MOVING FROM STRANGER TO “I TRUST YOU”

(1)

1

MOVING FROM STRANGER

TO “I TRUST YOU”

Explaining inter-personal trust development in the

sharing economy – The case of Airbnb

Elisabeth Pfeffer

(S2712989/B1070827)

Submitted in support of the degree

DDM Advanced International Business Management and

Marketing

Supervisors:

Dr. Jiyoung Shin (University of Groningen)

Professor Klaus Schoefer (Newcastle University)

(2)

i

Abstract

Trust is an essential component in making peer-to-peer sharing economy models work. The research aim is to investigate how inter-personal trust can develop in the online and offline environment of the sharing economy over time. The research design focusses on a qualitative analysis with semi-structured interviews of both Airbnb guests and hosts.

The key findings of this research are that guests and hosts slightly differ in their approach to trust development. Whereas structural assurances help guests to build initial trust for their hosts, hosts build initial trust through a general tendency to trust people. Additionally, inter-personal trust develops in stages while the medium used helps to move from one stage of trust to another. Additionally, culture, perceived risks, and experience were found to have an impact on inter-personal trust development.

(3)

ii

ACKNOWLEDGEMENTS

I would like to express my deep gratitude to my Groningen supervisor Dr. Jiyoung Shin and my Newcastle supervisor Professor Klaus Schoefer, for their guidance, positive encouragement and constructive criticism of this research work. They have helped to steer me in the right direction while allowing me to develop this research as my own work. I would also like to thank Dr. Martijn Keizer from the University of Groningen for his advice with my research design section. Furthermore, I would like to extend my grateful thanks to the participants in this study. I appreciate the time and effort that the interviewed Airbnb hosts and guests have invested in helping me with my research. Especially, Airbnb hosts have shown me much insight in how Airbnb is run from a host perspective and enabled me with new insights that I could have not gathered from a guest perspective alone.

Lastly, I would like to thank my boyfriend, as well as my friends and family for their moral support and encouragement throughout my study. I would have not been able to accomplish this research without them. Thank you!

(4)

iii

1. Introduction and background

Sharing, a century-old practice (Sahlins, 1972), has recently experienced a 21st_-century

(6)

2 which trust mechanisms may be important at various stages of trust development. It might be the case that trust changes over time due to the initial online context being moved to a facto-face offline context at a later stage, which does not occur in more traditional e-commerce settings. Additionally, communication methods used throughout the interaction between service provider and consumer might influence trust development.

Hence, this research provides ample opportunity to better understand the concept of trust in the sharing economy. Especially since trust represents the inner workings of the sharing economy, it is important to gain a better understanding of how trust develops in this context. Moreover, it has the potential to reveal differences to concepts of online trust in e-commerce. Overall, this research’s purpose is to is to investigate how inter-personal trust can develop in the online and offline environment of the sharing economy over time. The phenomenon of the sharing economy will be understood better by placing it into the context of the largest accommodation provider in the sharing economy, namely Airbnb and has thus been chosen as a case example for this research.

(7)

3

2. Airbnb and the international business environment

Airbnb is an international business, which provides over 5 million listings in 191 countries worldwide (Airbnb, n.d.). As other businesses of the sharing economy, it connects guests (renting the accommodation) and hosts (proving the accommodation) online via a standardized internet platform. Both parties are usually located in different countries, resulting in distance and thus increased uncertainty between guests and hosts. Two types of distance, geographic and psychic distance, can be attributed to this process.

Geographic distance describes the actual amount of distance between two locations. Since in the majority of cases1_{, both the host and guest are not located in the same country,}

information asymmetries arise, which might hamper trust establishment. For example, the location cannot be visited in advance, meaning that only the host knows the full details about the accommodation. In fact, the host could deceive guests with beautiful pictures of the accommodation but when the guests arrive, the apartment could look very different. On the flipside, the host also faces information asymmetries since he/she does not know the guest personally and whether he/she might damage their property upon arrival.

In addition to geographical distance hampering potential trust establishment between guest and host, psychic distance further complicates the guest-host interaction. Psychic distance can be defined as “the sum of factors preventing the flow of information from and to the market. Examples are differences in language, education, business practices, culture, and industrial development” (Johanson and Vahlne, 1977, p.24). Psychic distance results due to operating within multiple institutional environments, which are composed of different

1_{According to Joe Zadeh, vice president of product, two thirds of Airbnb bookings involves crossing}

(8)

4 regulatory, normative and cognitive forces (Scott, 1995). In turn, difficulties may arise due to these “rules of the game” being played differently in each host country. For example, Airbnb and their hosts have been penalized due to not adhering to local housing regulations, which restrict short-term rentals in cities such as Barcelona and New York, whereas cities such as Amsterdam have enabled Airbnb policies that allow short-term rentals (The Guardian, 2014). Differing regulations make it difficult for guests to know whether they are renting an illegal property. In turn, they might be thrown out of their apartment at any point of their stay or their insurance might not pay in case an accident occurs on site. Additionally, hosts also have to inform themselves about differing regulations and how to protect their house from damage, which depends on national as well as local housing regulations.

Moreover, normative forces such as culture play a role. According to similarity attraction- (Byrne, 1971) and social categorization theory (Tajfel, 1982), members of one’s in-group (i.e. same nationality, same language) are viewed in a more positive light whereas members of one’s out-group (different nationality, different language) are often associated with mistrust and suspicion (McAllister, 1995). Hence, certain biases against an out-group can complicate the establishment of trust between guest and host. In addition, cultural misunderstandings can occur either due to a lack of language commonality but also since different cultures may have different ways of interpreting messages online or have different expectations when they meet the host/guest.

(9)

5 the sharing economy, the next section will summarize the main findings on trust related concepts and identify the current research gaps surrounding trust in the sharing economy.

2. Literature review

Literature on trust in the sharing economy has been relatively scarce (Huurne et al., 2017). Some authors have created potential frameworks for investigating trust in the sharing economy by drawing on trust theories in the online context. For example, Hawlitschek et al. (2016) identified that trust in the sharing economy depends on three interrelated trust elements namely trust in the platform, trust in the peer and trust in the product. Furthermore, Möhlmann (2015) investigated drivers that lead to increased satisfaction of using the sharing economy and found that trust was one of the main predictors.

Most existing research has focussed on trust antecedents explaining initial engagement in the sharing economy. Especially, aspects that increase host trustworthiness online have been the main focus (Huurne et al., 2017). For example, while Abranova, Shavanova, and Fuhrer (2015) investigated the most effective response strategies when receiving negative reviews of guests, Ert, Fleischer, and Magen (2016) found that host’s pictures subconsciously influence perceived trustworthiness of the host. Additionally, Möhlmann (2016) found that reviews positively impact trustworthiness of the platform provider and that people trust established peer-to-peer platforms such as eBay more than more recent established platforms, e.g. Airbnb. Verification and signaling were also found to influence perceived trust. Hence, if hosts provided others with e-mail addresses, phone numbers or were integrated with social media accounts, they seemed more trustworthy (Teubner and Hawlitschek, 2017).

(10)

6 et al., 2015; Kim et al., 2015; Matzner et al., 2015; Mittendorf 2016). Also, the analysis of trust antecedents in the sharing economy has relied on quantitative approaches (Huurne et al., 2017), which is contrary to trust research in other online environments where various research methods such as case studies, interviews, experiments, and econometric analysis were employed (Gefen et al., 2008). Recent reviews point to the need of exploring the concepts of trust in the sharing economy in a qualitative manner to complement quantitative research (Huurne et al., 2017). In fact, qualitative research can help researchers uncover how risks in online transactions are perceived and evaluated and to find further relationships between trust and trust-related concepts (Grabner-Kräuter and Kaluscha, 2003). Furthermore, qualitative research can also help solve the problem of generic questions in quantitative research such as “Are renters in general reliable?”, which do not give room to evaluate “the type of action or the stakes that respondents have in mind” when making trust assessments (Ermisch et al., 2009, p.2).

(11)

7 have only very stereotypical information about the host such as a reputation score and a picture. Over time the richness of the medium might intensify by communicating via online/offline messaging after the initial booking and by eventually meeting face to face. Furthermore, Airbnb might influence trust development at different points in time through structural assurances such as guarantees.

Hence, this research paper will address the lacking literature of how trust develops in the sharing economy over time, in order to gain a better understanding of how trust can be facilitated at different points in time, online and offline (Grabner-Kräuter and Kaluscha, 2003). By doing so Luo and Zhang’s (2016) initial concept will be expanded to the for-profit sector of the sharing economy. Additionally, medium richness and social presence theory will be used to understand whether trust differs at different points in time due to the medium used. Lastly, additional mechanisms for trust development at different stages will be explored, highlighting the exploratory nature of this research. This paper especially seeks to analyse how trust can be established in the complex international environment as pointed out in paragraph 2. Thus, the three main research questions emerging are:

RQ1: How does inter-personal trust develop over time in the international context of for-profit

sharing economy platforms?

RQ1a: How does the medium used to interact between host and consumer help or hinder trust

development in the international context?

RQ1b: Which mechanisms support inter-personal trust development in the international

context of the sharing economy?

(12)

8

3. Proposition development

3.1 Inter-personal trust development

Inter-personal trust can be defined as “an expectancy held by an individual or a group that the word, promise, verbal, or written statement of another individual or group can be relied on” (Rotter, 1971, p. 444). According to research, inter-personal relationships develop over time. Lewicki and Bunker (1995) conceptualized that inter-personal trust develops in three stages (see Figure 1). The first stage is called calculus-based trust and involves people performing cost-benefit calculations about positive versus negative outcomes when trusting others. This kind of trust is rational and focusses on a transactional view of trust (Williamson, 1975, 1981). The authors point out that personal reputation would be one reason not to violate trust at this stage. Barber (1983) explains that in the stage of calculus-based trust, people may draw inferences of whether a party is trustworthy due to credible information about a person’s competence or intentions. In the case of Airbnb this could mean that when looking through profile information and reviews of the guest/host, people trust each other’s reputation2_{because both parties do not want to forgo a good reputation. Hence, control is of}

importance in calculus-based trust since guests and hosts know that they can be punished in the form of bad reviews, making it difficult for them to transact once again on the platform.

The second stage is called knowledge-based trust and relies on information rather than control or reputation. The better people know each other, the better they can make predictions about one another. Repeated communication can facilitate this predictability. In the case of Airbnb, people can message each other before or after booking an apartment thus

2_{Airbnb has a two-way reputation system in place where guests rate hosts and vice versa in order to inform}

(13)

9 resulting in the possibility to communicate with each other and form predictions. “It develops over time, largely as a function of the parties having a history of interaction that allows them to develop a generalized expectancy that the other’s behavior is predictable and that he or she will act trustworthy” (Lewicki and Bunker, 1995, p. 121). Hence, in contrast to calculus-based trust, knowledge-calculus-based trust relies on one’s own experience with an individual and focusses less on the control over the Airbnb guest and host.

The last stage of trust is called identification-based trust and focusses on the internalization of values of the trusted partner. It develops “as one both knows and predicts the other’s needs, choices, and preferences and also shares some of the same needs, choices and preferences as one’s own” (Lewicki and Bunker, 1995, p.123). The authors mention that colocation (living in the same house) can result in identification-based trust. Thus, in the case of Airbnb it could be the case that if guest and host are living in a shared accommodation, identification-based trust might occur. Nevertheless, only a small amount of people tend to end up in the identification-based stage since some parties may not find it relevant to build a

Figure 1 – Stages of inter-personal trust development (Lewicki & Bunker, 1995)

(14)

10 strong relationship or simply do not have the time or energy to devote to relationship development (Lewicki and Bunker, 1995). To summarize the above:

Proposition 1: Inter-personal trust between Airbnb guest and host develops in stages when transacting on Airbnb.

3.3 Medium used influencing trust development

(15)

11 (potential) that may be realized by some or remain an unfulfilled option for others” (p.10). In the same vein, people might also choose to forgo interactivity by not communicating through multiple channels, hence decreasing social presence due to higher information asymmetries.

Hence, the richer the medium used, the more interactivity potential it possesses. Since increased social presence aids trust development, the richness of and interactivity of the medium might influence the level of trust that can be formed (see Figure 2). For example, in the case of Airbnb, the initial cues such as reviews and the profile picture of the guest and host allow for some uncertainty reduction. Nevertheless, there is no feedback mechanism in the beginning. When engaging with a guest/host via messaging, the possibility of feedback increases the cues perceived, resulting in further trust. Lastly, when meeting offline, the guest and host meet face to face, which is the richest medium possible. Thus, at this stage, the development of trust is likely to be at its highest point since even more cues and information are available. In other words:

Proposition 2: Richness and interactivity of the medium used will increase social presence and

increase inter-personal trust.

In case people choose not to engage with their host via messaging or ignore for the most part to interact with the host face to face, medium richness and interactivity cannot be exploited and social presence would decrease, thus hurting trust development.

Proposition 3: Foregoing the richness and interactivity of the medium used will hamper

(16)

12

3.2 Factors influencing trust development

In addition to personal trust developing in stages, institutional trust may affect personal trust development. Hawlitschek et al. (2016) argue that trust in the peer is inter-related with trust in the platform in the sharing economy. This trust into the platform provider (trust in Airbnb) can also be called institutional trust and is different to trust in the Internet at large. McKnight et al. (1998) define institution-based trust by mentioning that it, “reflects the security one feels about a situation because of guarantees, safety nets, or other structures” (p.475). Some trust scholars (e.g., Zucker, 1986; Stewart, 2003) have found that institutional trust aids in transferring trust from one particular context to another context or entity. For example, Doney and Cannon (1997) explain in their research that consumers believe in the honesty of sales personnel due to them belonging to a reputable company. A similar mechanism seems to remain valid in the online environment. In fact, Hong and Cho (2011) found that trust in the platform provider of an e-marketplace can be transferred to the community of sellers in the B2C context. Hence, people believe that they can trust sellers online because of their association with the platform provider. Since the sharing economy shares similarities with the online B2C context, it may be the case that similar mechanisms of trust transfer apply in the C2C context. In fact, Airbnb provides several structural assurances such as the rating mechanism, host insurance and background checks of hosts/guests. Such structural assurances can form a safe context (institutional trust), which is then transferred to the sellers and buyers operating on the platform. In other words, structural assurances establish institutional trust, which is then transferred towards guests and hosts on Airbnb and thus facilitates the formation of inter-personal trust.

Proposition 4: Institutional trust (trust in Airbnb and its structure) will influence inter-personal

(17)

13 A potential model for trust development can be viewed below in Figure 2. It shows that inter-personal trust can develop over time and that it might be dependent on the richness and interactivity of the medium used. Additionally, structural assurances are expected to influence inter-personal trust development by forming institutional trust. Nevertheless, this model is expected to be expanded due to the exploratory nature of this research. This also explains why the exact causal relationships have not yet been depicted in the model below.

Institutional trust

Figure 2 - Potential model of trust development in the sharing economy

*The arrows are depicting that each medium builds upon each other and are not viewed stand alone in their impact on trust development

(18)

14

4. Research methodology

4.1 Research paradigm and design

The research paradigm applied is interpretative, which aims “to understand the complexities of the world through participants’ experiences” (Tuli, 2011, p. 103). It was chosen for several reasons. First of all, trust and social presence are social constructs, which vary depending on the interpretation of the individual. Hence, qualitative data helped to better comprehend the nature of trust in the sharing economy, by getting an in-depth understanding of the social construct of trust. Moreover, since little is known about trust development in the sharing economy, this research needed to explore the topic in-depth, which was only possible by conducting a qualitative study. Although some theories have been found to guide the study of trust development in the sharing economy, it remained critical to be open to other related concepts that might have had an influence on trust development. Lastly, qualitative data could explain how trust development unfolds by looking through the eyes of the person and thus helped discover subtle nuances that quantitative data is unable to detect especially with relation to how trust changes over time.

(19)

15

4.3 Data collection

The data collection used were depth semi-structured interviews. The advantage of in-depth interviews was the flexibility that it provided to the researcher due to departing from some questions in order to adapt to the interviewee (Legard, Keegan and Ward, 2003). Furthermore, in-depth interviews allowed for probing (Legard, Keegan and Ward, 2003), thus providing the opportunity to get an in-depth understanding of trust development in the sharing economy. Data collection was stopped as soon as a saturation point has occurred, meaning that no additional insights could be gathered from additional interviews (Baker, Edwards and Doidge, 2012).

The interviews were guided by making use of an interview guide (see Appendix Figure B and C). The interview guide was structured in a way to answer the research questions posed and followed a temporal approach. In other words, factors influencing inter-personal trust development such as structural assurances and the medium used were asked at both online and offline stages to verify the idea of inter-personal trust developing over time.

4.3 Sampling strategy

(20)

16 balanced sample of experience with different accommodation types to control for the impact that it might have had on the results. The chosen criteria and the justification for inclusion can be viewed in Table 1 and 2 in the appendix. Both guests and hosts were chosen to be interviewed since the investigation on inter-personal trust required a perspective from both sides, the guest and the host. Moreover, it was likely that the guest and the host perceived trust establishment different from one another.

4.4 Procedure

(21)

17

5. Description of sample

From September to October 2018 guests and hosts were contacted to ask for participation in this research and were subsequently interviewed.

Airbnb hosts were contacted via the Airbnb website. Out of 23 people contacted, 17 hosts agreed to an interview and six hosts declined yielding a response rate of 74%. During the interview process, three respondents were excluded since contact with them was lost. Nevertheless, two more hosts could be interviewed ultimately increasing the number of hosts interviewed to 16 hosts. The average age of the hosts was 42 years old. Out of the 16 hosts, six hosts were renting out a shared apartment, seven hosts rented it out privately and three hosts had experience with both. All Airbnb hosts were Dutch except for one British host. On average, hosts have rented out/shared their place on Airbnb for 2 years and 8 months. A summary of host characteristics including additional information can be found in table 5 in the Appendix.

14 Airbnb guests were contacted via phone or in person via convenience sampling. The average age of guests was 24 years old and all guests agreed to the interview yielding a 100% response rate. Out of the 14 Airbnb guests, four guests had experience with only shared accommodation, 6 with only private accommodation and 5 had experience with both. Guests were from diverse nationalities, including Austria, the Netherlands, Germany, France, Italy and Brazil. On average, guests used Airbnb 5.4 times prior to the interviews. A summary of guest characteristics including additional information can be found in table 6 in the Appendix.

(22)

18

6. Results

6.1 Motivation of using Airbnb

The two main motivations for hosts were to earn money and to meet other people. Logically following, Airbnb hosts that rent out their entire house/apartment are more concerned with the money-making aspect whereas hosts that share their apartment are keen on getting to know new people. For example, Airbnb host L mentioned: “Normally I would say to get in

contact with other people, but we are never here when people are coming to our house so it’s

not like that for us. For us it’s more like earning money, that’s the main goal of Airbnb

actually.” Airbnb hosts in shared accommodations enjoy meeting people from diverse

backgrounds, to listen to their stories and to understand their culture better. Some even mentioned that money is not the most important factor for renting out their Airbnb. In the statements of host C: “Well, of course, the money is one but not the most important (…) I think

the social aspect from being an Airbnb host is for me the most important. So, you can share something to someone else to experience the best of the city. I think it’s way more fun to be

in a house and be comfortable than to be stuck in a hotel at 3 or 4 km from the city.”

Similarly, Airbnb guests like to use Airbnb since it offers better value for money than the stay at a hotel/hostel. Guest C explained: “You can rent something that you have for

yourself, so you don’t have to share a bunkbed in a hostel or so, It’s quite cheap and you can

choose a nice location.” Especially the aspect that no other people stay at the apartment

intrigue people that want to travel with friends and have their own accommodation. Furthermore, Airbnb guests appreciate the fact that it is an actual home they can stay at, rather than a standardized room in a hotel, which makes it more of a local experience for them. For instance, guest B mentioned: “You get to know the actual country more than being

(23)

19

house of a person that is a member of the society. So, you get to know the culture more of the country.” Additionally, guests appreciate the local information they can get from Airbnb

guests. Guest E argued: “You also have a kind of partner there. Like the host you can ask him

questions about the city, he can recommend you tours or trips (…). In a hotel, you have not

that personal contact.”

6.2 Risks of using Airbnb

The primary potential risks associated with Airbnb for hosts is damage to their property and stealing. Furthermore, especially women mentioned potential physical abuse. Moreover, Airbnb hosts also explained that inconvenience for themselves can be a risk, which can occur if guests want to stay long periods of time or take up the room entirely. For example, host H explained: “It was more that we couldn’t have our own life anymore because he was in the

way. It’s ok if he wants to sit in the garden but him sleeping in it means that I cannot go there

and drink a glass of wine and talk to my husband.” Many Airbnb hosts also made clear that

they do not feel much at risk and that they barely had any negative experience, giving them a secure feeling to operate as an Airbnb host. For instance, host E explained: “I’m not really

worried that they make a lot of use of your stuff”. Similarly, host G mentioned: “I never felt

unsafe or that they are going to rob me”.

Guests identified four major risks when booking an accommodation on Airbnb. The biggest risk is that one might be disappointed with the apartments’ facilities due to false advertising. Moreover, some guests are worried that the Airbnb host may not appear upon arrival, thus leaving the guests stranded. For example, guest G mentioned: “You can get in a

not nice situation if you arrive in a place where you are supposed to be on holiday for four

days and you don’t have a place anymore and then you need to find a hotel and it’s going to

(24)

20 with the host either due to him being unfriendly/unhelpful or because he/she may physically harm guests.

6.3 Structural assurances of Airbnb

Proposition 4 projected that institutional trust will influence inter-personal trust through structural assurances by Airbnb. After conducting the interview, it became clear that institutional trust is of significant importance for the guest. For example, when asking the guests about their motivation and risks when using Airbnb, 9 out of 14 guests mentioned Airbnb’s security- and review system. For example, guest L explained: “Yeah I mean I think in

general it seems that Airbnb has quite a good security like management (…) so it seems like

it’s not that easy to upload a fake profile so to say.” Hence, even without pro-actively asking

how Airbnb can help guests trust people on the platform, guests immediately thought of Airbnb’s structural assurances. Thus, guests clearly seek out structural assurances such as reviews to reduce the information asymmetry between the guest and the host.

Furthermore, when asking guests how they decide which hosts to trust on the Airbnb platform, all 14 guests mentioned that they read through the reviews, showing the importance of the rating system to build initial inter-personal trust for guests. Reading the reviews is a deliberate process where guests start with looking at the star rating and if they deem it good enough, they carry on reading the review. Furthermore, half of the guests mentioned that they find the number of reviews important and would avoid a host that has little reviews over a comparable accommodation with a host that has more reviews. When asked why they trust the host at this initial stage the majority answers that reviews give them a sense of understanding whether other people liked the place, giving the host a certain kind of reputation. For example, guest D answered: “I think because they have recommendations

(25)

21

that. Like he is good in what he is doing.” Hence other’s people opinions build up a reputation

of the host, which helps to develop initial trust for the host. In other words, guests trust a host’s reputation because other people confirmed this reputation, which makes it easier to predict the behavior of the host.

Next to reviews, 9 out of 13 guests also find apartment pictures and their quality important to make inferences about the host. For example, guest M mentioned: “Once there

was a place and the sink, it was full of dishes. So, when I see this house I say ok he can’t even

wash it to take a picture, no.” Next to reviews and host names, guests also mentioned that a

superhost status, the profile picture and the description of the apartment/host is important for determining whether a guest would trust the host.

Hence, structural assurances such as reviews, apartment pictures, super host status, profile pictures and the description of the place indeed make it easier for the guest to build inter-personal trust. These structural assurances seem to be important for guests when they decide with which host they would like to stay with thus laying the base for inter-personal trust.

On the other hand, reviews are less important to Airbnb hosts. In fact, the interviews showed that only half of the hosts actively check reviews of the guests. Hence, they do not check reviews every time someone is booking with them and oftentimes even just check out of curiosity. Furthermore, even if guests do not have reviews, hosts would usually accept them, showing a stark contrast to guests who try to avoid hosts with no or limited reviews. Host D explained: “Yeah I notice that we get a lot of first timers and if they have more reviews

then it’s a lot easier to decide. But still we accept also first-timers. Yeah, we normally accept

(26)

22 Safety nets such as the Airbnb insurance do not have a large impact on Airbnb host. Often they do not really trust that Airbnb would follow up on the insurance promises or they only consider it as a factor with minimal influence. For example, host D mentioned: “Airbnb

says that they have insurance but I’m not really sure how much you can trust that. I mean if

your house burns down then Airbnb doesn’t cover that I’m pretty sure.” Additionally, host K

mentioned: “I try not to trust too much in that because if I wouldn’t trust the people, then I

wouldn’t trust the Airbnb insurance to make up for that. It’s just an additional thing so maybe

if something were to happen then yeah.”

Nevertheless, hosts explained that they do like if guests have some profile information since it makes it easier to trust guests. For example, host C explained: “Yeah I think it helps

but not consciously. It’s just, it’s good if you have a profile and fill in everything you ask. It’s

also, I’m an online marketer, so if you have a company with a Facebook page, with an

Instagram account and a good website, with all the credentials that you need and information

then you are trustable, and you are solid”. Half of the hosts also mentioned that their trust

decreases if a guest does not have a profile picture. For instance, host E mentioned: ”If there

is a picture then it is more like you can trust someone than if there is no picture because you

want to know who that person is”.

Lastly, Airbnb hosts also mentioned that in the beginning phase of Airbnb, they were much more likely to read all the reviews since they were less familiar with Airbnb. For example, host E mentioned that her and her husband would look through the reviews and select people rather than accepting everyone. In her statements: “The first half year we were

mainly just sensing and getting a feel how it works because it was also, hosting was new to us

(27)

23

we are going to select every one of them”. Additionally, host A explained: ”I did that [checking

profile information] first. When I started Airbnb that’s when I did all of that every single time”.

In a nutshell, structural assurances positively influence inter-personal trust development for both the guest and the host, especially in the initial trust formation stage. However, structural assurances are of such high importance to guests that without them they would not undertake bookings on Airbnb. On the other hand, Airbnb hosts are much more likely to let strangers stay at their house even if they do not have any reviews or profile picture, especially if the host has already gathered some experience with Airbnb.

6.4 Medium richness & interactivity

Proposition 2 stipulated that the richer the medium used, the higher the interactivity of the medium and the higher inter-personal trust would be. After conducting the interviews, it became clear that the richness of the medium used inherently determines the interactivity of the medium. If it is exploited, people perceive higher inter-personal trust. If it is not exploited there is not any base to exploit social presence, as indicated by proposition 3.

6.4.1 Medium: Online communication

Overall, 15 out of 16 hosts mentioned that online communication builds trust for them and is even more important than structural assurances such as reviews. In fact, almost all Airbnb hosts determined whether they would trust their guest by looking at the first message that guests send to the Airbnb host. They made clear that they do sometimes check reviews but that messages are more important to get a better understanding for the current situation. For example, host P mentioned: “You get a booking, but it starts with a message. That’s the first

thing you go to and if they write a longer story with 3 friends, we are from Munich, we study

this and this and we really need a break from our study and we have never been to Groningen

(28)

24

host us please? That’s enough.” Furthermore, host I explained: “I don’t always look at the reviews but yeah some people have reviews. You can read them if you want but most people,

they tell a story where they live, what they are doing, why they are coming to Groningen, to my place, like that.” By looking at the message and seeing why the people would like to stay

at their accommodation, hosts can gain more information about the guest such as the reason for their stay or with which party they are traveling with. Hence, information asymmetries decline. Host I explained: “You don’t know the people but if you write and hear what they are

doing. It makes me comfortable that I already knew the people a little bit. I don’t know them,

but it helps for me.” Hosts explained that also the feedback mechanism inherent in online

communication builds trust. For example, host K explained: “Well because the fact that they

have responded so it’s easy to make a request but then also respond to questions that I ask

and give normal feedback, that is, that raises the trust for me.” Furthermore, hosts explained

that it is easier to make inferences about the person when writing with them due to their writing style and speed of reply. Host H explained: “I think it helps. You get an idea whether

he is polite or not.” Also host L mentioned: “Yes of course. I mean when I’m sending a message

and they are responding really quickly and really that they are happy to come here, of course,

I am trusting people more and if I don’t get a message or it’s really slow, I’m not really happy,

then I’m trusting people a little less of course.”

Overall, 13 out of 14 guests mentioned that online communication helps them develop more inter-personal trust than just structural assurances such as reviews, confirming the importance of online communication to the guest. Similar to hosts, this is mostly due to the inherent two-way feedback online communication enables. For instance, guest E mentioned: “Now it is not just receiving information but also exchanging information like I

(29)

25 explained: “I’m always still suspicious of the internet, specifically if I don’t know the person

and you know that they have fake profiles and because you know things can happen. If

someone answers me I feel like ok there is actually a person I’m talking to”. Hence online

communication enabled the guest to exchange information with the host and to verify that an actual person is operating the Airbnb. This is in line with the proposition raised that increased interactivity builds social presence due to possible feedback mechanisms.

Furthermore, as with hosts, guests use certain cues to make inferences about the host. One of these cues is the writing style of the host. If hosts reply in a friendly, interested manner, then guests are more likely to trust the host. For instance, guest E argued: “There was some

information flowing back and forth which was nice because you can read the writing style

from the landlord. Is he extensively explaining things, is she even willing to put in some effort,

clearing his own parking lot so you can park there or inflating a third bed”. Furthermore,

guests perceive hosts who reply quicker to be more competent and to reduce uncertainty. Guest K mentioned: “If he doesn’t reply when I send a message then I’m scared that he will

never show up.” Interestingly, although guests mentioned that trust increases through

communication, trust is not necessarily broken if hosts do not reply immediately. Nevertheless, it does make people feel uncomfortable because of uncertainty increases.

(30)

26

6.4.2 Medium: Face to Face communication

Furthermore, the majority of guests and hosts explained that meeting their guests face to face increases trust even further.

To be specific, 13 out of 14 guests explained that face to face communication increases their inter-personal trust towards a host more than communicating online alone. Being able to meet face to face means moving the online conversation to an offline context with immediate and personal interaction, which increases inter-personal trust. The main reasons for this trust increase by the guest is an additional verification that the person is real as well as making predictions about the person through inter-personal cues such as shaking the hand and body language. For example, guest H mentioned: “I think personal contact always creates

trust and when you have a real impression of a person you can better estimate how he or she

is. So, I think I mean it’s just online communication so you can basically tell everything, you

don’t really have an impression of a person. How people are in real life does make a huge

difference because they have an impression, you see what they wear for clothes, how they

behave if they are polite so stuff like this, which has a huge impact on the image of a person.”

Additionally, guest M explained: “It is kind of like paralanguage. So, I think this is very

important because you can hear the tone of the voice and that makes everything different.

There is more room for interpretation in texts”. Thus, face to face contact increases the

(31)

27 to make them behave well. For example, host C mentioned: “I want to see them and check

them and not only to be sure for myself, but I also think that when people have seen me they

will be more timid. Because they have been in contact with me. So, they feel a little bit of pressure. If we don’t clean the house normal or proper then you will be mad or something and

he knows us because he has seen us.” Thus, hosts seem to use the face to face medium more

strategically than guests.

Four hosts (25%) did not feel that the initial face to face meetings increased their trust. This had mostly to do with people saying it does not really have to do something with trust but more with getting a gut feeling for the person or it being easier to talk to the person face to face. Nevertheless, the hosts that mentioned that the initial face to face encounter did not increase their trust mentioned that when the guests stayed for a couple of days, then trust would increase because of getting to know the person better. For example, host F explained:

“Sometimes I think wooah I wonder if I can really trust them but when they are in that room

they can come and go through the garden and when I sit in the garden and I start talking to

them, 9 out of 10 times, they appear to be very very nice people so it’s just the first impression

isn’t always the right one.”

Moreover, all guests that have shared the accommodation with a host also mentioned that inter-personal trust further increases when spending more face to face time during the upcoming days with their hosts. Obviously, the precondition for this is that the host and the guest spend enough time at home to interact with one another. In the statements of guest H:

”Yeah I think you get more personal contact because I think when you have a very short stay

you see the host one time, you get the key and then he maybe leaves and you maybe never

(32)

28 Similarly, the majority of hosts explained that trust usually increases over a couple of days. Nevertheless, this is not possible if the guest is not at home and thus not interacting with the host. For instance, host P clarified: ”Well, in general, we do what we do, and our guests do

what they do, we just live separate lives which happens to be under the same roof”.

Furthermore, trust oftentimes cannot increase if guests rent out a private house since the host is then not at the accommodation to interact with. For example, one out of 16 hosts mentioned that since he never sees the guests due to a number lock being in place, no increase in inter-personal trust takes place.

Thus, face to face meetings do increase inter-personal trust for hosts, especially when they have been living with the guest for a couple of days. The initial, first face to face meeting does also increase inter-personal trust for the majority of the hosts but not for everyone since the first impression might not be an accurate one.

(33)

29

6.5 Inter-personal trust developing in stages

Proposition 1 argued that inter-personal trust develops in stages. After conducting the interviews, it became evident that trust can indeed develop in stages but that there are clear differences between the antecedents of initial trust between guests and hosts. As pointed out in the previous sections, guests commence by making the decision to book with their preferred apartment/host. When making this decision, guests conduct a cost-benefit analysis of whom to stay with. By doing so, they consider the hosts reputation through looking at the reviews, by taking the superhost status into account as well as through looking at the profile pictures. If hosts have negative reviews indicating a somewhat worse reputation compared to other hosts, the costs would exceed the benefits and guests search for a different accommodation. For example, guest D explained: “I think I mostly look at the reviews. So if

there are a lot of reviews like in the 100 or pushing 100 and 4 or 5 stars I always kinda trust it.

Like when I was searching for Madrid I also saw one with only 1 review and I was like no I’m

not gonna even look at it. Hosts also realize that their reputation is crucial in attracting guests

and can feel forced to have a high ranking. For example, host K mentioned: “Yeah it is very

important because that’s what Airbnb is mostly about, the ranking, the amount of stars you

get, and I really hate how Airbnb does that because everything that is not 5 stars is bad. But also, if you are below 4.7 average then you are below what is expected of you. That’s also how

Airbnb communicates that.” Hence, Airbnb hosts can also feel pressured to perform at a

certain level, making it possible to control for host’s behavior. In other words, guests do enter the calculus-based stage when developing initial trust for the host.

(34)

30 example, host B explained the dilemma: “If you, for instance, think I’m not going to accept

people who are first time renters on Airbnb, (…) then I think more than half of my guests are

gone.” Furthermore, guests do not seem to worry about their own reputation meaning that

hosts cannot rely on the reviews to “control” for the behavior of the guests. Thus, hosts seem not to enter the calculus-based trust stage.

Nevertheless, what does influence a host’s initial trust perceptions are individual characteristics of the host. For example, Airbnb hosts mentioned that they have a general tendency to trust others which reduces the necessity to check every guest out via their profile information. For example, when hosts were asked why they trust guests on Airbnb, they answered: “I am a person who wants to trust everyone” (host F) and “I start with trust unless

I feel like I can’t trust. So that’s usually the way I am. I am very open in that sense. So, I always

start with trust and then I get the sense that I can’t trust that’s when that changes” (host A).

They also explained that one should not become an Airbnb host if one does not have trust in the society. For example, host C explained: “You need to trust the society and give them what

you got and hope that you will get it back.”

(35)

31 also his/her speed of reply leading to the development of knowledge-based trust. In the words of guest A: “I think it’s just like the contact that I have with them myself. Otherwise I

just have a description of the person that I’m going to stay with from other people who have been there, which is actually a good basis but it’s important that you get to know the person

yourself. So that you talk to them, or that you keep texting them and so yeah you find out

something about them, but you do it yourself.”

Similarly to guests, messaging serves as a reason to decrease information asymmetries for hosts and allows for the possibility of gaining the first-hand experience, which makes it easier to predict guest behavior. Hence, both guests and hosts enter the knowledge-based trust stage.

As previously pointed out, face to face interaction between guest and host further manifests knowledge-based trust due to getting additional information through inter-personal cues such as the handshake of the person, or whether they are smiling or not. Especially when getting to know the guest/host better, one can observe the host in various situations, which accumulates knowledge about the host. When meeting the guest face to face, behavior can be observed further and the knowledge-based trust grows, as pointed out in the previous paragraphs

A very limited number of guests/hosts mentioned that they experienced an even stronger trust mechanism, thus making it able for them to transcend to the last stage by Lewicki and Bunker (1995), which is called identification-based trust. Out of all 14 guests interviewed only 2 had such an experience. For example, guest A explained: “The people they really shared the

apartment with me like I was their roommate forever, not only for a couple of days but for a

(36)

32

anything and you spent some time together like cook together or hang out together. Guest J

also explained that as she lived with the host, they did not only talk about personal information but also adjusted to each other’s needs: “We have a good relationship and we

talk about our personality and this we trust much more each other. It’s kind of in a relationship.

It is a good relationship because we have our space, for example, some days I don’t want to

talk so much so I go to my room and that’s fine and it’s an amazing relationship.”.

Identification-based trust did only occur for a limited number of guests and also only if the host shared their apartment with the guest. Hence, guests that do not stay with their host are unable to reach this stage.

As is the case with guests, only a tiny percentage of hosts enters the identification-based stage with a guest and can only be achieved when both the guest and the host were sharing the apartment. Only two hosts experienced a deep friendship evolving with their guests. For example, one couple mentioned that the people staying at their apartment happened to move to the same village and needed a stay before moving in. Host D explained:

“So they stayed with us for three weeks, and it was just like we had a big family. When I got

home from work, she was already cooking here, dishwasher was cleaned out and the kids were

really nice and social. We heard very nice stories. And now they live in the same city, so we

visit them very often. We see them once a week at least, they are really good friends now.”

Additionally, host G mentioned “He was really wise, 32 years old or something and we had

such good conversations. Really wise and we told each other everything, we learned much

from each other and I showed him my country, and we went to my sister’s and we are even in

contact now. He got married a couple of weeks ago, he sent me some pictures. He’s going to

come here next time when he is working in Groningen.” As mentioned in the literature review

(37)

33 since they do not have the time to invest in such relationships. The majority of Airbnb guests and hosts mentioned this. For example, host M mentioned: “There are some people I think

that would book again if they came back to Groningen and that I would really like because I really like them, and they felt as friends, but life just goes on. And I got enough friends. It’s not

like I need friends, so I would not invest in it.”

6.6 Additional variables found/Exploratory part:

The analysis of the interviews also revealed new findings that were not previously included in the conceptual model and will be discussed in the following paragraphs.

6.6.1 The influence of similarity

Two factors were found that were translated into the concept of similarity: (1) culture and (2) shared values.

First of all, culture seems to have an impact on both Airbnb guests and hosts. In fact, especially the similarity and difference between the own culture and the culture of the guest and host seem to have an impact on inter-personal trust. Two of the guests mentioned that they trust hosts with a foreign name less than somebody that matches the name of the region they are traveling to. For example, guest I mentioned: “I had the choice between staying at a

host with a Russian name and someone else and then I was like I better go for the other one

which actually the location was worse than the Russian one. But since it just didn’t feel right,

not especially the Russian thing, but it is just weird that you see on an English page something

in Russian, which you can’t understand.” Similarly, guest G answered: “It’s a bit judgemental

but I have to say that probably especially in Sweden whenever the host was not a Swedish

person and let’s say that that kind of person had like an Arabic or kind of exotic surname, I

(38)

34 Similar to the guests, host N mentioned that she prefers to see the nationality on the profile of the guest and that she is against instant booking because: “Then you can also have

some East-Europeans and the whole house is empty.” Additionally, guests and hosts with

different cultures might find it difficult to share an apartment since barriers in communication may occur. For instance, host D explained: “Yeah but if you write with them, they were quite

well with English but with talking they didn’t understand some things. So there was something

wrong in the communication. So when we asked, they would always say yes. Just to be polite

and then they still didn’t know.” When asked if she will be more careful with having this

certain culture in their apartment again, she answered: “Yeah, I think we’ll think a second time

when we get them.”(…) “I think we are going to ask them a couple of extra questions just to

be sure.”

Furthermore, shared values made it easier to create initial trust for hosts on Airbnb. In fact, hosts mentioned that they believe that Airbnb guests have certain values or characteristics that make them trustworthy in comparison to people that would stay in a hotel. Host C explained: “It’s almost even every time we come back it’s almost as clean as we

left it when guests came. I think it’s a part of the social room. I think we have specific

eigenschappen (characteristics). I think we have some characteristics in this group.” Similarly

host H mentioned: “I think it’s also the same kind of people because when we talk about

Airbnb with our friends, they say I don’t understand that you do that (..) They think differently.

And I think that if you book through Airbnb that is just different kind of people. You trust

them.” Additionally, host D explained: I think well at least I have some idea that there are

more, the people that would look for our house via Airbnb that they are already a selected

group of people and they are different from the people that just use booking.com or use a

(39)

35

6.6.2 Active risk minimization

Secondly, it was discovered that hosts actively invest in risk minimization strategies to make it easier to trust the guest. In other words, while hosts have less possibility in actively choosing their guests, they have the possibility to strategically influence which guests are applying for their accommodation. In fact, Airbnb hosts use location, price and the profile information of the host to determine the target group that their offering attracts. For example, host E mentioned that he would not share his place with people in Amsterdam: “I think you don’t

really want to live with all the people coming to Amsterdam. They have a different

motivation/purpose to visit Amsterdam but if they want to visit Groningen they probably come

here to study or have something to do here.” Host L, who offers to rent out the whole place

agreed and further explained: “Maybe you saw lately the one post of a guy in Airbnb, where

it was totally trashed. But here in Groningen it’s not that party place. Maybe for students but

in Amsterdam it’s different than here.” Hence, Airbnb hosts in Groningen are less concerned

and thus also perceive lower risks for people coming to Groningen. This also makes it easier to trust people booking their apartments.

Additionally, almost a third of the hosts also mentioned that they use their pricing strategies to influence the kind of people that try to rent out their places. Hosts either temporarily change their prices for certain events or permanently choose for a higher price category. For example, host G explained that she raises her prices during KEI week where a lot of students come together to party. In her own statements: “You know here the KEI week.

Sometimes I make prices that high (…) because I don’t want that here, students, 16-,

17-year-old, partying all the time. I’m working so that won’t fit. So then I make it really high so it’s not interesting for them.” Other hosts explained that they chose for a higher price in general in

(40)

36

the other rooms in Groningen and that also attracts different people. Yeah that’s one of the reasons why we put our price a little bit higher.” Interestingly, host D mentioned that they are

aware of their low price and are thus more cautious when it comes to choosing the right guests for his Airbnb. In his words: “We are one of the cheapest houses in Haren so the risk is

that you draw the wrong kind of persons for the wrong kind of reasons, so the students that

look at a place in Groningen and the prices in Groningen are higher so then they go a bit further

away, they come here and instead of having a nice group of people in here, you get drunk

students.”

Additionally, the profile information of the host also helps to reduce their risk and attract the right target group. For example, host M explained that she is using a photo on Airbnb where she and her boyfriend are both on the picture to make people believe that they are staying with the both of them. In this way, she feels less at risk for men wanting to stay with her for the wrong reasons. She explained: “I left the profile with the two of us, me and

my boyfriend because I don’t want creepy man to think ooh there is a girl alone. That is

definitely about trust. It’s to protect myself because I thought if I put a picture of me then

maybe guys are thinking ahh that’s nice. No no no, don’t happen. People just think they book

with me and my boyfriend. That’s really that’s why I feel confident. As a girl alone I would not

do it, maybe even. So there’s the thing about trust.”.

6.6.3 Experience

(41)

37

trusting people more than in the beginning because when something is new then you first have to trust people and after one year of trusting people, maybe you are a lot more like oh

yeah it went good, all the several times so now it’s fine for me to leave the PlayStation there.”

Additionally, host A explained that she used to focus on reviews more in the beginning, but that experience changed that. In her words: “I did that first. I used to look at the person first.

When I started Airbnb that’s when I did all of that every single time. But then I had some

experiences and I realized there I no way of knowing until they are here. So I started trusting

because I had so many good guests. I can check their profile, but it doesn’t really say anything

until they are there.”

Furthermore, in the interview’s participants were asked whether they believe that one experience with a particular host/guest would influence how much they would trust their next host/guest. Interestingly, guests explained that if the experience was a positive one, they would oftentimes generalize and believe that all Airbnb hosts must be trustworthy. For example, guest F mentioned: “So like I have said at first I was sceptical about the whole Airbnb

thingy but then I did it once and twice and a third time and now I trust Airbnb hosts more

because it always works out for me.“ Nevertheless, when having a negative experience,

people explain that they would not believe that all Airbnb hosts are untrustworthy and they would continue using Airbnb. For instance, guest L explained: “I do think that every host is

different in their individual, but I do know since what happened in this one instance, I will be

a little bit more thorough in the future (…) I would still always use Airbnb again. Just because

it happens once, I don’t think that it will happen in every instance.” In fact, most guests

(42)

38

“I would probably think about hmm maybe I should read the feedback more and also see how the communication flows.” Hence, negative experience of a guest would lead to the more

active use of the structural assurances by Airbnb but not necessarily to a decreased trust perception of the person they are booking with. However, when the experience was positive people would immediately trust Airbnb hosts more.

Similar to guests, hosts also explained that they would not draw general conclusions of the Airbnb guest community because of a negative experience. In the statements of host P: ”Well the recent negative experience that made us not trust the next guest less but be more

cautious but that’s not specific on the next guest it’s just in general. But we had 70 guests or

so, so we have a basic general trust in people. And that hasn’t changed from before we started

with this until now it hasn’t changed, we still trust people.” Only if there were a pattern of

multiple negative experiences, hosts would consider stopping Airbnb. In the statements of host O: “I see this like an exception and it happens, and it wasn’t really a bad experience.

Maybe if you have a couple after each other then yes but otherwise no.” Some also mentioned

that they took some action or would take some action such as reading the reviews better or switching from instant booking3_{to instant booking with certain criteria. For example, host I}

explained: “It could be that it’s a big mess here. Then you think well I will screen the people

better, I want to read the reviews and all of that.”

In conclusion, the results can be summarized in figure 3 below. Five factors that influence inter-personal trust development could be identified. The five factors identified are (1) perceived risk factors, (2) individual factors, (3) institutional trust factors, (4)

3_{Instant booking means that guests can instantly reserve an accommodation for a certain date without hosts}

(43)

39 communication factors and lastly (5) time-based factors. The factors have been chosen to be arranged in a temporal format of 4 phases since this research concerns inter-personal trust development, which means that an ongoing process is taking place and ongoing trust mechanisms are better understood when placing them into their correct temporal phase. The four phases described are (1) before engaging with one another on the Airbnb platform, (2) the initial phase (antecedents) of inter-personal trust, (3) factors manifesting inter-personal trust-development during meeting online or offline and (4) outcomes of inter-personal trust development influencing their future trust perceptions.

(44)

40

7. Discussion

Trust has been named ‘the currency’ of the sharing economy, playing an integral part of the success of sharing economy providers such as Airbnb (Botsman and Rogers, 2010). In this research, four propositions with regards to inter-personal trust development were raised and could either be fully or partially confirmed in the findings, validating the importance of trust mechanisms in the sharing economy. Following from the analysis and its five factors identified, a new conceptual model has been developed and can be found at the end of the discussion in Figure 5. It will be used to embed the expected and new findings in one coherent model.

Expected findings

(45)

41 the general tendency to trust by Airbnb hosts. Hence, if Airbnb hosts have a general tendency to trust people, this disposition to trust is able to replace the calculus-based trust stage by building up swift trust. As soon as expectations of the guests are being confirmed through communication online or offline, hosts move into the stage of knowledge-based trust where predictions about the guest can be made. Furthermore, only a handful of people reaches the identification-based trust stage. Hence, to answer the first research question, inter-personal trust develops in stages with guests starting in the calculus-based trust stage and hosts initially developing swift trust through a disposition to trust. From these initial trust stages onwards, trust grows and develops to knowledge-based trust with a few exceptions even being able to reach the identification-based trust stage. These findings are presented in Figure 4 below.

Hence, this finding shows that in the online-context, Lewicki & Bunker’s (1995) stages model is largely applicable. Nevertheless, the calculus-based stage is only applicable to guests since hosts substitute the calculus-based trust stage by swift trust, which emerges from a

Calculus-based trust Disposition to trust Knowledge based trust Identification based trust HOST GUEST Stage 1 Stage 2 Stage 3 Figure 4 – Stages of inter-personal trust development

(46)

42 certain disposition to trust, thus enriching the theory of inter-personal trust development by showing that individual factors may have a significant influence on initial trust perceptions for hosts. Thus, individual factors, including a general tendency to trust, were decided to be included in the conceptual model. They were included to be influencing initial inter-personal trust perceptions since the interviews made clear that the general tendency to trust is the reason why hosts develop initial trust for the guests.

MOVING FROM STRANGER TO “I TRUST YOU”