1 Introduction

The premise for economic transactions up until the twentieth century was the ownership of products. However, as a result of the changing attitudes of consumers towards ownership of product, the premise for economic transactions in the twenty-first century are sharing, swapping, trading, or renting, paving way for collaborative consumption (Botsman & Rogers, 2010). Collaborative consumption is growing at an unprecedented rate and is fuelled by the wider adoption of information and communication technologies (ICT) by consumers, mostly through mobile phones. Mobile phones became the highest scale of consumer technology, representing around five billion unique mobile subscribers in 2017 (Gsmaintelligence, 2017). This digital upsurge is disrupting multi-layered companies, their supply chains, designers, manufacturers, distributors, wholesalers, and retailers. The power of the Internet, coupled with mobile phones, has unleashed new businesses. For instance, two of the world’s largest companies Airbnb and Uber in accommodation and taxi segment respectively own no inventory or vehicles(Goodwin, 2015; Tussyadiah et al., 2018). Instead, these companies provide a collaborative platform over the Internet to match individuals that need something to individuals that have something to offer, giving rise to a new type of peer-to-peer commerce, referred to as “collaborative consumption” or “sharing economy” (Tussyadiah & Pesonen, 2016). Market research firms speculate that the size of the sharing economy could increase exponentially to reach a market size of 335 billion USD by 2025 from just 15 billion USD in 2015 (PWC, 2015).

Airbnb has successfully positioned itself as prominent sharing economy platform in the tourism and hospitality sector with its homelike offerings, strong market share, and online presence. Airbnb has not only become a competitor of online travel agents, but also to established traditional hotel chains (Guttentag, 2015; Guttentag et al., 2018; Wang & Jeong, 2018). The valuation of Airbnb stood at 31 billion USD as at March 2017. This valuation is roughly same as the established hospitality chain Marriott International during that period (Wallenstein & Shelat, 2017a). Airbnb has ambitious growth plan to reach one billion users on its platform by 2028. To achieve this goal, it considers India as top three strategic markets (Pani, 2018). Boston consulting group conducted extensive market research on consumer attitudes towards sharing economy in three countries: Germany, India, and USA. Their survey on 3500 consumers found 93% of Indian consumers had used sharing economy platforms, a figure much higher than that for the USA and Germany. More than half of all mobile connections in markets like India involve smartphone, encouraging newly accessible consumer groups for innovations related to sharing economy (Wallenstein & Shelat, 2017a, 2017b, 2017c). India emerged as the second largest mobile market in the world with number of mobile subscribers reaching 1.1 billion as on July 2019 (Statista, 2019). Thus, emerging economies such as India is a promising market for sharing services like Airbnb (Li, 2016).

The tourism and hospitality industry is a pioneer of the sharing economy. Start-ups such as Airbnb allows tourists and residents to share their expert local knowledge and homes (e.g. as local tour guides). Developing countries like India with its unprecedented growth of smartphone users, provide a platform for sharing economy companies like Airbnb to achieve their growth objectives. However, there is a paucity of studies on the sharing economy in countries such as India. Cheng (2016) extant literature review on the sharing economy suggests that most of the existing empirical investigations have focused on the Western regions. The rapid growth of sharing economies in emerging regions can unfold unique findings that could be of paramount importance to both research and practice. Thus, the aim of this study is to understand Indian consumers’ intention to use the Airbnb platform. The research is grounded in the theoretical postulates of the UTAUT model. UTAUT is the most comprehensive theory in the information systems domain to understand individual technology adoption and use.

The remainder of this paper is structured as follows: Section 2 reviews extant literature on sharing economy with respect to Airbnb platform. This leads to theoretical basis and hypothesis development in Section 3. Research methodology constitutes Section 4 followed by results in Section 5. The penultimate Section 6 discuss findings emerging from this research investigation including separate sub-sections on implications for theory, implications for practice, limitations and future research directions. Finally, the paper ends with conclusion in Section 7.

2 Airbnb User Types, Theories Used, and Countries Examined

Google scholar search for keyword “Airbnb” yields 63,800 hits as of July 2020 indicating significant research interest on this collaborative consumption platform within twelve years of its formation in the year 2008. Dann et al. (2019) in their literature review on Airbnb found majority of Google scholar references do not represent genuine research on Airbnb (but rather brief mentions). The depth review of 243 articles revealed multifaceted nature of the platform comprising research on many domains, including Information and Management, Tourism/Travel/Hospitality, Law, and Economics (Dann et al., 2019). Later, Guttentag (2019) reviewed 132 peer-reviewed journal articles and divided them into six categories based on content analysis such as: 1) Airbnb guests, 2) Airbnb hosts, 3) Airbnb supply and its impacts on destinations, 4) Airbnb regulation, 5)Airbnb’s impacts on the tourism sector, and 6) the Airbnb company. Since, the focus of this study is on Indian travellers, a more in-depth review of various user types of Airbnb platform is necessary.

2.1 Airbnb User Types

Research on Airbnb user types can be broadly classified into two categories: 1) Airbnb guests – the travellers who book their accommodation, and 2) Airbnb hosts – the house owners who offer/lists them on the platform (Guttentag, 2019). The first stream ‘Airbnb guests’ - focus on guest/travellers/consumers interaction with the Airbnb platform and its impact. For instance, Bae et al. (2017) employed information processing theory as theoretical lens to conduct experimental survey among 411 Airbnb travellers to determine their purchase intention and sharing behaviour. The results revealed factors such as social distance, credibility of reviews, review breath, information usefulness, and adoption of reviews significantly influenced their purchase intention. Meanwhile, social distance reduced the equivocality and significantly influenced traveller’s credibility of shared experiences (Bae et al., 2017). In addition, numerous researchers have empirically examined question on travellers’ motivation to use Airbnb platform repeatedly highlighting the utilitarian and experiential benefits of the platform (Guttentag, 2019). For example, Guttentag et al. (2018) study on over 800 Airbnb users segmented respondents into five different segments – Money savers, Home seekers, Collaborative Consumers, Pragmatic Novelty Seekers, and Interactive Novelty Seekers based on their distinctive motivation characteristics.

Airbnb hosts - the second stream of research focus on host (sellers) interaction with the Airbnb platform and its impact. For instance, Ert et al. (2016), studied host (sellers) photo on guests (consumers) decision making. Host attributes play vital role in delivering high quality peer-peer accommodation services as consumers are exposed to larger risks than monetary loss (Lauterbach et al., 2009). Such instances include but are not limited to report of Airbnb hosts sexual assault on nineteen year old during his stay in Madrid (Lieber, 2015b), and attack on guest Mike Silverman by his host’s dog during his stay in Salta, Argentina that made the guest to spent two nights in the hospital (Lieber, 2015a). The above instances demonstrate the extent of risk exposure to guests in sharing economy platforms; making trust and reputation of the host pivotal for proper functioning of such platforms. Ert et al. (2016) argue host attributes and product attribute affect potential guest purchase decision. Host attributes comprise of three major things: trustworthiness, attractiveness based on hosts photo (visual information), and reputation based on host review rating (non-visual information). The empirical results through controlled experiment and Airbnb data in Stockholm, Sweden found hosts with more trustworthy photo could commend higher price for their listing and have more probability of getting booking.

2.2 Theories Used, and Countries Examined

This study employed keywords such as “Airbnb” AND “guests” OR “consumers” OR “travellers” in Scopus database to locate papers relevant to the focal phenomenon – Airbnb guests (Patil et al., 2020). The keyword search returned 160 initial articles out of which only 45 focused on consumer/travellers/guests were fully downloadable. Researchers in collaborative consumption arena employed as much as fourteen unique theories/models as theoretical lens across these 45 studies to examine travellers/consumers/guest’s usage of Airbnb platform. Table 1 provides summary of dominant theories that are employed on two or more such instances. The most dominant theory to examine Airbnb guests’ interactions on the platform was Theory of planned behaviour with three examinations. Researchers used this theory to examine both travellers purchase (Chatterjee et al., 2019) and repurchase intention (Mao & Lyu, 2017) on the platform. While, Price emerged as the significant determinant of purchase intention; Attitude and subjective norms as significant determinants of repurchase intention (Chatterjee et al., 2019; Mao & Lyu, 2017). Four theories/models such as Attribution theory, Prospect theory, Social exchange theory, and Extended unified theory of acceptance and use of technology (UTAUT2) jointly occupied second position by serving as theoretical lens on two instances each. Researchers employed attribution theory to examine tourists switching intention in Malaysia a new outcome variable (Tiamiyu et al., 2020a); while the remaining three theories/model examined travellers behavioural intention to use the platform (e.g., Aruan & Felicia, 2019). Tiamiyu et al. (2020a) results on Malaysian travellers revealed price unfairness negatively affects psychological engagement, which in turn negatively affects tourists’ switching intention. Furthermore, three were nine theories/models such as: 1) Attachment Theory (Yang et al., 2019), 2) Innovation Diffusion Theory (IDT) (Wang & Jeong, 2018), 3) Information processing Theory (Bae et al., 2017), 4) Mean-end chain theory (Aruan & Felicia, 2019), 5) Mehrabian-Russell model (So et al., 2020), 6) Push–pull–mooring (PPM) model (Yan et al., 2019), 7) Reactance Theory (Teubner & Graul, 2020), 8) Stimulus-organism-response theory (Mody et al., 2017), and 9) Technology acceptance model (TAM) (Wang & Jeong, 2018) that researchers employed on one instance each to examine travellers use of Airbnb platform. Finally, the extant literature review found majority of the existing studies focused on developed western countries such as the USA (Lee & Kim, 2018; Li et al., 2019; Mao et al., 2020), Canada(Guttentag et al., 2018; Liang et al., 2018), and Germany (Amaro et al., 2019; Möhlmann, 2015) with just one study focusing on India (Chatterjee et al., 2019).

Table 1 Dominant Airbnb theories/models
Full size table

In summary, the literature review reveals Theory of planned Behaviour as the most used theoretical lens and USA as country most examined on issues pertinent to the Airbnb guests. This uncovers significant research gap to examine this focal phenomenon using unified theory of acceptance and use of technology (UTAUT) often rated as most comprehensive theory in understanding individual technology acceptance in Indian context (Tamilmani et al., 2018, 2020).

3 Theoretical Basis and Hypothesis Development

Information systems once considered as an applied discipline has come of age with native theories to emerge as reference discipline for other domain on its own right (Baskerville & Myers, 2002). First native theory in the information systems discipline is Davis (1989), TAM based on theory of reasoned action (TRA). This parsimonious model with more than 40,000 Google Scholar citations emerged as dominant theory beyond IS domain to understand user acceptance of technology and have been integrated with multiple theories/constructs across range of individual (Childers, Carr, Peck, & Carson, 2001), organisational (Wixom & Todd, 2005), and societal (Carter & Weerakkody, 2008) contexts. To that extent, within tourism and hospitality sector, Wang and Jeong (2018) integrated TAM and IDT to evaluate user’s adoption of Airbnb platform. Besides TAM, with more than 20,000 Google Scholar citations another popular native IS theory is unified theory of acceptance and use of technology (UTAUT).

3.1 UTAUT Based Theories

Understanding “why” individuals embrace information technologies and “why not” is one of the most mature research streams in the contemporary IS research arena and has been continuous puzzle for research and practice (Venkatesh et al., 2007; Williams et al., 2009). In addition to TAM, a plethora of models including diffusion of innovation (DoI) theory, IDT, and model of personal computer utilization were proposed to explain individual adoption of IS/IT on variety of technology attributes and contextual factors. This multitude of context and theory presented new challenge of plurality to IS researchers until emergence of Venkatesh, Morris, Davis, and Davis (2003) comprehensive unified theory of acceptance and use of technology (UTAUT) that eliminated similar/redundant constructs through exhaustive review, mapping and integration of eight dominant technology adoption models (see Venkatesh et al., 2003 for review). The focal phenomenon of UTAUT was organisational users of technology primarily driven by their extrinsic motivation emphasizing on the utilitarian value.

3.2 Revised UTAUT Model

UTAUT – the most comprehensive theory in IS arena is not without limitations. Venkatesh et al. (2016) literature synthesis on 1267 UTAUT citations found just 62 studies (approximately 5%) used the theory while the remaining 1205 studies (comprising 95%) just cited the article for generic reasons without using UTAUT model or its constructs. These findings revealed researchers scarcely employ UTAUT model on its entirety with majority of studies omitting the moderator relationships. To overcome shortcomings of UTAUT, Dwivedi et al. (2019), re-examined the model using combination of meta-analysis and structural equation modelling (MASEM) techniques. The findings from 1600 data points spread across 162 prior studies revealed individual differences variable attitude partially mediated the effects of all four UTAUT exogenous variables to behavioural Intention and had direct effect on use behaviour. In addition, the study found significant association between facilitating conditions and behavioural intention that was not part of the original model. Finally, the study excluded moderators of UTAUT model as none of its guiding theories such as TPB and TAM included one. Moderators become relevant only if significant variation existing among individuals examined in same context and hence, they are not universally applicable. The MASEM based UTAUT model is robust alternative to examine individual technology adoption and use as it addresses the shortcoming of UTAUT (Dwivedi et al., 2019). Henceforth, this study will refer to the re-examined model as meta-UTAUT. The extant literature review suggests scarce utilization of robust information system theories in tourism and hospitality sector. This study aims to fulfil this research gap through adaption of meta-UTAUT model to examine consumer adoption of Airbnb in India. Figure 1 depicts the proposed research model with appropriate hypotheses.

Fig. 1
figure 1

Proposed research model (Source: Adapted from Dwivedi et al., 2019)

Full size image

3.2.1 Performance Expectancy

Performance expectancy refers to the extent, which an individual believes that using a technology will help him or her to improve productivity of underlying task (Rana et al., 2015 and b). The extant literature has acknowledged similarities of this construct to others such as usefulness and extrinsic motivation (Davis et al., 1989, 1992) and job-fit and outcome expectations (Compeau & Higgins, 1995). Meta-UTAUT model found technology capabilities such as performance expectancy as a significant predictor of individual attitudes towards using the underlying technology, in addition to their intention. Tourism and hospitality researchers have also reported significant impact of perceived usefulness on consumer’s attitude towards using technologies such as hotel reservation websites (Morosan & Jeong, 2008) and Airbnb (Wang & Jeong, 2018). Since, emergence of UTAUT model in 2003, a reasonable number of studies have reported significant direct effect of performance expectancy on consumer intention (Escobar-Rodríguez & Carvajal-Trujillo, 2013; Escobar-Rodríguez & Carvajal-Trujillo, 2014; Rana et al., 2015 and b). Based on the preceding discussions, this study proposes the following hypotheses:

  • H1: Performance expectancy will positively influence Indian consumer’s attitude to use Airbnb platform.

  • H2: Performance expectancy will positively influence Indian consumer’s intention to use Airbnb platform.

3.2.2 Effort Expectancy

Effort expectancy refers to the extent of ease associated with individual’s technology use (Venkatesh et al., 2003). Davis et al.’s (1989) TAM model found perceived ease of use a root construct of effort expectancy as significant determinant of individual’s attitude towards using technology. Along similar lines, meta-UTAUT found attitude to partially mediate the effect of effort expectancy on behavioural intention (Dwivedi et al., 2019). Researchers in tourism and hospitality sector found mixed results for perceived ease of use/effort expectancy and attitude path relationship. For instance, research on online travel communities (Casaló et al., 2010) and hotel front office systems (Kim et al., 2008) found ease of use as significant determinant of individual’s attitude. On contrary, Wang and Jeong (2018) research on Airbnb platform and Chang and Caneday (2011) research on tourist’s use of Webgis platform found ease of use as a non-significant determinant of consumer’s attitude. In addition, prior literature in hospitality sector has confirmed the significant impact of perceive ease of use on perceived usefulness towards consumer’s adoption of technology (Kim et al., 2008). Based on above discussion, this study proposes following hypotheses:

  • H3: Effort expectancy will positively influence Indian consumer’s attitude to use Airbnb platform.

  • H4: Effort expectancy will positively influence Indian consumer’s performance expectancy to use Airbnb platform.

3.2.3 Social Influence

Social influence refers to the extent to which consumers perceive that important others in their life (e.g., family and friends) believe they should use a particular technology. Prior literature examined significant relationship between subjective norm (a root construct of social influence) and consumer attitude towards use of technology based on theory of planned behaviour. Few such instances include traveller’s intention to use green hotels in Taiwan (Teng et al., 2015) and repurchase intention on sharing economy platform Airbnb (Mao & Lyu, 2017). Meta-UTAUT model reaffirmed the significant relationship of social influence on attitude and inclusion of this relationship improved the explanatory power of the theoretical model. In addition, Venkatesh and Bala (2008) proposed TAM3 to evaluate various determinants of perceived usefulness and perceived ease of Use. TAM3 found subjective norm as significant antecedent of perceived usefulness among organisational user of technology. Internalization process is basis for this association that involves individuals’ incorporate referents belief into their own belief structure (Warshaw, 1980). For instance, Lu et al. (2005), research on consumer intention to adopt wireless mobile technology found social influence as significant antecedent of perceived usefulness. Hence, this study proposes following hypotheses:

  • H5: Social influence will positively influence Indian consumer’s attitude to use Airbnb platform.

  • H6: Social influence will positively influence Indian consumer’s performance expectancy to use Airbnb platform.

3.2.4 Facilitating Conditions

Facilitating conditions refers to the extent to which consumers believe that organisational support and technological infrastructure are available to them for using a particular technology (Venkatesh et al., 2003).Taylor and Todd (1995) acknowledged the similarity of facilitating conditions with perceived behavioural control in their decomposed theory of planned behaviour. Venkatesh et al. (2003), argued the presence of performance and effort expectancy as predictors of behavioural intention in UTAUT model will affect the role of facilitating conditions. Based on the assumption they hypothesized and found facilitating conditions as non-significant predictor of behavioural intention and significant direct predictor of use behaviour (Rana et al., 2016). However, unlike organisational settings the facilitating conditions such as technology applications, vendors, devices used to assess technology and run a application can vary for each individual. To accommodate such scenarios, Venkatesh et al. (2012) in their extended UTAUT model (UTAU2) focused on consumer setting included the link between facilitating conditions and behavioural intention expecting it to behave more like perceived behavioural control from TPB. Numerous studies on consumer technology adoption validated the significant effect of facilitating conditions on behavioural intention (Rana et al., 2016). Tourism and hospitality sector reported mixed findings for this path relationship. While, So et al. (2018), found perceived behavioural control (a root construct for facilitating conditions) as significant predictor of consumer behavioural intention among Airbnb consumers in the US; Mao and Lyu (2017), study on the same country and platform found the effect to be non-significant for consumer repurchase intention. Based on above discussion, this study proposes the following hypothesis:

  • H7: Facilitating conditions will positively influence Indian consumer’s attitude to use Airbnb platform.

3.2.5 Hedonic Motivation

‘Hedonic motivation’ is defined as the fun or pleasure derived from using technology and it is an important determinant of consumers technology acceptance and use (Baabdullah et al., 2019; Brown & Venkatesh, 2005). The primary driver of hedonic motivation is intrinsic motivation where user perform activity for end in itself. Hedonic motivation is the most important theoretical addition to the UTAUT2 as it integrated the much-needed affective component into largely cognition based UTAUT. It plays significant role in determining consumer behaviour intention focused on hedonic outcomes than utilitarian outcomes (Tamilmani et al., 2019). However, recently researchers have found the significant effect of hedonic motivation on effort expectancy that in turns played vital in determining consumer use of technologies even in case of utilitarian technologies (Koenig-Lewis et al., 2015; Robin et al., 2016). Hence, this study proposes following hypothesis:

  • H8: Hedonic motivation will positively influence Indian consumer’s effort expectancy to use Airbnb platform.

3.2.6 Attitude

Dominant adoption theories most often employed attitude as their mediating variable to predict consumer behavioural intention. Attitude towards behaviour is the extent to which individuals have positive or negative evaluation about the behaviour under question. This leads to favourable attitudes towards behaviour with desirable consequence and unfavourable attitudes towards behaviour with undesirable consequence(Ajzen, 1991; Dwivedi et al., 2017). Despite, the major role of attitude among UTAUT guiding theories, the final UTAUT model excluded the construct. Contrary to UTAUT, Dwivedi et al. (2019) established the significance of attitude through MASEM and reinstated attitude in their model paving way for better explanation of variance on consumer use of technology. Tourism and hospitality researchers have widely affirmed that individual’s positive attitude towards new technology ultimately results in higher likelihood to use the underlying technology (Amaro & Duarte, 2015; Park et al., 2017; Wang & Jeong, 2018). The current study assesses consumer intention to use Airbnb platform in Indian context and thus propose following hypothesis:

  • H9: Attitude will positively influence Indian consumer’s behavioural intention to use Airbnb platform.

3.2.7 Self-Efficacy

Social cognitive theory (SCT) developed by Bandura (1986), is one of the most powerful theories to determine human behaviour comprising self-efficacy as one of its core construct. Self-efficacy is defined as the degree of individual’s judgement on their ability to use a technology (e.g., Computer, Websites) in order to accomplish their job or task (Bandura, 1986). Compeau and Higgins (1995), later expanded SCT with specific focus on self-efficacy to the context of individual technology use. Since then, self-efficacy was found to be significant direct determinant of behavioural intention on range of technologies such as: citizen e-government adoption (Lallmahomed et al., 2017); Consumer internet banking adoption (Alalwan et al., 2015). In addition, self-efficacy emerged as significant indirect determinant of behavioural intention through attributes such as: Affect and Outcome expectation in citizens e-government adoption(Rana & Dwivedi, 2015)); Effort expectancy in nurses adoption of electronic patient record (Maillet et al., 2015); and consumer motivation to book online hotel deals using travel intermediaries (Chen et al., 2016). In UTAUT model, though the effects of self-efficacy were significant during initial stages, the effects became non-significant over time in determining individual intention to use technology. Thus, Venkatesh et al. (2003), excluded self-efficacy in their UTAUT model stating effort expectancy capture the effect of self-efficacy. However, researchers continue to find significant direct effect of self-efficacy on resident’s attitudes towards support for tourism (Wang & Xu, 2015), and hotel guests recycling intention (Grazzini et al., 2018) within the tourism and hospitality sector. Given the preceding discussion on importance of self-efficacy and considering Indian consumers are still in early stages of adoption to sharing economy platforms, the study proposes following hypothesis:

  • H10: Self-efficacy will positively influence Indian consumer’s behavioural intention to use Airbnb platform.

3.2.8 Trust

Trust between two or more parties determine their future actions. Prior research has demonstrated, trust as the most important factor for establishing relationships, both of interpersonal and commercial nature (McKnight & Chervany, 2001; Waseem et al., 2018). Trust serves as central part of many economic transactions arising from human necessity of understanding their social surroundings to identify how others behave. However, the social environment is very difficult to comprehend, since individuals are free agents and as such, their behaviours are not necessarily rational or predictable (Gefen et al., 2003). The very nature of internet enabled collaborative consumptions makes it difficult to regulate its social environment through rules and customs. In such scenarios, individual adopt trust as an integral part of their social complexity reduction strategy (Luhmann, 2018). In the other words according to Reichheld and Schefter (2000, p. 10): “Price does not rule the Web; trust does” (p. 10).

The mediating role of internet adds further complexity to trust in collaborative consumption. Under such circumstances, trust splits into two forms rooted as: trust on the platform (guests and platform) and trust on the seller (guests and hosts)(Han et al., 2016; Tussyadiah & Pesonen, 2018). Mittendorf (2016), research on Airbnb platform found interaction between trust on platform and trust on host such that trust on former could subsequently increase trust on later by mitigating uncertainties. Instances of trust on platform as significant predicator include: Chinese consumers purchase intention in peer to peer to accommodation platform (Ye et al., 2019) and consumers active request on the Airbnb platform (Mittendorf, 2018). Hence, this study proposes following hypothesis:

  • H11: Trust will positively influence Indian consumer’s behavioural intention to use Airbnb platform.

4 Research Methodology

This study employed quantitative research methodology and used convenient sampling technique for data collection through self-administrated online surveys from June to August 2018 in India ((Saunders et al., 2007; Wang & Jeong, 2018). Researchers distributed survey instrument using both paper-print and web-based survey approaches to minimize selection bias and maximise the number of respondents (Evans & Mathur, 2005). Table 2 depicts all the nine constructs and their corresponding measurement items adapted for consumer adoption context of sharing economy platform Airbnb from the dominant information systems literature (e.g., Davis et al., 1989; Venkatesh et al., 2003; Venkatesh et al., 2012). Venkatesh et al. (2003), while developing UTAUT model selected only four highest loading items for each construct from their measurement model. This study adapted a similar approach, choosing three measurement items for each of the nine constructs with a factor loading of >0.60. Table 5 depicts that the factor loading for all the 27 measurements items are well and above >0.60 with attitude having all values >0.80 level. In order to measure the unobserved constructs indirectly through measurement items, the research employed seven-point Likert scale with scores ranging from 1 (=“Strongly disagree”) to 7 (=“Strongly agree”) for the survey. Tourism and hospitality sector researchers have adapted Likert scale measurement items for consumer adoption surveys on Airbnb (Wang & Jeong, 2018) and online tickets purchasing in low-cost carriers to name a few (Escobar-Rodríguez and Carvajal-Trujillo (2014). Researchers conducted pre-test of the survey among 25 UK based Indian students to increase the content validity of the measurement items. The pre-test results helped to modify wordings in the survey that respondents reported unclear while answering them.

Table 2 Survey constructs and measurement items(s)
Full size table

5 Results

5.1 Respondents Demographic Profile

This study obtained 301 valid responses from Indian consumers to evaluate their intention to use Airbnb platform. Table 3 depicts demographic characteristics of the survey respondents. Majority of the respondents were males at 59.1%, while the remaining 40.9% comprised of female respondents. In terms of age, cumulatively 81.4% respondents were 29 years or younger that increased to staggering 90% in terms of respondents 34 years and below. As far as the education of the respondents is concerned, many of them 53.2% were graduates followed by 25.9% at least completing high school education. Understandably, majority of the respondents 63.5% were students followed by the next largest group 23.3% represented by private sector employees.

Table 3 Demographic characteristics of respondents
Full size table

5.2 Descriptive Statistics

Table 4 depicts mean and standard deviation of nine constructs alongside their all twenty-seven constituent measurement items. Overall, the mean of all the measurement items were well above four almost touching the value of five in some cases. This indicates that Indian consumers responded favourably to use of sharing economy platform Airbnb at large. In terms of standard deviation (S.D), the values of majority of the items 23 out of 27 were stable with values below 1.5. However, the S.D. Values for remaining four measurement items: one each from Attitude (AT1) and Behavioural Intention (BI1); and the remaining two from self-efficacy (SE1 and SE3) were above 1.5. The higher values underscores divergence among respondents forming varying opinions about the same underlying question for these attributes. The next step involved reliability analysis of measurement scales through calculation of Cronbach’s alpha. The value of Cronbach’s alpha provides indication on the internal consistency of the items measuring the same construct time and again (Hair et al., 1992). Cronbach’s alpha value for all the construes (see Table 4) were over and above the recommend value of 0.70 emphasizing on their high internal consistency in measuring the respective constructs (Hair et al., 1992; Nunnally, 1978). Cronbach’s alpha value ranged from minimum of 0.737 for Facilitating Conditions (FC) to maximum of 0.851 for Attitude (AT) indicate relatively high reliability of the constructs.

Table 4 Mean and Standard Deviation of constructs
Full size table

5.3 Measurement Model

Confirmatory factor analysis (CFA) was conducted to test the convergent and discriminant validity of constructs. Researchers adapted Anderson and Gerbing (1988), three ad hoc tests such as standardised factor loadings, composite reliabilities, and average variance extracted to measure the convergent validity of latent variables. The level of association among measurement items and a single latent variable determines their standardised factor loadings. The factor loading of all the measurement items measuring a latent variable were high with values >0.60 (see Table 5). In addition, composite reliability an indicator similar to Cronbach’s alpha had values beyond minimum threshold of 0.70 (Hair et al., 1992; Kapoor et al., 2015; Nunnally, 1978). Finally, average variance extracted (AVE) value that measures variation explained by the latent variable to random measurement error ranged from 0.577 for performance expectancy (PE) to 0.784 for Attitude (AT). These estimates of AVE were relatively higher than the stipulated lower limit of >0.50 (Fornell & Larcker, 1981). The significant value for all the three tests confirms high convergent validity for all measurement scales and respective latent construct.

Table 5 Results of confirmatory factor analysis (CFA)
Full size table

Researchers again adapted Anderson and Gerbing (1988) recommendations for testing discriminant validity. Latent variables qualify discriminant validity test, when the factor correlation among a pair of latent variables is always less than the square root of AVE of each variable in the factor correlation matrix. Evaluation of discriminant validity as depicted in Table 6 reveals that square root of AVE shown in bold fonts across the diagonal is always greater than the correlation value for any pair of variables. For instance, the correlation value of 0.670 between performance and effort expectancy is less than the bold square root of AVE values shown along the diagonal of these variables 0.760 and 0.844 respectively. In other words, the proposed research model satisfy discriminant validity condition for all latent variables, which states for a variable to be considered distinct from other variables the square root of AVE of the variable should be greater than its correlation with other latent variables (Smith & Barclay, 1997).

Table 6 Factor correlation matrix
Full size table

[Legend: AVE = Average Variance Extracted, CR = Composite Reliability, FL = Factor Loading].

Finally, Table 7 depicts the fits statistics of the measurement model from CFA. The test of measurement model resulted in χ2 value of 408.135 with 284 degrees of freedom and a significant probability value of less than 0.001. Bentler (1995), proposed the χ2 over degrees of freedom ratio as better measure of model fit with values ranging from three to one. The ratio of the proposed measurement model 1.437 is well within the stipulated limit. Besides, this study also reports descriptive fit statistics that compares a specified model, typically an independent model to show the superiority of the proposed model. The fit statistics reported in this study include Adjusted Goodness of Fit Index (AGFI), Comparative Fit Index (CFI), and Root mean Square Error Approximation (RMSEA). Anderson and Gerbing (1988), found CFI as one of the most stable and strong fit indices with acceptable values of greater than or equal to 0.90 (>0.90) (Hoyle, 1995). In the case of AGFI, the acceptable value should be greater than or equal to 0.80 (>0.80). Finally, although values of values of less than are equal to 0.10 (<0.10) are acceptable for RMSEA, Hair et al. (1998), recommends models should possess RMSEA value of lesser than or equal to 0.08 (<0.08) for better fit. Table 7 illustrates all measurement model indices are well within the recommended level.

Table 7 Model fit summary measurement model
Full size table

[Note: Square root of AVE on diagonals in bold; p > 0.05: Non-significant; *p < 0.05; **p < 0.01; ***p < 0.001].

5.4 Structural Model Testing

This study conducted structural equation modelling (SEM) analysis using AMOS Graphics and SPSS Version 25 software to test hypothesis from H1 to H11. Like measurement model, analysis involved calculation of goodness of fit indices for the structural model. Results revealed significant probability value of less than 0.001, with χ2 value of 702.05 and 303 degrees of freedom yielding desirable chi-square to degree of freedom ratio of 2.317. Table 8 depicts that fit statistics for other indices were also within the threshold value such as AGFI = 0.829, CFI = 0.902, and RMSEA = 0.066.

Table 8 Model fit summary structural model
Full size table

Researchers after establishing adequate fit indices for the structural model proceeded to path examination of all eleven proposed hypotheses. Table 9 depicts the results of structural model path analysis. This study found support for ten out of eleven proposed hypotheses from H2 to H11; except for hypothesis from performance expectancy to attitude (H1) that was surprisingly found non-significant. The independent constructs effort expectancy, social influence, and facilitating conditions significantly influenced attitude (i.e., H3, H5, and H7); Meanwhile, Performance expectancy, self-efficacy, and trust significantly influenced behavioural intention (i.e., H2, H10, and H11). Moreover, Attitude significantly mediated the effects of various independent constructs to behavioural intention (H9). In addition, social influence and effort expectancy significantly influenced Performance expectancy (H6 and H4), while hedonic motivation significantly influenced effort expectancy (H8) of Indian consumer intention towards adoption of sharing economy platform Airbnb. In terms of significant paths, facilitating conditions was the strongest predictor of Attitude with path coefficients value of 0.851 and as predicted attitude was the strongest predictor of behavioural intention with path coefficients value of 0.295.

Table 9 Hypotheses test results
Full size table

[Legend: Significance of p value: p > 0.05 (N.S): Non-significant; p < 0.05: *; p < 0.01: **; p < 0.001: ***; R2: Variance explained].

Figure 2 depicts the validated research model emerging form structural model results with path coefficients and significance of each relationship. In addition, the research model also reveals the variance explained on each of the four dependant constructs (Performance expectancy, effort expectancy, attitude and behavioural intention). This research model explained as much as 65% variance of Indian consumer intention to adopt sharing economy platform Airbnb.

Fig. 2
figure 2

Validated research model

Full size image

6 Discussion

The proposed model explained about 65% variance on Indian consumer intention to use Airbnb platform. This is much higher in comparison to Wang and Jeong, (2018) study that explained about 60% of variance on USA consumer intention to Airbnb platform – a rare stu the tourism and hospitality sector that employed dominant information system theories such as TAM and IDT as their baseline model. The strong variance explained by the current study is due to inclusion of attitude from the meta-UTAUT model, which emerged as the strongest predictor of consumer intention to use Airbnb platform.

The most intriguing finding was the non-significant path relationship between performance expectancy and attitude (H1). Though, this is contrary to the meta-UTAUT model, performance expectancy was found to be significant director predictor of behavioural intention at p < 0.05 level (H2). Davis et al. (1989), TAM model offers plausible explanation for this pattern that hypothesized perceived usefulness (a root construct) of performance expectancy as determinant of behavioural intention over and beyond indirect effect through attitude. Davis et al. (1989) argues’ failure to activate individual’s affective conditions fully while deciding to use system will affect their adoption decision. Therefore, attitude would not be expected to fully capture individual’s performance considerations on their intention. In such scenarios, individuals form intentions towards behaviour that they believe would improve the performance of underlying task over and above their positive or negative feelings evoked toward the behaviour. Davis et al. (1989) research on 107 MBA students over two different times found Perceived usefulness to be significant predictor of intention on both time period and significant on attitude only during initial stage but found to be non-significant predictor during later stage of technology use. The demographic profile of current study reveal 90% of respondents were of below 34 years either millennials or post-millennials comprising mostly students well versed with technology (Dimock, 2018). Given the preceding discussion, it would be a safer argue that sharing economy platform such as Airbnb does not necessarily require attitude formation as consumers use them for productivity gains and the platform usefulness directly influence consumer behavioural intention. Prior literature provide empirical support for both non-significant effect of Performance expectancy/perceived usefulness on attitude (Bajaj & Nidumolu, 1998; Curran & Meuter, 2005) and significant effect of Performance expectancy on behavioural intention on individual technology adoption. Few such significant instances within tourism and hospitality sector include research on tourist adoption of smartphone apps (Gupta et al., 2018), and consumer intention to reuse mobile apps for making hotel reservations (Fong et al., 2017).

Performance expectancy was the only non-significant determinant of attitude (H1) while the rest three antecedents such as effort expectancy (H3), social influence (H5), and facilitating conditions (H7) all had significant effect underscoring the central role of attitude as a mediator in determining Indian consumer intention to use sharing economy platform Airbnb. In other words, attitude formation is priori condition for some individual beliefs (effort expectancy), normative beliefs (social influence), and control beliefs (facilitating conditions) in order to determine Indian consumers behavioural intention towards Airbnb. This is not surprising, as number of existing studies have found empirical evidence for these associations both in the tourism and information systems literature. Researchers found significant impact of ease to use/ effort expectancy on consumers attitude formation towards hotel booking website (Morosan & Jeong, 2008) and Airbnb platform in U.S (Wang & Jeong, 2018). In addition to attitude, effort expectancy emerged as significant determinant of performance expectancy. EE➔PE significant association originated from their root constructs perceived ease of use and perceived usefulness in Davis et al. (1989), TAM model that argued systems easier to use can save individuals more time which can be in turn effectively employed to accomplish more work for the similar efforts. In other words, Indian consumers expect their interaction with sharing economy platform Airbnb to be less complex with easier to use interface ultimately enabling them to book accommodation much faster and efficient.

Moreover, social influences significant impact on attitude was not surprising as individuals can refine their attitude towards sharing economy platform based on information gathered from friends, family, and important others experiences with similar platforms (Mao & Lyu, 2017; Teng et al., 2015). In addition to significant direct relationship with attitude, social influence also had higher indirect effect on attitude through performance expectancy (H6). The strength of SI➔PE (0.324***) was much higher than the strength of SI➔AT (0.260**) relationship. The above results imply that Indian consumers perceive their own behaviour to be highly affected by important referents in their life. Teo et al. (2008), work on preservice teacher’s attitudes towards computers found similar pattern with subjective norm having stronger significant influence on perceived usefulness than attitude towards using computer.

The final antecedent of attitude, facilitating conditions also had significant impact in determining consumer attitude and was the strongest with path value of 0.851***. Prior research has reported significant indirect effects of facilitating conditions on attitude and direct effect on consumer intention to use technology. For instance, Dieck and Jung (2018), research on tourist intention to use mobile augmented reality (AR) applications found availability of better hardware, uninterrupted internet, and longer battery life significantly influenced tourist’s usefulness and ease of use, that ultimately determined their attitude towards using AR. Similarly, Fong et al. (2017), research on consumer intention to reuse mobile apps for making hotel reservations found facilitating conditions such as support services to play vital role in predicting consumer behavioural intention. Along similar lines, the current research found facilitating conditions as the strongest determinant of attitude based on meta-UTAUT model.

This research included three additional constructs such as hedonic motivation, self-efficacy, and trust that were not part of meta-UTAUT model and found all of them to be significant on their respective dependant variables. First, Venkatesh et al. (2012), integrated “hedonic motivation” an affective component into largely cognition based UTAUT as determinant of behavioural intention to be more relevant to consumer context driven by their intrinsic motivation. However, the current study mapped hedonic motivation only to effort expectancy based on Venkatesh (2000) grounded on anchor and adjustment framework that states enjoyable system will be perceived more easier to use. Moreover, consumer use of Airbnb is more of utilitarian purpose rather than hedonic. HM➔EE(H8) path emerged as the second most significant path with path value of 0.816*** underscoring the fact that enjoyable technologies are do perceived as easier to use. Second, the extant literature in tourism and hospitality has mostly examined the role of self-efficacy on consumer behavioural intention indirectly thorough their attitude (Wang & Chen, 2015). This research went a step further and found self-efficacy as direct determinant of consumer behavioural intention.

Finally, trust emerged as the second strongest direct determinant of Indian consumer intention to use sharing economy platform Airbnb with path value of 0.254**. Prior research found trust to play dominant role among both hosts (damage to property) and guests (safety concerns such as: hidden camera; burglary; rape; robbery) in sharing economy platform (Liang et al., 2018). Moreover, individuals adopt trust as an integral part of their social complexity reduction strategy in sharing economy platforms (Luhmann, 2018). For instance, Mittendorf (2016) found guest trust on Airbnb platform could improve their trust on Airbnb host to mitigate uncertainties. Furthermore, researchers have found lack of trust on the Airbnb platform as single most inhibitor negatively affecting consumers overall attitude towards the use of platform (So et al., 2018). A number of studies have established the significant impact of trust towards consumer behavioural intention to use sharing economy platform (Mittendorf, 2018; Wang & Jeong, 2018). To that extent, this study reaffirms the role of trust as one of the most dominant predictor of Indian consumer intention to use Airbnb platform. Finally, attitude the core construct of meta-UTAUT model still emerged as the strongest predictor of consumer behavioural intention with path coefficient value 0.295***.

6.1 Contributions to Theory

The confrontation among tourism and hospitality sector researchers often to “pick and choose” constructs from multitude of theories for their investigations demonstrates dearth of native theories. Researchers have made efforts to provide integrated models by drawing different constructs from multiple theories (So et al., 2018) or used dominant theory from another domain as baseline model and extended them to tourism sector (Wang & Jeong, 2018). However, these studies have not exploited the capability of UTAUT that is well citied and comprehensive model in the information systems arena to understand individual technology adoption and use. To that extent, this study adapted Dwivedi et al. (2019), meta-UTAUT as baseline model and extended the model with trust, hedonic motivation, and self-efficacy. Researchers tested the proposed model in emerging country (India) that explained 65% variance of Indian consumer intention to use Airbnb platform much higher in comparison to similar studies that explained 60% and 56% variance on U.S (Wang & Jeong, 2018) and Korean consumer (Bae et al., 2017) respectively on similar platforms. The proposed model addresses the issue of both rigour and relevance through adaption of comprehensive meta-UTAUT as baseline model and extending it with context specific variables. Thus, this research contributes to the tourism and hospitality researchers by providing research model to readily choose their construct minimizing their confrontation in operationalizing their constructs during research model design.

6.2 Implications for Practice

This first comprehensive research on Indian consumer intention to adopt Airbnb offer number of implications for the practitioners. This research model offers holistic understanding on various attributes affecting Indian consumers intention to book accommodation on Airbnb platform. This is paramount importance to Airbnb marketers as they consider India in list of top three strategic market to reach their ambitious plan of one billion users on its platform by 2028 (Pani, 2018). To achieve this goal, marketers should assess each significant variable from this research model and focus on the specific attributes depending on the stage of the marketing plan to allocate necessary resources.

This research demonstrates individual beliefs performance and effort expectancy directly influence their intention and attitude respectively. This underscore the degree of importance that individuals place on the usefulness and ease of using platforms to accomplish their underlying task. Airbnb should leverage on these attributes by providing accurate representations of user requirements to website developers and product designers. This can result in reduced user ambiguity and delivery of an engaging platform with sleek design and robust user interface (Dwivedi, et al., 2019). In terms of normative beliefs, social influence significantly affected both usefulness and consumer attitude towards using the sharing economy platform Airbnb. In collectivist country like India, individuals place more emphasis on their referents opinions though internalization process and incorporate their referents belief into their own belief structure(Warshaw, 1980). Referents could be friends, neighbours, family members, existing user of product/service and inspirational figures in the media such as sports heroes or movie stars (Brown et al., 2002). Researchers have demonstrated that potential consumer’s interaction with current consumers can significantly reduce the uncertainty of using product/service and improve the quality of customer-firm relationship. This provides an additional medium for organisations to influence prospective consumers through referral programs and/or advertisement campaigns(Kulviwat, Bruner II, & Al-Shuridah, 2009).

Moreover, individual control belief facilitating conditions emerged as the most significant predictor of Indian consumer attitude towards adoption of Airbnb. Prior research has demonstrated that facilitating condition such as help desks, training programs, and adequate training programmes could potentially improve Indian consumer’s adoption towards e-government services (Rana et al., 2016). In addition, emerging economies present unique challenges such as low internet speed, intermittent connection, and storage capacity as majority of them access internet through low-end smart phones. To overcome such problems popular social media company Facebook introduced new application “Facebook lite” –a slimmed down version of Facebook less than One Megabyte (1 MB) in size that is faster, smaller, and lighter on data consumption targeting users from developing countries(Gibbs, 2015). Airbnb should consider similar strategies such as providing basic mobile version of their application in country like India with large mobile subscribers.

Notwithstanding the impact of attitude, trust emerged as the second most significant predictor of consumer intention to use sharing economy platform Airbnb. Consumers increasingly trust online reviews also popularly known as electronic word of mouth (eWOM) that allows them to evaluate and compare products/services through massive amount of user-generated content (UGC) available over the internet guiding them to make purchase decisions (Chevalier & Mayzlin, 2006; Ismagilova et al., 2017, 2019). Prospective travellers take the opinions and narratives provided in reviews by actual travellers into consideration before making travel arrangements including accommodation, underscoring the necessity of Airbnb to design review and trust building mechanisms on the platform. Though, Airbnb’s reciprocity review mechanism (guest and hosts rating each other) is criticized for high five star ratings (Ert et al., 2016). Bridges and Vásquez (2018) research on micro-analytic comparison of positive reviews found that reviewers communicate negative experiences using more nuance and subtle cues, potentially reducing uncertainty for future guest and hosts. Hence, Airbnb marketers in emerging economies like India should make necessary efforts to build the review mechanism effective and trustworthy to maximize the potential in the market.

6.3 Limitation and Future Research Directions

Readers should interpret findings of this research in light of following limitations. First, despite developing comprehensive research model, this study validated the model using cross sectional data that limits the generalizability of the results. In future, researchers should employ longitudinal data collection techniques to improve generalizability. Second, this study measured only consumer intention to use sharing economy platform Airbnb. Prior research had found that consumer intention not necessarily translate into usage behaviour of technology platforms. Thus, future research should measure consumer intention as well as actual use of platform (Wu & Du, 2012). Third, this study evaluated consumer adoption towards Airbnb focusing only on the attributes of platform. However, tourism and hospitality researchers had demonstrated impact of host attributes on the platform towards guests accommodation booking decisions (Ert et al., 2016). Future researchers should evaluate interactions among hosts and guest’s interaction on the platform in emerging economies. Despite these limitations, this comprehensive study on Indian consumer’s intention to use Airbnb offer number of implications for research and practice alike.

7 Conclusion

This study examined Indian consumer’s intention to use sharing economy platform Airbnb employing Dwivedi et al. (2019)‘s, meta- UTAUT model as theoretical lens and extended the model with Trust, self-efficacy, and Hedonic motivation constructs to be more relevant to consumer context. Second, focused on India that is set to become fourth largest travel & tourism economy in the world by 2027 (WTTC, 2017). Finally, empirically tested the model with 301 Indian consumers to evaluate their intention to use sharing economy platform Airbnb. The findings revealed attributes such as effort expectancy, social influence, and facilitating conditions as significant indirect determinant of consumer behavioural intention through attitude. However, attitude formation was not necessary for performance expectancy that emerged as significant direct predictor of consumer intention to use Airbnb platform. Besides attitude, self-efficacy, and trust emerged as significant different predictors of Indian consumer intention to use Airbnb explaining as much as 65% variance. In doing so, this study provides comprehensive research model to examine consumer adoption of sharing economy platform.