The Effects of Perceived Risk, Brand Credibility and Past Experience on Purchase Intention in the Airbnb Context

Abstract

This study investigated the main effects of perceived risks, brand credibility and past experience on intention to stay at Airbnb places and the interaction effects of past experience with perceived risks and brand credibility on intention to stay. A survey research method was utilized in South Korea. Our study is the first study to have applied and supported Jacoby and Kaplan’s six constructs in explaining consumers’ risk perceptions in the Airbnb context. One of the primary findings of this study was that past experience played a significant moderating role in the Airbnb context. Social risk and psychological risk had negative effects and performance risk had a positive effect on Airbnb users’ intention to stay. With regard to non-users, psychological risk had a negative effect on intention to stay. These findings indicate that Airbnb users consider more specific risks based on their prior experiences of Airbnb use. This study also revealed that brand credibility had the strongest influence on both Airbnb users’ and non-users’ intention to stay. Consistent with previous research, this finding indicates that brand credibility is a critical variable in consumer decision-making for the intangible and heterogeneous products like accommodation. Based on these findings, theoretical and managerial implications were provided, and limitations and future research were discussed.

1. Introduction

The emergence of peer-to-peer platforms has radically transformed the tourism and hospitality industry [1]. Information and communication technology (e.g., smartphones, Web 2.0, mobile technology) helps people easily share access to resources, such as transportation (i.e., ride shares), accommodation (i.e., short-term rentals) and food (i.e., peer-to-peer dining) with one another [2]. One of the most heavily discussed sectors of the peer-to-peer platforms is home-sharing lodging, and Airbnb is the leading platform in this sector [3].

Airbnb places can be found in more than 220 countries and 100,000 cities worldwide, representing seven million Airbnb listings with 750 million guests served in 2020 [4]. Results of market research suggest that consumers who have used Airbnb in the last three to five years have used it again [5]. Airbnb wields a strong influence on the tourism and hospitality industry. In 2017, the Airbnb company was already valued at $30 billion while the market capitalization of Hilton and Marriot was hovering around $20 billion and $34 billion respectively [6].

Despite the fast-growing popularity of Airbnb in the tourism and hospitality industry, there are concerns for guests. Airbnb places are not required to be inspected by the government for safety standards or by the Airbnb company for room cleanliness and facilities, while most hotels are inspected by governments, and they regularly hire private inspectors to retain public trust. Some Airbnb guests have claimed that a host did not show up, a host canceled the reservation shortly before check-in, an Airbnb listing did not match the description or lacked promised amenities/items, the property of a listing was not generally clean, and the number of bedrooms or bathrooms in the property did not match [7]. Recently, significant concerns related to safety, security and discrimination issues have emerged. For instance, an American man was sexually assaulted by his transsexual Airbnb host during a trip to Spain [8]. A Melbourne IT specialist lost $9000 on a fake listing of a Hawaiian villa [9]. An Asian American woman was turned away by an Airbnb host during snowstorm because of her race [10]. These incidents have stirred an international debate about Airbnb’s responsibilities as a company [11]. These horror stories should negatively influence potential customers’ perceptions about Airbnb. However, there has been little research conducted about consumers’ perceived risk and its impacts on Airbnb use.

Since Airbnb is a new type of accommodation, consumers may focus on potential risks that they may face if they use the platform. Several researchers have suggested that brand credibility is a central cue in the consumer decision-making process, especially with high-risk products, because perceived credibility towards a brand reduces perceived risk of consumption [12,13,14,15]. Researchers in travel and tourism have also found that prior knowledge, specifically past use experience, is one of the primary variables which influence consumers’ risk perceptions and their subsequent decision behavior [14,16]. Therefore, this study investigates the effects of consumers’ perceived risk, brand credibility and past experience on intention to stay at an Airbnb place. This study first identified dimensions of perceived risk. Secondly, it examined the main effects of perceived risk, brand credibility and past experience on intention. Finally, this study examined whether past experience moderated effects of perceived risks and brand credibility on intention.

2. Literature Review

2.1. Theoretical Background

Accommodation is often considered as a high-risk product due to intangibility and heterogeneity [17]. Intangibility implies that sets of performances of individuals’ experience at accommodation cannot be seen or touched in the same manner in which tangible goods can be sensed [18]. Heterogeneity portends the potential for high variability in the performance of accommodation [18,19]. For example, the quality and essence of a hotel can vary from producer to producer, from customer to customer, from destination to destination, and from day to day [19]. Therefore, consumers experience difficulties in evaluation of accommodation before they make a purchase decision and they have a higher level of risk perception in their decision-making [20].

To reduce consumers’ perceived risks caused by intangibility and heterogeneity, major hotel chains guarantee a familiar and standardized quality of service throughout the world. However, the Airbnb company is an online intermediary and it does not guarantee safety, legality and quality of each host’s services. In addition, the Airbnb company provides various types of rooms such as a shared room (e.g., couch surfing), host-resident private rooms (e.g., mini-bed and breakfast) and a host-absent entire place (e.g., vacation rental) [21]. Besides, Airbnb guests are not able to physically examine the Airbnb properties until they make the reservation due to the Airbnb policy. Therefore, consumers’ decision-making on an Airbnb place generates high uncertainty as to their outcomes and thus consumers’ perceived risk is more important in their decision-making [22,23].

When travelers make decisions, especially for high-risk products, they use information search and processing strategies for risk reduction [17,19]. The case-based vacation planning theory [24] suggests that cases obtained through the learning experiences are stored in memory, retrieved, modified, reused in similar contexts, and restored in memory for the future use. Through these repeating processes, travelers understand which information sources and attributes are useful in similar decision-making situations [25]. In decision-making regarding accommodation, people tend to value brands over various performance attributes in order to maximize ease of information justification [19]. Several researchers also suggest that people consider brand images in their decision-making on hotels as the risk reduction strategy [17,26]. Additionally, people often count on their memory formed by their actual use experience because their memory is trustworthy and credible [16,27,28,29,30]. In addition, past experience can offset consumers’ risk-reduction needs [17]. Based on those assumptions, a conceptual framework was suggested in Figure 1.

2.2. Perceived Risk

Travel risk is defined as the possibility of experiencing danger while traveling [31] or the consciousness of insecurity and knowledge of the likelihood of damage while traveling [32,33]. Park et al. [23] defined it as a consumer’s belief about the potential uncertainty associated with negative outcomes in a purchase situation. They also suggest that consumers’ perceived risk in the mobile shopping context is often caused due to discrepancies between consumers’ pre-purchase evaluation and actual product qualities.

It has been identified that perceived risk is a multidimensional construct [16,34]. According to Cunningham [35], risk has been conceptualized as involving two elements: uncertainty and consequences. The former one refers to the subjective possibility of loss if the results of an act were not favorable and the latter one refers to individual’s subjective feelings of certainty that the outcome will be unpleasant [35,36]. Jacoby and Kaplan [37] focused on studying about adverse consequences and suggested six different types of losses: performance, financial, physical, psychological, social, and time losses [38,39]. This study also applies Jacoby and Kaplan’s [37] six constructs to understand consumers’ perceived risks toward Airbnb use.

Performance risk is defined as the possibility of not getting what is expected or the service not performing as expected [40]. The performance risk is related to situations whereby experiences at the Airbnb places may not be consistent, the photos/locations/descriptions of the listings posted on the Airbnb platform do not match the ones in real life, the property of a listing was not generally clean, and the number of bedrooms or bathrooms in the property did not match the information provided on the platform [7,41,42]. These situations often happen to Airbnb customers because Airbnb has no control over the conduct of hosts [43] and there are no specific hospitality standards implemented among Airbnb listings [41,44].

Financial risk means the possibility of losing money or not attaining the best possible monetary gain [45]. In a study by Mitchell and Greatorex [26], financial risk turned out to be the most important risk for hotel services. Specifically, for services purchased online, financial risks were part of the top three risks perceived by consumers [46]. Although Airbnb listings are said to be cheaper than most hotels, some consumers still perceive a lack of cost savings [2]. In fact, Airbnb customers pay a guest service fee of 6–12% to the Airbnb company, a security deposit which is set by a host and a value-added tax if the customer stays in the European Union, Switzerland or Norway [47]. In addition, some of the Airbnb listings request a cleaning fee [47]. The Airbnb company also takes a three percent of service fee from the host for each reservation to cover the cost of processing the transaction and this service fee increases the cost and postentially the price for customers [47].

By applying Jacoby and Kaplan’s [37] definition, this study defines physical risk as the possibility that staying at an Airbnb place may not be safe and may lead to harm or injury. Compared to hotels or traditional bed and breakfasts, Airbnb listings are unregulated [48] and there seems to be no health and safety regulations that the Airbnb hosts have to follow [49]. Kosoff [43] also argues that Airbnb does not enforce specific safety standards and it is up to hosts to make sure they are providing a safe rental. Thus, security and personal safety is one of the concerns in the use of Airbnb [41,50]. As discussed earlier, several horror stories in the use of Airbnb have been distributed to potential customers through news media and social networking sites. In addition to cases like the sexual assault and the fake listing, there are more cases related to the physical risk. A host drugged his guests [51]. A man died in a freak accident during an Airbnb stay [43]. A Canadian woman died in an Airbnb listing in Taiwan because it had a broken heater which filled the room with poisonous gas [52].

Several researchers have suggested that an individual perceives social risk when he/she faces a choice situation to the extent that he/she believes his/her peers may evaluate them negatively due to a certain purchase they make [38,53]. Social risk in this study is defined as the possibility that staying at an Airbnb place will have a negative effect on how others think of the Airbnb use [37,45]. Based on the definition, it can be suggested that staying at an Airbnb place may entail social risk when a consumer’s family members, friends or peers do not approve of them using the said service. Pires et al. [46] found that social risk was the second most important risk perceived by consumers for both low and high-involvement services purchased online.

Based on Garner’s [45] definition, this study refers to psychological risk as the possibility that staying at an Airbnb place will have a negative effect on a person’s peace of mind. Psychological risk is dominant when consumers purchase a product which is expensive, complex and difficult to judge [54,55]. Pires et al. [46] also found that psychological risk was ranked the highest for services purchased online. As discussed earlier, Airbnb accommodation is a high-risk product due to heterogeneity and intangibility. In addition, Airbnb users should make purchases through its mobile application and pay in advance. In some cases, getting a refund is difficult or complicated. According to customer reviews on social networking sites [10,21], it seems that there is a chance to have a dishonest host who does not provide appropriate services as promised and on the contrary he/she files a false claim and/or leaves negative reviews about the customer on the Airbnb platform. All these incidents make customers feel disappointed with Airbnb. Moreover, potential customers who read reviews written by actual users of Airbnb may feel higher levels of psychological risk in using Airbnb.

Time risk means the possibility of wasting time, taking too much time and/or putting too much effort in using Airbnb to book accommodation [45]. Mitchell and Greatorex [26] found that for hotel services, time risk was the second most important risk for consumers. As the Airbnb provides a wide variety of accommodation choices, it may be time-consuming for consumers in order to look for offers that are trustworthy [50]. Additionally, since not all Airbnb listings have an instant book feature, it would require more time and effort to send messages to the host before being able to reserve a room [44]. To make a reservation of an Airbnb place, a potential guest needs to register, manage his/her profile with a picture of himself/herself and information about his/her interests, and send messages to the host, which includes an introduction of himself/herself, the purpose of the visit and a respectful intention of booking the host property [56]. The Airbnb hosts then decide whether to accept or reject the booking [1]. If the host does not accept the booking, the Airbnb users need to spend more time to repeat the same process again with a different host for a different Airbnb place.

Risk perception has been suggested as one of the most significant determinants of travelers’ purchase intention, especially in the online/mobile contexts [17,57,58]. According to Sonmez and Graefe [58], risk perception is the strongest predictor of safety concerns. Previous research has also confirmed that perceived risk directly negatively influences travelers’ choice of destination as well as their propensity to visit or avoid certain destinations [16,58]. Accordingly, the following hypothesis was proposed:

Hypothesis 1. 

Perceived risks are negatively associated with intention to stay at an Airbnb place.

2.3. Brand Credibility

Erdem and Swait [59] first defined brand credibility as the believability of the product position information contained in a brand. Erdem and Swait [59] and Erdem et al. [12] suggest that brand credibility entails the consistent delivery of what is promised. According to Erdem and Swait [59], the concept of credibility includes two main dimensions, trustworthiness and expertise, and trustworthiness affects consumer choice and brand consideration more than expertise. For this reason, this study focuses on trustworthiness.

In travel and tourism research, brand credibility has often been examined as a critical variable in study of travelers’ decision-making behavior [14,60]. For example, when consumers make a decision about accommodation, which is a high-risk product, they cannot readily evaluate service quality until they actually consume the services and therefore they put more effort on judging levels of brand credibility rather than judging various service attributes of each brand [19].

Importance of brand credibility can be explained by the signaling theory that a brand can serve as credible signals of the unobservable product quality because it symbolizes the cumulative efforts of a supplier’s continuous marketing communication strategies [60,61,62]. The signaling theory is based on the assumption that a different level of product information which flows between consumers and suppliers causes a problem of information asymmetry and this information asymmetry causes consumer uncertainty about the quality of services [60,61,63]. In this circumstance, a credible brand can reduce difficulty in decision-making by diminishing the cognitive effort required to evaluate a service provider, thus increasing the possibility of purchase intention [60]. Wang and Yang [64] found that brand credibility had direct positive effects on purchase intention. Baek et al. [61] also found that brand credibility positively influenced brand purchase intention. Based on the prior research, the following hypothesis was proposed:

Hypothesis 2. 

Brand credibility is positively associated with intention to stay at an Airbnb place.

2.4. Past Experience

Prior knowledge plays various roles in affecting consumers’ risk perception, brand credibility and their subsequent decision behaviors [14,16]. According to Kerstetter and Cho [14], prior knowledge includes three dimensions: familiarity, past experience and expertise. Our study focuses on past experience (i.e., prior Airbnb use experience) because it is one of the strongest variables influencing consumers’ decision-making process, especially when consumers purchase high-risk travel products [14,65]. Raju and Reilly [66] suggest that travelers are likely to make a biased choice based on their past visit experiences. Dependency on past experience in decision-making can be explained by the fact that people first recall past experiences when making travel decision as they trust their own memories the most [19]. Therefore, we suggest the following hypothesis:

Hypothesis 3. 

Past experience is positively associated with intention to stay at an Airbnb place.

Several studies have suggested that previous travel experience moderates the effects of risk perception or brand credibility on purchase intention [29,30,67]. Sharifpour et al. [16] also suggest that there is a significant interaction effect of past experience and risk perception on intention to visit. With regard to risk perceptions, more experienced travelers perceive less risk on certain risk factors [16]. Fuchs and Reichel [17] revealed differences among first-time and repeat visitors in relation to the types of risks they perceived [16]. For example, first-time visitors were more concerned about human-induced risks, psychological risk, food safety and weather risk, whereas repeat visitors were more concerned with financial risk, performance risk related to the service quality and physical risk. In addition, it is assumed that the relationship between brand credibility and consumers’ purchase intention may vary in terms of users/non-users [12]. A few studies have also implied that levels of brand credibility differ between first-time users and repeat users and thus exhibit different purchasing behaviors [12,14]. Airbnb users may have confidence in booking the Airbnb places again because they rely on retained prior knowledge formed by their actual experiences of using Airbnb. As discussed earlier, brand credibility is formed as a result of a supplier’s continuous marketing communication. Therefore, past experience may have a stronger effect than brand credibility for Airbnb users. However, nonusers feel uncertain about staying at Aribnb places, so credibility of Airbnb may act as key information in their purchase decision-making [27,68,69]. Kerstetter and Cho [14] also suggest that first-timers tend to use commercial information sources to make purchase decisions. Therefore, we assume that past experience of Airbnb moderates the effects of risk perceptions or brand credibility on intention.

Hypothesis 4. 

There are significant interaction effects of perceived risks and past experience on intention to stay at an Airbnb place.

Hypothesis 5. 

There is a significant interaction effect of brand credibility and past experience on intention to stay at an Airbnb place.

3. Methods

The study population was individuals from South Korea who recognized the Airbnb brand and/or had stayed at Airbnb places. An online survey research method was utilized for this study. The sampling frame was 96,600 panel members of a panel company in South Korea. By applying the nonprobability convenience sampling method, the panel members were invited to participate in the study through an email with the online survey link. They survey link was distributed in October 2017 for a week. As the screening question, the study participants were asked if they recognized Airbnb and only those who clicked yes were able to move to the next question, which asked if they had stayed at an Airbnb place at least once. The study participants who had not stayed but recognized Airbnb or those who had stayed at an Airbnb place met the sample requirements and were able to complete all of the questionnaires. The panel members who completed the survey received certain points which would be converted into money.

In this study, past experience was defined as prior Airbnb use experience. In terms of the past experience, the study participants who had not stayed at an Airbnb place, but recognized Airbnb were treated as the Airbnb non-users and coded as “0” and those who had stayed at an Airbnb place at least once were treated as the Airbnb users and coded as “1”. The survey instrument consisted of four sections: perceived risk, brand credibility, intention and demographic profiles. The measurement items were adopted from previous research and most of them were modified in order to fit in the Airbnb context. Perceived risk consisted of six dimensions with 21 items. Four items to measure performance risk were taken from Finley [41], Guttentag [44] and Jung [42]. Four items to measure financial risk were taken from Ha [70] and Kim et al. [71]. Three items for physical risk were taken from Jose [51], Lieber [72] and Stone [48]. Three items for social risk and three items for psychological risk were taken from Kim et al. [54]. Four items to measure time risk were taken from Guttentag [44] and Kim et al. [54]. Five items to measure brand credibility were taken from Ayeh et al. [73] (2013). Three items to measure intention to stay at an Airbnb place were taken from Cronin et al. [74]. All items were measured by using a 7-point Likert-type scale in which “1” means strongly disagree and “7” means strongly agree.

4. Results

Among the 20,600 invited panel members, 5065 attempted to respond to the survey. Screening questions were asked for identifying if the respondent recognized about the Airbnb brand or if the respondent had stayed at an Airbnb place. Then, 759 panel members were qualified to continue to respond to the survey and 737 members completed the survey. A single case was excluded because the respondents had identical Internet Protocol addresses. Out of 736 data, 415 were from the Airbnb users (56.39%) and 321 were from non-users (43.61%).

The data collected were analyzed using SPSS 22.0. In terms of the Airbnb users, about 65.3% of respondents were female (

Table 1. Demographic Profile of Respondents.

	Total (n = 736)		Airbnb Users (n = 415)		Airbnb Non-Users (n = 321)
Personal Characteristics	Frequency	%	Frequency	%	Frequency	%
Gender
Male	260	35.3	144	34.7	116	36.1
Female	476	64.7	271	65.3	205	63.9
Age
20–29	276	37.5	184	44.3	92	28.7
30–39	243	33.0	126	30.4	117	36.4
40–49	172	23.4	80	19.3	92	28.7
50–59	45	6.1	25	6.0	20	6.2
Level of Education
High school graduate	83	11.3	46	11.1	37	11.5
University/college graduate	547	74.3	292	70.4	255	79.4
Masters or Ph.D. graduate	102	13.9	75	18.1	27	8.4
Others	4	0.5	2	0.5	2	0.6
Monthly Household Income
–$1749	60	8.2	25	6.0	35	10.9
$1750–$3499	215	29.2	105	25.3	110	34.3
$3500–$5249	223	30.3	106	25.5	117	36.4
$5250–$6999	132	17.9	95	22.9	37	11.5
$7000–$8749	73	9.9	54	13.0	19	5.9
$8750	33	4.5	30	7.2	3	0.9
Place of Residence
Seoul	298	40.5	187	45.1	111	34.6
Incheon, Gyeonggido	213	28.9	113	27.2	100	31.2
Daejeon, Chungcheongdo, Gangwondo	55	7.5	18	4.3	37	11.5
Daegu, Gyeongsangbukdo	50	6.8	30	7.2	20	6.2
Busan, Gyeongsangnamdo	85	11.5	50	12.0	35	10.9
Gwangju, Jeollado, Jejudo	35	4.8	17	4.1	18	5.6

). Slightly less than half (44.3%) of respondents were between ages of 20–29 and approximately one-third (30.4%) were between ages of 30–39. Most respondents had a bachelor’s degree (70.4%) and 18.1% of respondents had a master’s degree or Ph.D. About one-fourth of respondents (25.3%) had monthly household income between US $1750 and $3499, another one-fourth (25.5%) had monthly household income between $3500 and $5249, and 22.9% had monthly household income between $5250 and $6999. Slightly less than half (45.1%) of respondents lived in Seoul. In terms of non-users, 63.9% of respondents were female. About two-fifth of respondents (36.4%) were between ages of 30–39. Most respondents had a bachelor’s degree (79.4%). About one-third of respondents (34.3%) had monthly household income between US $1750 and $3499 and 36.4% of respondents had between $3500 and $5249. In terms of residence, 34.6% of respondents lived in Seoul and 31.2% lived in Incheon or Gyeonggi-do.

The results of factor analysis and reliability analysis indicated six factors with 21 items for the perceived risk (

Table 2. Results of Factor Analysis and Reliability Analysis for Perceived Risk.

	Factor Loadings	Eigen-Value	Variance (%)	Cronbach’s Alpha
Performance Risk		3.472	16.535	0.903
I worry that the Airbnb place/listing will not match the photos posted online	0.888
I worry that the Airbnb place/listing will not match the descriptions posted online	0.864
I worry that the Airbnb place/listing would not be clean	0.825
I am concerned that the Airbnb host would treat me unkindly	0.670
Financial Risk		2.960	14.093	0.889
I am concerned about whether an Airbnb place is more expensive than hotel rooms in the same area	0.867
I am concerned that an Airbnb place is overpriced considering the quality	0.818
I am concerned if the price of an accommodation on the Airbnb website is more expensive compared to other travel websites	0.759
I am concerned that staying at an Airbnb place could involve financial losses	0.613
Physical Risk		2.421	11.531	0.924
I am concerned that staying at an Airbnb place/listing would lead to something bad happening to me	0.837
I worry that the Airbnb host may do something bad to me	0.790
I am concerned that it may not be safe to stay at an Airbnb place/listing	0.749
Social Risk		2.726	12.983	0.940
Staying at an Airbnb place/listing will adversely affect others’ opinion of me	0.852
I would be thought of as foolish by people whose opinion I value if I stay at an Airbnb place/listing	0.830
The thought of staying at an Airbnb place/listing causes me concern because some friends would think I was just being showy	0.823
Time Risk		3.468	16.513	0.944
Using Airbnb to book accommodation takes a lot of effort	0.867
Using Airbnb to book accommodation will be a waste of time	0.803
Using Airbnb to book accommodation will take too much time	0.800
Using Airbnb could lead to an inefficient use of my time as I have to send messages to the host	0.794
Psychological Risk		2.644	12.591	0.936
The thought of staying at an Airbnb place/listing makes me feel psychologically uncomfortable	0.837
The thought of staying at an Airbnb place/listing gives me a feeling of unwanted anxiety	0.818
The thought of staying at an Airbnb place/listing causes me to experience unnecessary tension	0.815

KMO = 0.917; Bartlett’s χ² = 14,949.574, p < 0.001; Total Variance = 84.245%.

). The Kaiser-Meyer-Olkin (KMO) value was 0.917, Bartlett’s tests were significant at p < 0.001 and the total variance extracted by the factors was 84.245%. All factor loading scores were greater than 0.613 after an orthogonal rotation using the Varimax method, and the Eigenvalues were all greater than 2.421. The Cronbach’s Alpha score were 0.889 and higher.

Table 3. Results of Factor Analysis and Reliability Analysis for Brand Credibility.

	Factor Loadings	Eigen-Value	Variance (%)	Cronbach’s Alpha
Airbnb is trustworthy	0.933	4.241	84.829	0.955
Airbnb is reliable	0.932
Airbnb is sincere	0.922
Airbnb is honest	0.918
Airbnb is dependable	0.899

KMO = 0.906; Bartlett’s χ² = 3906.287, p < 0.001; Total Variance = 84.829%.

presents results of factor analysis and reliability analysis for brand credibility. The KMO value was 0.906, Bartlett’s tests were significant at p < 0.001, the total variance extracted by the factors was 84.829%, all factor loading scores were greater than 0.899, the Eigenvalue was 4.241 and the Cronbach’s Alpha score was 0.955. In terms of intention, the KMO value was 0.762, Bartlett’s tests were significant at p < 0.001, the total variance extracted by the factors was 86.918%, all factor loading scores were greater than 0.924, the Eigenvalue was 2.608 and the Cronbach’s Alpha score was 0.924 (

Table 4. Results of Factor Analysis and Reliability Analysis for Intention.

	Factor Loadings	Eigen-Value	Variance (%)	Cronbach’s Alpha
I will stay at Airbnb places.	0.938	2.608	86.918	0.924
I will recommend my friends to stay at Airbnb places.	0.935
When I make decision on accommodation, I will choose Airbnb places.	0.924

KMO = 0.762; Bartlett’s χ² = 1694.420, p < 0.001; Total Variance = 86.918%.

). All results above indicated that the questionnaire used in this study reached relatively good construct validity.

Before conducting the multiple regression analysis, the mean centering method was applied because it made results of the interaction effects slightly more meaningful and easier to interpret. Independent variables were mean centered to avoid the potential of multicollinearity in the interaction terms [75]. As the results, the VIF values were less than 3.00 (VIF_{performance risk (PR)} = 2.367, VIF _{financial risk (FR)} = 2.109, VIF_{physical risk (PhR)} = 2.360, VIF_{social risk (SR)} = 2.378, VIF_{time risk (TR)} = 2.552, VIF_{psychological risk (PsR)} = 2.515, VIF_{brand credibility (BC)} = 1.497, VIF_{past experience (PE)} = 1.292, VIF_PRxPE = 2.231, VIF_FRxPE = 1.992, VIF_PhR×PE = 2.252, VIF_SR×PE = 2.219, VIF_TR×PE = 2.355, VIF_PsR×PE = 2.307, VIF_BC×PE = 1.273. All these results indicate that multicollinearity is not a potential problem in the data set.

Multiple regression analysis was conducted to examine main effects and interaction effects of independent variables on intention (

Table 5. Results of Multiple Regression Analysis on Intention.

	Total		Airbnb Users		Airbnb Non-Users
	Beta	t	Beta	t	Beta	t
Performance Risk (PR)	0.085	2.662 **	0.170	3.641 ***	0.016	0.320
Financial Risk (FR)	−0.002	−0.068	−0.075	−1.595	0.062	1.389
Physical Risk (PhR)	0.080	2.522 *	0.083	1.787	0.093	1.774
Social Risk (SR)	−0.093	−2.903 **	−0.173	−3.189 **	−0.033	−0.721
Time Risk (TR)	−0.057	−1.727	−0.070	−1.244	−0.056	−1.205
Psychological Risk (PsR)	−0.153	−4.653 ***	−0.115	−2.357 *	−0.221	−4.246 ***
Brand Credibility (BC)	0.706	27.792 ***	0.737	21.538 ***	0.706	17.868 ***
Past Experience (PE)	0.162	6.854 ***
PR × PE	0.057	1.829
FR × PE	−0.061	−2.074 *
PhR × PE	−0.012	−0.376
SR × PE	−0.051	−1.656
TR × PE	0.001	0.040
PsR × PE	0.058	1.826
BC × PE	−0.046	−1.982 *
Model Summary
Adjusted R2	0.683		0.595		0.597
ANOVA Regression F	106.632 ***		87.779 ***		68.586 ***

Note: * p < 0.05; ** p < 0.01; *** p < 0.001.

). The model was significant (F = 106.632 at p < 0.001) and R²_adj was 0.690. Performance risk (β = 0.085, p < 0.01), physical risk (β = 0.080, p < 0.05), brand credibility (β = 0.706, p < 0.001) and past experience (β = 0.162, p < 0.001) were significantly positively associated with intention. Social risk (β = −0.093, p < 0.01) and psychological risk (β = −0.153, p < 0.001) were significantly negatively associated with intention. Brand credibility had the strongest effect, followed by past experience and psychological risk. There were significant negative interaction effects of FR × PE (β = −0.061, p < 0.05) and BC × PE (β = −0.046, p < 0.05). The study results supported Hypothesis 2 and Hypothesis 3 and partially supported Hypotheses 1 and 4.

Because the significant interaction effects were found, further analyses were conducted. The data were separated into two groups: the Airbnb user group and the Airbnb non-user group. The multiple regression analysis was conducted with each of these two groups. In terms of the Airbnb user group, the model was significant (F = 87.779 at p < 0.001) and R²_adj was 0.602. Brand credibility (β = 0.737, p < 0.001) and performance risk (β = 0.170, p < 0.01) were significantly positively associated with intention and social risk (β = −0.173, p < 0.01) and psychological risk (β = −0.115, p < 0.05) were significantly negatively associated with intention. Brand credibility had the strongest effect, followed by social risk, performance risk and psychological risk. In terms of the nonuser group, the model was significant (F = 68.586 at p < 0.001) and R²_adj was 0.605. There are only two factors which were statistically significant. Brand credibility (β = 0.706, p < 0.001) was significantly positively associated with intention and psychological risk (β = −0.221, p < 0.001) was significantly negatively associated with intention. Brand credibility had the strongest effect, followed by psychological risk.

5. Conclusions

5.1. Discussion

This study investigated effects of consumers’ perceived risk, brand credibility and past experience on intention to stay at an Airbnb place. Based on the literature reviewed, this study identified six dimensions of perceived risk: performance risk, financial risk, physical risk, social risk, time risk and psychological risk. It also examined main effects of perceived risks, brand credibility and past experience on intention to stay and interaction effects of past experience with perceived risks and brand credibility on intention to stay.

The results of this study support the finding of previous studies [16,34] that perceived risk is a multidimensional construct. Our study results also support Jacoby and Kaplan’s [37] six-constructs model of perceived risk. There were a few studies which examined risk perceptions in the Airbnb context [76,77]; however, those studies did not apply Jacoby and Kaplan’s [37] entire six constructs. Our study was the first study which applied and supported the whole six constructs in explaining consumers’ risk perceptions at the Airbnb context.

The initial assumption of this study was that consumers’ risk perception negatively affected intention to stay at Airbnb places. The study results indicate that social and psychological risks of Airbnb negatively affect intention to stay. However, performance and physical risks of Airbnb are positively associated with intention to stay. Recently, Yi et al. [77] studied about effects of perceived risks in the Airbnb context and they also found that physical and performance risks were positively associated with behavioral intention and desire to stay at Airbnb places. Their results and our results in the Airbnb context can be explained by the unique characteristics of Airbnb, heterogeneity. The Airbnb company is an online intermediary platform which matchmakes between hosts and consumers and it introduces various types of rooms provided by the hosts to consumers. It appears that consumers understand the variety of Airbnb places and thus they are likely to tolerate potential physical and performance risks they may face when they make purchasing decisions on the Airbnb places. Additionally, these results can be explained by the unique characteristics of travel motivations. Some travelers consider new ways to travel and unique local experiences as a core part of their identity and these travelers are core customers of Airbnb [78]. When they seek new ways of making their trip, the trip increases the level of risk, but those travelers are still willing to make risky choices because they know the fun inside the risk [77].

One of the primary findings of this study was that past experience played a significant moderating role and this study was the first which examined the moderating role of past experience in the Airbnb context. Additionally, this study suggests when the moderating effects of past experience are investigated especially in the Airbnb context, risk perceptions should be examined as multidimensional constructs. Our study findings are consistent with the findings of previous research that visitors differ from non-visitors in terms of considering certain risks in their purchase decision-making [16,17]. According to our study, social risk and psychological risk have negative effects and brand credibility and performance risk have positive effects on Airbnb users’ intention to stay. With regard to non-users, brand credibility has a positive effect and psychological risk has negative effect on intention to stay. These findings indicate that the Airbnb users consider more specific risks based on their prior experiences of Airbnb use while non-users consider the psychological risk in their decision making. It appears that the Airbnb users understand that Airbnb places are heterogeneous across locations; therefore, they deliberate various specific risks. Several travel researchers have also suggested that experienced travelers are more critical of specific types of information and consider various types of risks in their decision-making [14,17].

This study revealed that brand credibility had the strongest influence on intention to stay at Airbnb places. This finding is consistent with previous research which revealed that brand credibility is a critical variable in study of travel decision-making [14,60]. Based on the case-based vacation planning theory [24,25], Jun et al. [19] states that consumers cannot readily evaluate service quality of accommodation until they actually consume the services and thus they put more effort on judging levels of brand credibility rather than judging various performance attributes of each brand in order to maximize ease of information justification. Our results can also be explained by the signaling theory. According to this theory, a credible brand reduces the difficulty of making decisions especially for intangible and heterogeneous products, such as accommodation, and thus the brand credibility increases purchase intention [60]. The theory suggests that a brand can serve as credible signals of the unobservable service quality because it symbolizes the cumulative efforts of a supplier’s continuous marketing communication strategies [62].

Given the findings of the current study, practical implications are suggested. The Airbnb company needs to enhance its brand credibility. Considering that brand credibility has the strongest influence on intention to stay at Airbnb places, it is recommended that enhancing brand credibility should be the priority in its marketing strategy. Heterogeneity is one of the unique characteristics of Airbnb. Consumers use Airbnb to find a place to stay instead of finding a hotel because Airbnb provides various options to choose from to fit their budget, preferences and desired locations [78]. The heterogeneity of Airbnb, however, increases consumers’ perceived risks. Therefore, consumers count on the brand, Airbnb, to be there for them when they face risky situations. For this reason, brand credibility of Airbnb is very important and meaningful compared to other accommodation. To make the brand trustworthy, the Airbnb company should establish stronger policy to solve customers’ problems. According to iProperty Management [79], 82% of people who had a problem with their Airbnb stay mentioned that they had a problem with Airbnb’s customer service and 57% of those who mentioned poor customer service also said that it was their primary complaint. Providing high-quality customer service related to guest safety (e.g., scams, unsafe conditions, fake listings, discrimination, sexual abuse) is one of the most urgent ways to gain its customer’s trust and boost its brand credibility. This implication can be applied to other intermediary platforms when the services are especially intangible and heterogeneous. In addition, managers of boutique hotels or bed and breakfasts may consider this implication due to their heterogeneous characteristics [19].

It should be noted that consumers’ perceived risk is a crucial determinant in their decision making, especially for intangible and heterogeneous products. The Airbnb company should investigate the effects of particular perceived risks rather than examining overall perceived risk, and establish means to mitigate and alleviate those risks. The Airbnb company should be aware that Airbnb users commit a detailed assessment of the various risks in their decision making. Therefore, marketing managers needs to have delicate communication strategies with Airbnb users. Both Airbnb users and non-users are concerned about situations like that staying at an Airbnb place makes them feel psychological discomfort, unwanted anxiety and unnecessary tension. The Airbnb company needs to provide guarantees to its guests that the Airbnb guests will be taken care of by the company in case something happens during their stay. The referral marketing and enhancing positive word-of-mouth from the Airbnb users will make potential users to alleviate their psychological risk [80]. To make Airbnb’s intangible products tangible, marketing managers may encourage Airbnb users to share photos of the Airbnb place while they are staying. Photos of an Airbnb place shared by the consumers are more trustworthy than those shared by the hosts. To reduce consumers’ perceived risks caused by heterogeneity, information frequency will be essential as consumers require greater amount of information in their decision making to reduce their perceived risk [80].

5.2. Limitations and Future Research

This study has the following limitations. First, our study sample showed an over-representation of females and the 20 to 39 age group. Interpreting and generalizing our results to other samples should be performed with caution.

Second, the data were collected in South Korea and it only represented the views of South Korean consumers. Airbnb is not popular in South Korea compared to other countries like USA, France, Italy, UK and Spain. Future research should replicate this study by using samples from other countries or other nationalities. In addition, it will have interesting results if we can compare effects of perceived risks between countries where Airbnb is popular and where Airbnb is not yet popular. In addition, people can stay at Airbnb places in more than 220 countries in the world. This fact may lead to comparisons of cultural differences. Specifically, by comparing effects of perceived social risk within various cultural groups can be more interesting results could be found. For example, as Airbnb is not popular in South Korea, Korean consumers appear to believe that their peers may evaluate them negatively due to the Airbnb use; therefore, their perceived social risk significantly negatively affect their intention to stay at Airbnb places. People in other countries where Airbnb is popular may not consider perceived social risk as important for them.

Third, this study appears to show that when Airbnb users are more concerned about performance risk, they are more likely to intend to stay at Airbnb places. As discussed above, this may be explained by the heterogeneity of Airbnb. However, this should be reinvestigated with various samples because it can be argued that the conclusions derived from the study are country-specific. In addition, it was found that the Airbnb users considered more specific risks based on their prior experiences while non-users considered the psychological risk only in their decision making. Even though this study discussed about those findings with similar results of the previous research, the results might be different by consumer type (e.g., if the consumers are frequent travelers or not, if the Airbnb users stayed at an Airbnb place in Korea or abroad). It is recommended that a replication of this study should be conducted with different types of countries or different types of consumers to verify the generalizability of the results.

Finally, the data for this study were collected before the coronavirus (COVID-19) outbreak emerged. Therefore, the study results on risk perceptions, brand credibility and intention to stay at Airbnb places might be different if data were collected now. Until a vaccine or remedy for COVID-19 is found, consumers’ major concern for decision-making regarding accommodation will be about cleanliness and hygiene of accommodation and the flexible cancellation policy the accommodation suppliers provide. For future research, we need to add additional risk factors related to those issues. Additionally, future research should investigate perceived health risk and travel anxiety as other significant variables which affect intention to stay.