Jan (2020-21)	224	49.55	Not applicable
	Jun–Sep (2021)	228	50.45	Not applicable
Type of Travel
	Domestic	318	70.35	Not applicable
	International	134	29.65	Not applicable
*Multi-group analysis

5. Data analysis and results

5.1. Control variables

Spector (2021) recommended using control variables to eliminate the effect of extraneous factors on the main observed relationships. Since customer demographics, such as age, gender, education, marital status, and income (Lu and Pas, 1999) can have possible intervening effects on the outcomes of the study (see Table 1), they were modelled as control variables.

5.2. Initial checks

Before evaluating the psychometric properties of the measures, the normality assumption of the variables was evaluated by examining skewness and kurtosis for the entire dataset. The skewness value of each variable was found to be between −2 and +2, and the kurtosis values were between −7 and +7, indicating sufficient univariate normality. To check for a lack of CMB in the data, the post-hoc Harman's one-factor test was run using the principal component analysis in SPSS24, where all items were forced to load onto only one factor. The largest factor explained only 23.56% of the variance of the dataset, implying a lack of CMB.

5.3. Psychometric properties (Dataset 1)

Next, the psychometric properties of measures were evaluated using confirmatory factor analysis (CFA) in the CB-SEM using AMOS23 with the first dataset. The reliability values of each construct, measured through Cronbach's alpha (α) and composite reliability, exceeded the cut-off level of 0.7 (Nunnally, 1978; see Table 2). This implies that all multi-item scales used in this research were internally consistent. Following a reliability check, the constructs were evaluated for empirical validity, reflected through convergent and discriminate validities (Bagozzi and Yi, 1988). To evaluate convergent validity, besides factor loadings (>0.70), the criterion for average variance extracted (AVE; >0.50) was deployed (Hair et al., 1998) and was found to be in order.

To evaluate the discriminate validity, the Fornell and Larcker (1981) criterion was used, whereby the AVE of each construct was compared with the maximum squared correlation (MSC) of that construct with the other constructs in the model. In all cases, the AVE values were found to be larger than the corresponding MSC, implying discriminant validity. Table 2 provides the results for all items, descriptive statistics, as well as the results of the convergent and discriminant validity analyses. The overall model fit was evaluated by examining the values of chi-square/degrees of freedom (χ²/df), goodness of fit index (GFI), comparative fit index (CFI), normed fit index (NFI), Tucker–Lewis index (TLI), and root mean square error of approximation (RMSEA), and was found to be satisfactory with χ²/df = 3.84, GFI = 0.91, CFI = 0.91, NFI = 0.91, TLI = 0.91, and RMSEA = 0.06.

5.4. Hypothesis testing (Dataset 2)

After the measurement properties of the constructs, through CFA, were re-evaluated with the second dataset (N = 228) and were found to be satisfactory, the structural model was evaluated. The fit indices of this model were appropriate with χ²/df = 3.91, GFI = 0.90, CFI = 0.91, NFI = 0.91, TLI = 0.91, and RMSEA = 0.06,. The path from pre-stay SRSS to ESS was found to be significant (β = 0.738, p < .05); hence, H1 is supported. This means that as customers plan their trip, they expect a higher hotel star-rating to indicate enhanced safety measures. Next, pre-stay ESS was found to significantly influence during-stay CSS, and satisfaction (β = .588, p < .05; β = 0.478, p < .05). Hence, hypotheses H2a and H2b are supported. Further, the path from pre-stay SRSS and satisfaction was also significant (β = 0.801, p < .05). Hence, hypothesis H3 is also supported. It implies that higher pre-stay hotel ratings create a positive customer satisfaction towards the hotel amenities, once customers check in at the property. Also, during the stay, CSS, a validation of pre-stay expectations, is found to positively influence satisfaction (β = 0.491, p < .05). Thus, hypothesis H4 is supported. Finally, as theorised, such satisfaction has a positive influence on the customers' RI (β = 0.767, p < .05), which supports hypothesis H5. For examining the serial mediation hypothesis (H6), Model 6 in Haye's PROCESS macro in SPSS24 was deployed. The indirect effect of hotel SRSS on RI, through ESS, CSS, and satisfaction, was significant (β = 0.121, p < .05); hence, the mediation hypothesis H6 is also supported.

5.5. Multigroup analysis (control variables; Dataset 2)

Since, modelling categorical variables (non-latent) in CBSEM is an extant challenge (Kupek, 2006), to evaluate the effect of control variables on the model, the multi-group analysis (MGA) protocol in AMOS23 was deployed. This is because all the control variables (traveller demographics) are categorical in nature. Each of the control variables was made binary and restricted to two categories with approximately similar respondent numbers for easy execution in the MGA, as shown in Table 1. The MGA in AMOS23 tests the difference in specific paths across the model signified by the critical ratio (CR), which is a t-value. A CR value above 1.96 suggests the paths are different across the two groups. It was found that for all the control variables, the paths in the model were not significantly different and the CR was below 1.96 for all paths (see Table 3 ). Hence, the control variables were found to have no significant effects on the model.

Table 3.

Control variables.

Main Hypothesis		Path Value	PD Gender (M-F)	PD Age (HA-LA)	PD Education (HE-LE)	PD Income (HI-LI)	PD Marital Status (MA-SI)
H1	SRSS → ESS	.738*	.022	−.013	.013	.034	.033
H2a	ESS → CSS	.588*	.012	.035	.005	.037	−.018
H2b	ESS → satisfaction	.478*	−.034	−.022	−.027	−.024	−.031
H3	SRSS → satisfaction	.801*	.031	.014	.004	.029	.012
H4	CSS → satisfaction	.491*	.042	−.014	−.018	.019	.011
H5	Satisfaction → RI	.767*	.005	−.032	.026	−.016	.015

*significant at 95% level of significance; PD: Path difference for the two categories of a variable.

M: Male, F: Female, HA: High Age, LA: Low Age, HE: High Education, LE: Low Education, HI: High Income, LI: Low Income, MA: Married, SI: Single.

6. Discussion

This study comprises a temporal examination of the effect of pre-consumption (here pre-stay) variables such as the expectation of safety service (ESS), a hygienic need for a traveller, shaped through the star-rating of a hotel as a signal of safety (SRSS), on during-consumption (here during-stay) variables such as confirmation of safety services (CSS), satisfaction, and revisit intention (RI). The context of the study was travellers in India undertaking travel during the COVID-19 period in India, more specifically just after the deadly second Delta wave (second half of 2021). The research design was cross-sectional with data for the study collected at two points in time, pre-stay and during-stay, with the respondents staying at the same hotel they planned during the pre-stay phase.

Our results reveal that customers who had positive perceptions of a hotel based on its star-rating consider it to be safer to stay during the COVID-19 period. This implies that guests build a mental representation of hotels based on star-ratings, and that these ratings influence service expectations regarding the measures taken by the hotel to ensure customer safety (Rajaguru and Hassanli, 2018). Thus, star-ratings serve as a credible classification model that reflects the expected quality standards in the hospitality industry (Dioko et al., 2013; Serrano et al., 2014).

A primary finding of this study pertains to the effect of ESS (pre-stay) on CSS and satisfaction towards the hotel. The perception towards safety is a stable mental inclination embedded in the traveller's knowledge about safety practices to be followed within a servicescape (Tasci and Sönmez, 2019). Our findings show that as expectations, shaped through hotel ratings, increase, they are not only used as a rubric to evaluate against the existing safety measures once the customer checks in, but such expectations also shape the satisfaction towards the hotel with regards to its safety provisions (Hitzeroth and Megerle, 2013). Alternately, it can be argued that the expectations themselves are manifestations of prior experiences, from the same or different service providers, and serve as a mechanism to reinforce or weaken customer satisfaction with a specific experience (Slattery et al., 2012).

This study also confirms that the perceived star-rating of a hotel has a significant positive relationship with CSS. This implies that after customers arrive at the hotel, they constantly evaluate the hotel's protocols designed to provide a safer environment. Consequently, customers who are concerned about their safety prefer hotels that mitigate the risk of infection. CSS, in alignment with the star-rating of a hotel, reinforces their belief in such ratings. This also means that risk-averse customers are willing to pay a higher price for a higher-star hotel, with the hope that they will receive superior safety-oriented services through corresponding measures for infection mitigation.

Next, this study finds that satisfaction with the stay has a significant positive relationship with RI. This indicates that positively constructed satisfaction leads to strong similar future behaviours, while lack of it, formed due to negative experiences, reduces the propensity of the customer to reengage (Jalilvand et al., 2012). It also means that if the hotel administration follows all necessary safety processes to meet customer safety expectations, customers will form positive dispositions towards the hotel, which will in turn foster RI. Additionally, the serial effect of pre-stay perceptions of a hotel's safety, based on its star-rating, on the during-stay experiences, satisfaction, and RI, supports similar findings in the recent literature arguing for this unique pathway connecting a hotel's star-rating as a source of customer loyalty during a pandemic (Zhu et al., 2021). Finally, this work also establishes the universal applicability of this model irrespective of the traveller characteristics, like gender, age, marital status, education, and income, as none of these control variables was found to have a significant moderating effect on the model.

7. Theoretical contributions

This research makes several theoretical contributions. Recent research (e.g., Atadil and Lu, 2021) focuses on customer perceptions towards safe stays in hotels during the COVID-19 pandemic. These works have explored the dimensions underlying a safe hotel image – for example, perceptions of hygiene control, medical preparedness, use of self-service technology, and privacy – as well as the implications thereof for the visit intention of the customers. However, none of these works has moved beyond pre-travel perceptions to investigate during-travel evaluations, satisfaction and the implications of these evaluations for customer revisit intentions. Underpinned by Herzberg's dual motivation perspective, and unique integration of the signalling theory and the ECM, this study proposes that the official star-rating of a hotel serves as a signal that shapes the pre-stay ESS, and that this temporally affects the during-stay CSS, satisfaction, and RI. The finding confirms that unless the hotel's safety attributes, reflected by the star-rating, are adequate and fulfil the hygienic need of safety during a pandemic, customers will not be motivated to stay at the property (Herzberg et al., 2007). The temporal format of the research design, with data collected from the same user at two points in time, contributes to the emerging literature in the domain of hotel safety (e.g., Jung et al., 2020; Filieri et al., 2020, 2021).

There is little literature on the signalling impact of the star-rating of a hotel on its safety services (Sekar and Santhanam, 2022). Ariffin and Maghzi (2012) and Setiawan et al. (2019) argue that a higher star-rating leads to higher customer expectations towards service at the hotel facility. The current research implies that the hotel star-ratings are used to first form expectations and then to judge the service quality and that these expectations and judgements are important for satisfying the psychological and physiological needs of customers (Cser and Ohuchi, 2008; Huang et al., 2018). Overall, the star-rating of a hotel appears to be a source of customer expectations that impacts perceptions of the actual service provided by the hotel (Huang et al., 2018; Israeli, 2002; Serrano et al., 2014). The present work thus establishes the importance of the official star-rating of a hotel as a signal of overall hotel service quality, and a key driver of customer satisfaction (during stay; Malik et al., 2020; Sozen and O’Neill, 2020; Tajeddini et al., 2021).

Previous studies have indicated that customers dislike the risk implicit in hotel selection and feel the opacity that comes from service providers selling poor-quality services even under a high star-rating. Works by Wu and Cheng (2022) and Serrano et al. (2014) have argued that such hospitality classifications, like official star-ratings, third-party ratings (e.g., Trivago), or tier systems (e.g., Airbnb), tend to create information asymmetry, as such facilities over-promise and under-deliver. Our research establishes that beyond the absolute star-rating, it is the perception of services behind a specific rating that motivates customers to choose a property, and that reinforcement of expectations through actual experiences generates positive guest satisfaction as well as re-engagement intentions. This is especially true if the customer need in context is safety. Limited previous attention to this domain is understandable, as, besides the few studies looking at man-made disasters and terrorist attacks (e.g., Zenker and Kock, 2020), the safety of travel was not a priority research focus before the COVID-19 pandemic. Hence, this work, as a way forward, argues for more research on leveraging the visible traits of a hotel, including its star-rating, to create and deliver on the perception of having adequate safety mechanisms to limit the spread of disease during hotel stays (Contreras and Mep, 2020).

8. Practical implications

Based on the findings presented here, the study offers several practical contributions. Given the importance of hotel star-ratings in setting safety expectations, perceived experiences, and revisit behaviour, hotel (chain) management could deploy niche marketing campaigns conveying the message about their safety services. Notices indicating that star-ratings are earned through exemplary safety services, or that the safety record of the hotel, in terms of infections, is reflected in its star-rating, can be disseminated through traditional, electronic, and social media. The message can also be reinforced through endorsement by celebrities and reputed doctors, or testimonials by previous customers who emphasise the safety aspects of the hotel (Sigala, 2020).

Additionally, the hotel star-ratings have played an important role in attracting customers during or after the COVID-19 pandemic. The ratings reflect the importance of safety expectations and satisfaction of stay experiences, evaluated through safety-oriented amenities of the hotel (Quintal et al., 2010). Such evaluations and outcomes are universal irrespective of the traveller's demographic profile. Hence, hospitality firms must focus on raising their service standards, especially those related to safety, to upgrade their star-ratings periodically. Since star-ratings are also determined by the overall service amenities at the hotel, government bodies may also consider introducing a separate star-rating or certification based solely on the safety services of the hotel. This star-rating should be based on trustworthy information about the safety protocols followed by the hotel, such as its sanitisation processes, levels of vaccination among staff members, contactless check-in and check-out options, limitations to the number of customers in the hotel during a pandemic wave, the provision of separate services for vaccinated and non-vaccinated customers, assurances that vaccinated staff follow all safety protocols, and the presence of other technology-enabled services at the hotel. Such services can also help the hotel achieve successful containment in case of an outbreak, thus ensuring that it does not turn into an infection hotspot (Huang et al., 2020; Joo and Woosnam, 2020).

Our findings depict a strong effect of safety-related expectations on the confirmation of those through actual safety-related experiences and satisfaction. Hence, hotel managers need to deliver on highlighted safety features, as such promised benefits will influence the orientation of customers towards the specific features that deliver those benefits. Further, given that safety perceptions have a positive influence on satisfaction, safety features that are highlighted to protect customers from infection act as a powerful tool to build satisfaction, which increases the likelihood that customers will revisit the hotel. In addition, promoting the safety aspects of a hotel (as reflected in its star-rating) and delivering on promises not only helps customers form positive perceptions towards the hotel but also becomes a source of customer loyalty. This is because customers attempt to converge their expectations and actual consumption experiences to shape their experiences; hence, diligent and precise delivery of safety-based promises is critical to inducing customer loyalty.

9. Limitations and future research directions

Despite its theoretical and practical contributions, this research is subject to some limitations that should be addressed in future studies. First, as the focus of this study was on the impact of pre-stay expectations on during-stay experiences, ideally respondents should have been contacted for the second time during their stay at the hotel, rather than at a time after their visit, when some early memories of the stay, if long, would have faded. However, due to the limitations of online data collection that were mandated, as COVID-19 restrictions at the hotels meant physical data collection was not possible, the second contact wave could be conducted only immediately after respondents had completed their stay. Thus, they had to remember the experiences they had while staying at the hotel. Further, the study, though temporal, is cross-sectional by principle as no repeated measures are involved. Future studies should attempt to replicate the results using a pure longitudinal research design, where the contact questionnaire includes the same variables, related to the expectations and confirmation of those, measured before and during the stay, respectively.

Second, this research focused solely on future hotel revisit intention due to concerns related to safety from infection. While the safety of customers at a hotel is important, customer experiences at a hotel are also driven by the other services offered by the hotel. Though in this study, satisfaction measured pertains to safety features only, it is possible that other hotel amenities and services, not directly related to pandemic safety, also contributed to customers’ satisfaction. Future research should therefore investigate the during-stay experiences of customers, as a combination of safety-related and general services, to identify the unique effects of each factor on overall satisfaction and, in turn, a more robust revisit intention measurement.

Finally, the study focused on only one attribute of the hotel, the star-rating, and its implications regarding customer safety perceptions. However, safety perceptions of a hotel servicescape are made up of complex combinations of not only multiple attributes and services at the hotel itself but also other factors, such as customer reviews, social media brand engagement, prior experiences, hotel location, co-customer behaviour, and a few others. For example, even the best hotels in a city facing a large-scale outbreak will be considered unsafe. Hence, future studies should include a variety of pre-stay hotel-related information, beyond star-ratings, to gain a more comprehensive understanding of safety expectations.

Declaration of competing interest

None.

Data availability

Data will be made available on request.