Effect of motivations and engagement with eWOM on hotel queries

Abstract

The study of antecedents of eWOM queries has received less attention than the effects of these queries on purchase decisions. Considering the academic interest recently aroused in engagement with eWOM, the aim is to analyse, in the hospitality context, the effects that motivations to use eWOM have on engagement to eWOM and how this engagement influences adoption of eWOM queries. Based on the double dimension of receiving-sending both for motivations and engagement, the causal relationship chain ‘motivations-engagement-eWOM adoption’ is estimated in a sample of 393 hotel guests through a structural equation model. The effects of both motivations on engagement and, in turn, engagement on adoption are confirmed. The novelty of this work lies in the study of motivations and engagement, both from the dual perspective, as antecedents of the use of eWOM searches. These results have important academic and business management implications.

Keywords:

• Motivations
• engagement
• eWOM
• adoption
• hotels

JEL CODES:

• M30; M31

Introduction

The Internet penetration rate is 64.2% worldwide, 87.7% in Europe, and 92.5% in Spain (Internet World Statistics, 2020). Online consumers like to share their shopping experiences through multiple media channels, from the most conventional to those based on web technology. In turn, consumers consider that comments between buyers are the most important source of information when making a purchase decision (Flanagin et al., 2014). The information that is shared online is generated from a behaviour known as electronic word-of-mouth (hereinafter eWOM). eWOM is the term used to refer to general behaviour related to online comments made by consumers, and can be approached from two perspectives: receiver and sender (Kanje et al., 2020). Most of the works address the sender’s approach (e.g., Fine et al., 2017; Line et al., 2020; Shen et al., 2020) and fewer contributions are focused on receiving or searches resulting from eWOM (e.g., Book & Tanford, 2019; Kim et al., 2011).eWOM behaviour is an important key in the tourism industry and, especially, in the hotel industry (Serra & Salvi, 2014). The risks inherent to travel decisions (e.g., prices, seasonality, etc.) are factors that encourage consumers to seek information before making a reservation. Most consumers trust the comments of other users and use them as a source of information (Gretzel & Yoo, 2008). There is empirical evidence on the effects that eWOM queries have on various aspects of consumer behaviour (e.g., Ladhari & Michaud, 2015), such as sales (e.g., Park & Lee, 2009), purchase decision (Parikh et al., 2014), or purchase intention (e.g., Abedi et al., 2019). However, there are fewer studies focused on analysing which variables influence this type of queries (e.g., Hatzithomas et al., 2016). Therefore, the study of the drivers that contribute to the eWOM query and that may intervene in its effectiveness on the purchase decision will be key for hotels to better adapt their strategies to the online environment.

Among these variables, engagement with eWOM is an area of ​​study that has attracted great academic interest in recent years (e.g., Gvili & Levy, 2018; King et al., 2014; Wang & Kubickova, 2017). However, the empirical evidence on the antecedents and consequences of this type of engagement is still limited (e.g., Yusuf et al., 2018). The study of motivations towards eWOM is also essential in order to understand what factors influence this behaviour and to help companies empathise with consumers in creating value (Fine et al., 2017). There is more research into the motivations from the sender’s point of view (e.g., Xu, 2018) than from the receiver’s perspective (e.g., Parikh et al., 2014) and, furthermore, there is no empirical evidence that provides a two-sided approach and addresses both the motivations to consult eWOM and the motivations to write eWOM.

In summary, the gaps that have motivated this work are the following: 1) less attention to the background of the eWOM queries than to their effects, 2) less attention to the study of eWOM from the receiver’s point of view than from the sender’s approach, 3) scant empirical evidence on engagement with eWOM, and 4) scant empirical evidence on motivations for using eWOM that simultaneously address motivations for seeking and submitting eWOM. Mainly, the deficiencies detected in the literature are the lack of research on the antecedents of eWOM queries and the scarcity of empirical evidences focusing on the double sender-receiver approach to explain how these queries are formed.

From the above arguments, this work addresses the study of the adoption of eWOM queries, understood as the influence that have on the purchase decision. With this aim, this proposal will try to complete the aspects of the literature related to the variables that affect eWOM queries, from both sender and receiver approaches. Thus, the objective pursued is to study the effect that motivations and engagement with eWOM have on the degree of adoption of eWOM queries. To do this, it is intended to analyse the chain of relationships ‘motivations towards the use of eWOM - engagement with eWOM - adoption of the eWOM query’ considering that motivations and engagement are two-dimensional variables formed by the query dimension and the send dimension. Therefore, we frame the final variable to be explained (adoption of eWOM), from the receiver’s perspective, while we find ourselves looking at the explanatory variables (motivations and engagement) from the dual perspective of the receiver and the sender.

The novelty of this work lies in the study of motivations and engagement as antecedents of the eWOM queries made by consumers to assist them in making decisions about hotels. In addition, given that the literature is still lacking in works that address the antecedents of eWOM queries from the dual receiver-sender perspective, it adds a two-sided approach by treating motivations and engagement from both viewpoints. The results of this research will provide relevant conclusions from the academic and practical points of view. At an academic level, we are contributing to research on motivations and engagement with eWOM and its effects on the purchase decision derived from eWOM queries. On a practical level, the results can help to improve the hotels’ lines of action in relation to the design of communication with guests aimed at improving the way they make recommendations and increase their effectiveness in terms of booking processes.

Theoretical review

eWOM behaviour

eWOM behaviour can be defined as any statement (positive or negative) about a product (or service, company, brand) that is posted on the Internet by consumers (current, potential, or former) (Hennig-Thurau et al., 2004). The wide variety of general online media, as well as sites specifically intended for business comments (e.g., Google Reviews, Yelp, etc.) or tourism (e.g., TripAdvisor, Booking, Trivago, etc.) permits rapid dissemination of a high volume of objective, subjective, and emotional (e.g., emoticons) information that remains accessible for a long time to a multitude of consumers (Munar & Jacobsen, 2014).

Following an overall approach, eWOM behaviour can be analysed from the perspective of both the receiver and the sender of the information (Kanje et al., 2020). The former is a pre-purchase behaviour and focuses on seeking information that helps with the decision, whereas the latter is post-purchase behaviour in which the individual makes and disseminates comments to help others, settle a score, or reduce dissonance.

The growth of eWOM behaviour has brought about a change from the classic e-commerce model to a new social commerce model (Shanmugam et al., 2016). Consumers have the power to generate content on social media by conducting evaluations, criticisms, reviews, comments, enquiries, recommendations, etc. (Hajli & Sims, 2015), so social factors are now key determinants of transactions. In addition, consumers now take on the role of ‘prosumers’ and become a fundamental part of a business’s image (Siuda & Troszynski, 2017). Given the potential influence that eWOM has on purchase decisions, interest in social commerce research continues to grow (King et al., 2014; Yusuf et al., 2018).

Engagement with eWOM

To delve into the variables that influence eWOM behaviour and its impact on consumer decisions, it is necessary to consider the participation or engagement that the individual exhibits in regard to that behaviour. Since the Internet is the virtual environment in which consumers interact and exchange information, social support becomes an important social value. There are different theories that have supported the study of eWOM communication, such as the Elaboration Likelihood Model (ELM) (Petty et al., 1981), the Theory of Reasoned Action (TRA) (Ajzen & Fishbein, 1977), and the social support model (Huang et al., 2010).

According to ELM, consumers are affected by the information they receive either via the central route (consumer’s decision based on rational criteria) or the peripheral route (less cognitive effort involved in the consumer’s decision) (Sussman & Siegal, 2003). TRA considers ‘attitude-intention-behaviour’ relationships and explains the process by which a consumer adopts information when exposed to social norms, reviews, or comments in the online medium (Fan et al., 2013). However, in our opinion, the theory of social support offers the most adequate explanation of engagement with eWOM. According to this approach, consumers look for social resources that offer both emotional and information-based support. Emotional support implies the capacity to trust another person thereby generating a feeling of being loved, accepted, or belonging to a group (Pfeil & Zaphiris, 2009). Information-based support involves providing messages in the form of ratings, recommendations, or notices to help other consumers (Liang et al., 2011). Therefore, the eWOM query behaviour that a consumer exhibits may be based on the search for social sources that help them, from an emotional and rational perspective, to make purchase decisions.

Based on this theory, in general terms, engagement refers to the emotional reactions caused by a stimulus to the individual and that allow them to form a link or relationship with said stimulus (Kapoor & Kulshrestha, 2011). In the commercial context, it is a psychological state that is derived from co-creative and interactive experiences that the individual has with a brand or a company (Brodie et al., 2011). Van Doorn et al. (2010, p. 254) defines consumer engagement as ‘a behavioural manifestation toward a brand or firm, beyond purchase, resulting from motivational drivers’. It is shared that this type of engagement is multidimensional in nature. From an integrative approach, it is considered that consumer interactions contain three dimensions: the cognitive (referring to the processing and development of the interaction), the affective (linked to emotions) and the behavioural (related to the time and effort that the consumer invests) (Hollebeek et al., 2014). Since online social media facilitates these interactions, the behavioural dimension of engagement would cover participation in eWOM-related activities. Therefore, from the concept of consumer engagement comes the concept of engagement with eWOM.

Engagement with eWOM is based on the desire to request or share eWOM information from or with other consumers, respectively (Yusuf et al., 2018). The literature on this construct is scarce and relatively new. Following the behavioural approach of engagement (Hollebeek et al., 2014), there is no firm agreement on the type of behaviours that make up engagement with eWOM. When reference is made to engagement with a particular website, this engagement includes several types of behaviours related to comments, percentage of clicks, number of users, visits, likes, etc. (Barger et al., 2016; Coelho et al., 2016). Chu and Kim (2011) differentiate between opinion seeking behaviour, opinion issuing behaviour, and the opinion transmission behaviour of others. However, the most generalised position is to characterise engagement with eWOM by two types of behaviour (e.g., Fine et al., 2017; Gvili & Levy, 2018):

• The behaviour of receiving or consulting eWOM is based on the search for information by consumers who are usually engaged with the message (Calder et al., 2009). In this case, the consumer’s engagement with eWOM is passive (Muntinga et al., 2011).

• The behaviour of sending, or disseminating eWOM, represents two actions: the action of creating and giving an opinion, more typical of consumers who are experts or well-informed about the product (Barnes & Pressey, 2012), and the action of forwarding or passing on the information of others (Sun et al., 2006). In this case, it is considered that the consumer’s engagement with eWOM is active (Muntinga et al., 2011).

Motivations towards eWOM

Motivations represent a need or desire that triggers a response in the individual (Hennig-Thurau & Walsh, 2003), and for this reason they can be key determinants both in the search for and in the dissemination of eWOM. Along these lines, studying the motivations that individuals have towards the use of eWOM is especially important for hotel managers, since it can help them to improve aspects that encourage positive comments from eWOM senders and lead eWOM receivers to consult these comments.

The literature on the motivations towards eWOM is not very abundant, especially in the context of hotels. In general, there is a lack of agreement in the identification and classification of motivations. In addition, some contributions are confusing when it comes to clearly focusing on the type of motivations, either from the perspective of the eWOM receiver or of the sender. Most research analyses the motivations for writing opinions (e.g., Kim & Jang, 2019; Munar & Jacobsen, 2014; Munzel & Kunz, 2014; Shen et al., 2016; Xu, 2018) and fewer authors evaluate the reasons behind the consultation of eWOM (e.g., Parikh et al., 2014). No contributions have been found that address the motivations from both perspectives.

Therefore, considering the dual receiver-sender perspective, the motivations for using eWOM can be divided into two types:

• Motivations for consulting eWOM. The work of Schiffman and Kanuk (1978), one of the first studies on motivations, differentiates between motives of self-involvement, of involvement with the product, and of involvement with others. Following this approach, Hennig-Thurau and Walsh (2003) propose a series of motives related to risk, purchase security, reduction of dissonance, the social component, and learning about products. To these motivations, Goldsmith and Horowitz (2006) add reasons related to the convenience and timeliness of the information found and Parikh et al. (2014) add the search for novelty and trust in the queries. More recently, Srivastava and Kalro (2018) identify reasons such as improving decision-making, reducing risks, counteracting the negativity bias, as well as social reasons, curiosity, economic incentives, and validation of beliefs. However, one of the most accepted classifications is that of Kim et al. (2011), which summarises the motivations as convenience, social support, and risk reduction.

• Motivations for sending eWOM. One of the most prominent works on this type of motivations is that of Hennig-Thurau et al. (2004), which identifies the motivations related to utility or convenience, post-purchase, personal satisfaction, and emotions. Cheung and Lee (2012) study five groups of motivations labelled as selfish, collective, altruistic, moral, and self-efficacy. The contribution by Yoo et al. (2013) distinguishes between intrinsic motivations, linked to curiosity, personal goals, or the search for emotions, and extrinsic ones, associated with social needs. Munzel and Kunz (2014) propose motives derived from positive and negative experiences, motives for social bonding and individual benefit. Fu et al. (2015) analyse the motivations of altruism, reciprocity, and selfishness. Along the same lines, Yen and Tang (2015) and Shen et al. (2016) propose the search for recognition, concern for other consumers, a desire to help the company, economic incentives, and social benefits.

Proposed model and hypothesis

As previously mentioned, academic research understands that engagement with eWOM is a multidimensional construct (Calder et al., 2009; Hollebeek et al., 2014), although it is usual to conceptualise it as two-dimensional, that is, formed by the behaviours of consulting eWOM and sending eWOM (Gvili & Levy, 2018). In addition, it has been highlighted that engagement is a more complex behavioural manifestation than purchasing, which is caused by motivations (Van Doorn et al., 2010; Verhagen et al., 2015). Therefore, consumer motivations towards the use of eWOM can make an important contribution to engagement. Following the two-dimensional approach to engagement (Gvili & Levy, 2018), we consider that the motivations towards the use of eWOM should also include this dual approach that includes the motivations to consult and to send eWOM.

With this approach, the following model is proposed that represents the chain of relationships ‘motivations towards the use of eWOM - engagement with eWOM - adoption of the eWOM query’ (Figure 1). Considering that this adoption refers to the degree to which consumers modify their behaviour using the suggestions and comments from eWOM queries (Filieri & McLeay, 2014; Rani & Shivaprasad, 2018), it is therefore intended to study the effect of the motivations on engagement and in turn the effect of this engagement on the adoption of the eWOM.

Figure 1. Proposal model. Source: own elaboration.

Effects of motivations on engagement

The literature has highlighted that engagement is generated from motivations (Barger et al., 2016; Verhagen et al., 2015). With regard to this matter, the most relevant empirical evidence is linked to the effect of motivations on consulting and sending behaviours.

Regarding the effect of motivations to send eWOM, Shen et al. (2016) confirm that the intention to carry out or write eWOM depends on the five groups of motivations mentioned above (recognition, concern, help, economic incentives and social benefits). According to Fu et al. (2015), there is a direct and positive effect between the motivations to send and the attitude towards sending eWOM. In particular, in the field of tourism, the work of Fine et al. (2017) reveals that the intrinsic and extrinsic motivations of tourists to disseminate information via eWOM have a positive relationship with this behaviour.

Regarding the effect of motivations to consult eWOM, there are some contributions related to the reasons for consulting eWOM (e.g., Kim et al., 2011; Parikh et al., 2014). However, there is very little empirical evidence that demonstrates the causal effect of these motivations on the intention or behaviour to carry out consultations via eWOM (e.g., Moliner-Velázquez et al., 2021).

The possible effect that both types of motivations can have on engagement with eWOM lies in the essence of its concept (Yusuf et al., 2018). This engagement is part of a series of consumer motivations that represent concerns towards themselves, such as the possible benefits and costs, and towards others, such as social needs (Van Doorn et al., 2010). Therefore, in accordance with the previous empirical evidence, we understand that the general motivations towards the use of eWOM, both sending and receiving, will have a significant effect on engagement with eWOM. The more motivated the consumer is towards eWOM behaviour in general, the more involvement they will want to have and the higher the degree of engagement they will feel with this type of behaviour. Considering the dual dimension of motivations and engagement, we propose (Figure 1):

H₁: Motivations to consult eWOM positively influence H_1a) engagement to consult eWOM, and H_1b) engagement to send eWOM.

H₂: Motivations to send eWOM positively influence H_2a) engagement to consult eWOM, and H_2b) engagement to send eWOM.

Effects of engagement on the adoption

It is shared in the literature that eWOM behaviour notably influences purchase intentions (Baber et al., 2016; Ladhari & Michaud, 2015; Sharifpour et al., 2016). However, to rationalise this impact, it is necessary to consider the level of engagement that the individual has with this behaviour (Yusuf et al., 2018). Analysing the desire to seek or share information with other consumers will be key to better understand its effects on consumer behaviour. In this sense, we believe that engagement with eWOM will have a significant effect on the purchase decision. This purchase decision will be conditioned by the degree of influence that the eWOM query has had. Therefore, it is assumed that engagement with eWOM will have a positive effect on the adoption of the eWOM query.

Some works can be found in the literature on the effect of engagement with certain facets of consumer behaviour. For example, according to Correia et al. (2017), brand engagement indirectly influences the behaviour to send eWOM. However, empirical evidence on the relationship between engagement with eWOM and purchase intention is very scarce. According to Yusuf et al. (2018), in the context of online purchases, there is a positive and significant relationship between engagement with eWOM and the consumer’s purchase intention. Furthermore, in the tourism field, no contributions have been found on the effects of the consumer’s engagement with eWOM.

Based on these results and the role that engagement plays in the development of a behaviour, we consider that the more consumers feel that they have engaged with eWOM behaviour, the greater their involvement and connection with this type of communication. Therefore, the greater the engagement, the greater the influence that the eWOM queries have on the purchase decision, that is, consumers are more likely to adopt the eWOM query. Continuing with the two-dimensional approach to engagement, we formulate (Figure 1):

H₃: H_3a) Engagement to consult eWOM, and H_3b) engagement to send eWOM positively influence the adoption of the eWOM query.

Methodology

Measurement instrument and field work

A quantitative empirical investigation was developed and subsequently applied to carry out research into hotel services using face-to-face surveys. The questionnaire was prepared from a set of scales carefully selected from the literature and adapted to our context, in which 7-point Likert scales were used. Regarding the motivations towards the use of eWOM, the scale of motivations to consult is made up of 3 items, adapted from the work of Kim et al. (2011) and based on Hennig-Thurau and Walsh (2003), and Goldsmith and Horowitz (2006), and the motivations to send scale is made up of 5 items adapted from Hennig-Thurau et al. (2004). To measure engagement with eWOM, 3 items from the Gvili and Levy (2018) proposal were adapted to measure engagement to consult and 2 items were adapted from the work of Kim and Cha (2002) to measure engagement to send. Finally, the adoption of the eWOM query was measured through 2 items adapted from Filieri and McLeay (2014). The survey was in both Spanish and English, to ensure that all guests were able to understand and answer the questionnaire. Following Douglas and Craig (2007) recommendations, a two-stage pilot test was carried out. Firstly, the questionnaire was given to five scholars of marketing, experts in hospitality and digital buying behaviour, to improve the sequencing and wording of the items. Then, a pre-test was done with 15 hotel guests—who match the target population criteria- to verify that it functioned correctly. Based on the feedback, question order was modified, the items were better adapted to the context, and scale sensitivity was verified.

The population under study is defined as the guests of hotels located in the geographical environment of the Valencian Community. The study included hotels in two of the main cities in the province of Valencia: Gandía and Valencia. The province of Valencia is a benchmark for Spain’s hospitality industry as it is the second largest autonomous community in terms of accommodation offer and employees (INE, 2022).

An initial list was drawn up from Spain’s official hotel guide¹ and the hotel directory of the Valencian Tourism Agency.² This list was completed with information from the SABI (Iberian Balance Analysis System) and DUNS100,000 databases. Following the previous studies, from the final list, various categories of hotels were selected: two 5-star hotels, twenty-two 4-star hotels, and eighteen 3-star hotels, since higher-class establishments are more likely to invest in technology and share information with potential guests through social media (Ruiz-Molina et al., 2011).

For the purposes of completing the questionnaire, authorisation was requested from the establishment to conduct interviews in the hotel reception area. The questionnaire was conducted by a survey company in hotel lobbies during mornings and evenings. A nonprobability convenience sampling approach was used. 1175 respondents were intercepted, obtaining 393 complete and valid questionnaires (33.45% response rate). The main characteristics of the sample are shown in Table 1.

Table 1. Sample profile.

Type of trip		Gender
Vacations	36.6%	Male	35.7%
Business	22.0%	Female	64.3%
Others	19.4%
Frequency of stay		Age
Once a year	44.1%	Mean	43.80 (± 14.8)
2-4 times a year	36.1%
>4 times a year	19.8%	Nationality
Education		Spanish	88.5%
No studies	1.8%	British	1.8%
Elementary	13.7%	French	1.5%
Secondary/college	15.3%	Italian	1.5%
Professional training	18.6%	U.S.	0.5%
3-year degree	14.8%	Other	6.2%
5-year degree	33%
Postgraduate	2.0%
Others	0.8%

Source: own elaboration.

Data analysis

The data was analysed using different statistical techniques to test the research hypotheses. Firstly, a confirmatory factor analysis (first-order measurement model) was estimated with robust maximum likelihood to validate the factor structure of constructs and the psychometric properties of measurement scales (reliability and validity). Internal consistency of the constructs was evaluated considering two indicators: composite reliability coefficient and the variance extracted for each scale. Analysis of the scales also included scale construct validity (convergent and discriminant) for the factors which make up the latent variables following the criteria of Steenkamp and Van Trijp (1991), and Fornell and Larcker (1981).

Secondly, a structural equation model (SEM) was estimated. This type of causal modelling enabled joint consideration of the measurement of the constructs and the prediction to evaluate the effects of the latent variables without contamination from measurement errors (Bentler, 2006). Robust maximum likelihood from the asymptotic variance-covariance matrix was again applied due to the lack of multivariate normal distribution of the data. All estimations were carried out with EQS 6.2.

Dimensionality, reliability, and validity of the measurement scales

A first-order measurement model was estimated to analyse the dimensionality and validity of the proposed measurement scales. This estimation reached acceptable fit indices (Table 2). The internal consistency was evaluated through Cronbach’s Alpha and composite reliability (CR), whose minimum threshold is 0.7 (Anderson & Gerbing, 1988), and the variance extracted from each of the scales (AVE), whose value must exceed 0.5 (Fornell & Larcker, 1981). All these indicators, listed in Table 2, exceed the recommended limits. In relation to the process of refining the scales, we must indicate that an item was eliminated from the motivations to write eWOM as it presented an insufficient load to its latent factor, resulting in a scale made up of four indicators.

Table 2. Reliability and validity of the measurement scales.

Scales	Items	λ Standardised (t-Stat)
Motivations for consulting eWOM α = 0.879; CR = 0.892; AVE = 0.741	I read information about hotels from other consumers’ online reviews to find solutions for my problems	0.610
	Customer reviews help me to make the right buying decisions	0.981** (12.13)
	I read online reviews about hotels to benefit from others’ experiences before I book a hotel room	0.942** (12.02)
Motivations for sending eWOM α = 0.927; CR = 0.924; AVE = 0.756	I write comments because this way I can express my joy about a good buy	0.855
	I write comments because I want to help others with my own positive experiences	0.975** (21.96)
	I write comments because I can tell others about a great experience	0.902** (16.22)
	I write comments because I want to save others from having the same negative experiences as me	0.727** (12.40)
Engagement to consult eWOM α = 0.985; CR = 0.984; AVE = 0.953	I usually read recommendations on hotels on social media	0.954
	I am pleased to read on social media about experiences other people have had with hotels	0.995** (50.49)
	I am open to receiving other people’s opinions on hotels	0.980** (43.10)
Engagement to send eWOM α = 0.974; CR = 0.966; AVE = 0.934	I will tell through the Internet good things about this hotel	0.937
	I will recommend this hotel to other people via the Internet	0.995** (17.26)
Adoption of the eWOM query α = 0.996; CR = 0.994; AVE = 0.988	I closely followed the suggestions in online reviews and went to the recommended accommodation	0.996
	The information in online reviews has motivated me to purchase the recommended accommodation	0.992** (81.68)

**: significant at 99%.

Chi² _Sat-B.(d.f.=69)=117.70 (p-value = 0.0002) -Chi² _Sat-B./d.f.=1.71-; RMSEA = 0.055; CFI = 0.994; IFI = 0.994; GFI = 0.912; AGFI = 0.864; BB-NFI = 0.991; BB-NNFI = 0.992.

Source: own elaboration.

In the next step, the validity of the scales was tested. Specifically, the following were contrasted: (1) content validity, since, as indicated in the previous section, all measurement items were adapted according to a rigorous review of the literature; (2) convergent validity, since, as shown in Table 2, the standardised factor loadings are 99% significant (Steemkamp & Van Trijp, 1991); and (3) discriminant validity, since the linear correlation between each pair of scales is less than the square root of the AVE of the scales involved as Table 3 shows (Fornell & Larcker, 1981).

Table 3. Discriminant validity of the measurement scales.

Scales	1.	2.	3.	4.	5.
1. Motivations for consulting eWOM	0.861
2. Motivations for sending eWOM	0.464	0.869
3. Engagement to consult eWOM	0.373	0.324	0.976
4. Engagement to send eWOM	0.269	0.351	0.330	0.966
5. Adoption of the eWOM query	0.415	0.285	0.407	0.319	0.993

The elements of the main diagonal represent the square root of the AVE.

Source: own elaboration.

Results and discussion

After validating the measurement scales, a causal model was estimated in order to test the hypotheses on the chain of relationships ‘motivations towards the use of eWOM - engagement with eWOM - adoption of the eWOM query’. The estimated path coefficients are provided in Table 4, showing that they are all significant, which allows us to contrast the research hypotheses in an affirmative way.

Table 4. Causal model estimation.

Causal relationships	Standardised coefficient	t-Stat	Hypothesis
Motivations for consulting eWOM → Engagement to consult eWOM	0.283**	3.12	H1a contrasted
Motivations for consulting eWOM → Engagement to send eWOM	0.147*	1.98	H1b contrasted
Motivations for sending eWOM → Engagement to consult eWOM	0.193**	2.35	H2a contrasted
Motivations for sending eWOM → Engagement to send eWOM	0.293**	3.14	H2b contrasted
Engagement to consult eWOM → Adoption of the eWOM query	0.345**	4.73	H3a contrasted
Engagement to send eWOM → Adoption of the eWOM query	0.211**	3.16	H3b contrasted

*: significant at 95%; **: at 99%.

Chi² _Sat-B.(d.f.=72)=138.26 (p-value = 0.000) -Chi² _Sat-B./d.f.=1.92-; RMSEA = 0.062; CFI = 0.993; IFI = 0.993; GFI = 0.898; AGFI = 0.850; BB-NFI = 0.989; BB-NNFI = 0.991.

Source: own elaboration.

With regard to the effect of motivations on engagement, on one hand, the motivations to consult eWOM have a significant influence on both the engagement to consult eWOM (γ = 0.283**) and to send eWOM (γ = 0.147*) and, furthermore, the motivations to send eWOM also have a significant effect on both types of engagement (γ = 0.193** and γ = 0.293**). The effects of motivations have not been widely investigated in the literature. For example, there is some empirical evidence regarding the effects of motivations to consult eWOM on the consulting intention or behaviour (e.g., Moliner-Velázquez et al., 2021) and also on the effect that motivations to send have on eWOM dissemination behaviour (e.g., Fine et al., 2017; Shen et al., 2016). Although no studies have been found that confirm the causal relationship between motivations and engagement, our results are in line with certain authors who understand that engagement is generated by motivations. (e.g., Barger et al., 2016; Verhagen et al., 2015; Yusuf et al., 2018).

With regard to the effect of engagement on the adoption of the eWOM queries, the results indicate that this adoption depends significantly on the engagement to consult eWOM (β = 0.345 **) and the engagement to send eWOM (β = 0.211**). There is some evidence in the context of services of the influence of engagement with eWOM on purchase intention (e.g., Yusuf et al., 2018), however, no previous contributions have been found that confirm the relationship between engagement and adoption. Therefore, the results obtained allow progress in the study of this relatively recent area with immense potential to explain different facets of eWOM behaviour (Gvili & Levy, 2018).

Comparing the two types of motivations, towards the consultation and towards the sending of eWOM, the results demonstrate differences in the relationships of the ‘motivations-engagement-adoption’ chain. Regarding the motivations-engagement relationship, the motivations towards consultation are the most important predecessor of engagement with eWOM consultation, meanwhile motivations towards sending are the most important predecessor of engagement with eWOM sending. In regard to the engagement-adoption relationship, engagement with eWOM consultation is the one that has the greatest effect on the adoption of eWOM queries.

In summary, all the theoretical relationships proposed have been confirmed (see Figure 1). The results also allow us to advance in the study of the adoption of eWOM queries. Firstly, both perspectives, receiver and sender, are closely connected. There is also a stronger link between the variables related to each approach, between query-related variables and between sending-related variables. From the receiver’s point of view, there is a stronger relationship between the motivations towards the consultation, the engagement with consultation and the adoption of eWOM queries. From the sender approach, there is stronger relationship between the motivations towards sending and the engagement with eWOM sending. Despite the link between both perspectives, these results suggest that there may be important differences between the eWOM query process and the eWOM sending process, which poses a challenge for further research on this two-sided approach.

Conclusions and implications

The objective of this work was to study the effect that motivations towards the use of eWOM and engagement with eWOM have on the degree of adoption of eWOM queries, that is, on the influence that these consultations have on the purchase decision. This objective has been motivated by the limitations that have been detected in the literature on the study of motivations and engagement and their relationships with eWOM.

In this way, the literature review has revealed the need for further research exploring the antecedents of eWOM queries and their effects on purchase decisions. Along these lines, we make a relevant contribution to the advancement of academic research on eWOM behaviour focused on the study of motivations and engagement and their contribution to adoption of information through eWOM. Although there is evidence of the effects that eWOM queries have on purchase intention or decision (e.g., Abedi et al., 2019; Parikh et al., 2014), none of the previous works has addressed these relationships within the same investigation nor has it focused on the study of the antecedents or causes of eWOM queries from the dual receiver-sender perspective.

This study opportunity has led us to carry out empirical research in the tourism sector and, in particular, in the context of hotels. The results have confirmed that the motivations towards the use of eWOM positively influence engagement with eWOM and this engagement also contributes to the adoption of the eWOM query. It is concluded that motivations and engagement are fundamental factors in explaining to what degree the eWOM that consumers consult influences their purchase decision. These relationships represent interesting lines of research to take a more in depth look at the conditions of purchase based on factors linked to the online environment.

On the one hand, motivations have turned out to be an important variable that can contribute to purchase intention. In this line of study, although the literature highlights that the information shared on the Internet about experiences with products and services is a powerful source of information (Flanagin et al., 2014), even surpassing the influence of other promotional actions, the eWOM query is not by itself a sufficiently effective behaviour if there are no specific motivations towards the use of eWOM (Kim et al., 2011). That is why, in the current context of electronic commerce with a marked social character thanks to eWOM (King et al., 2014; Yusuf et al., 2018), studying the motivations that consumers have both to consult eWOM and to send it will permit a better understanding the mental processes of the consumer prior to purchase.

On the other hand, consumer engagement with eWOM behaviour has emerged as another key antecedent to purchase intent. The study of engagement offers extensive research opportunities to advance the understanding of the factors that influence the purchase. According to Van Doorn et al. (2010), engagement depends on the motivations of the individual, so they are antecedents that must be addressed jointly. In addition, the empirical evidence on the contribution that engagement with eWOM has in regard to the queries and its influence on the purchase is very limited (e.g., Yusuf et al., 2018). It is therefore necessary to advance along these lines by applying motivations and engagement within the context of accommodation experiences.

From a business point of view, if in tourism there is evidence that eWOM queries influence decisions (e.g., Fakharyan et al., 2012; Jalilvand & Samiei, 2012), the results of this work can inspire managers to design actions that stimulate motivations towards the widespread use of eWOM to influence engagement and thereby contribute to other consumers’ choice of a service. In this sense, online communication strategies could be adapted and improved taking into account the different motivations that individuals may have both to consult and to write or disseminate eWOM. Regarding the motivations to search and consult, it could be effective to improve aspects such as the perception of time savings, convenience, purchase risk, or belonging to a group. To do this, it would be necessary to develop actions aimed at simplifying booking systems and facilitating access to information about other customers’ experiences. Regarding the motivations for writing and disseminating eWOM, it could also be useful to increase the perception of assisting other consumers and the usefulness of the recommendations, as well as highlighting the importance of social needs.

Future lines

To advance in this line of research, we propose to continue the study of motivations and engagement with eWOM, in regard to the two aspects of consulting and sending, and in the context of accommodation. This dual approach would allow eWOM behaviour to be addressed from the overall perspective of the complete consumer experience. The eWOM queries that consumers use to decide on a purchase generate expectations that can influence their satisfaction and, in turn, this satisfaction can also condition the eWOM comments that they post later. Therefore, jointly analysing the consultations and the dissemination of eWOM will be key to identifying differences between both behaviours and common antecedents.

At a methodological level, the analysis could be applied to other types of tourist accommodation (e.g., apartment rentals, campsites, Airbnb, or home swaps). To improve the extrapolation of results, the sample could also be extended to other geographic areas on a national level. Different measurement scales could also be used that make it possible to gather information on other significant dimensions of eWOM behaviour. For example, the dimensions of convenience, social support, and risk reduction from the classification of motivations to consult by Kim et al. (2011) could be differentiated; and the proposal by Shen et al. (2016) would help address dimensions of motivations for writing related to extroversion, altruism, economic and social benefits, and helping the company. Finally, online forwarding of eWOM (Chu & Kim, 2011; Muntinga et al., 2011; Sun et al., 2006), understood as a behavioural consequence of eWOM related to the action to transmit information to others, could also be added to the dimensions of engagement to send.

In addition, the effects that the COVID-19 pandemic is having on the profitability of the tourism sector must also be considered. 2020 has been described as the worst year in the history of tourism (UNWTO., 2021). In regard to the Spanish hotel industry in particular, revenue at the end of 2020 had fallen by 66% compared to 2019 (Statista, 2021). This health crisis has accelerated the digital transformation of companies and has caused significant changes in consumer behaviour, especially in pre-purchase processes, such as new decision criteria, greater rationality of purchases, etc. All these changes highlight the fragility of certain sectors (Mele et al., 2021), especially those related to information and communication technologies, such as the tourism sector. This new, uncertain, and highly competitive context mean that some of the motivations or reasons that lead consumers to make eWOM queries may be altered. Credibility of the eWOM comments consulted (Rahman & Mannan, 2018), and engagement with eWOM before and after purchase (Yusuf et al., 2018) are variables that are currently of particular interest in helping to deepen the study of the motivations towards the use of eWOM. Therefore, these variables could be analysed as antecedents of the motivations towards the use of eWOM.

It would also be interesting to address the study of variables linked to service innovation. Recent works have investigated the relationship between hotel innovation and competitiveness during and after the pandemic (Gössling et al., 2021; Sharma et al., 2021; Shin & Kang, 2020). To this regard, information and communication technologies are key so that eWOM behaviour can spread information among consumers about the innovations that a hotel implements. Therefore, the study of the perception of innovation and its communication through the consultations and dissemination of EWOM would also help to improve knowledge of antecedents to purchase decisions in the current context of the pandemic.

Disclosure statement

The authors declare no conflict of interest.

Additional information

Funding

This research has been developed within the framework of the project Grant PID2020-112660RB-I00 funded by MCIN/AEI/10.13039/501100011033, the Grant for Consolidated Research group AICO/2021/144 funded by the Conselleria d’Innovació, Universitats, Ciència i Societat Digital of the Generalitat Valenciana and the Funding for Special Research Actions of Universitat de Valencia (Reference no.: UV-INV-AE-1553911).

Notes

1 2009 print edition and subsequent revised editions available at http://www.tourspain.es/es-es/Paginas/index.aspx

2 http://comunitatvalenciana.com/viaje/alojamiento/hoteles