How do various trust mechanisms affect users’participation in Airbnb？

曲馨儀 易欽洲

Introduction

Airbnb， as one of the best-known sharing economy businesses all over the world， is a company based in San Francisco operates an online marketplace and hospitality service for people to lease or rent short-term lodging. The company has more than 5 million lodging lists in 81，000 cities and 191 countries and has facilitated over 300 million check-ins （Wikipedia.com， 2018）.

The platform has already been used to locate emergency housing for people influenced by more than 90 natural disasters in 20 countries， which rendered assistance to tens of thousands of people suddenly homeless overnight. In some cases， it can help the community respond to people in need faster than the government. The company then set a more ambitious target： helping assist some of the 65 million people who had been displaced by conflict， natural disaster， or disease around the world （Meixler， 2018）. How does business interaction or free offer happen as the key to private home is given to strangers？

The "sharing economy" is a rapidly developing and transformational new business model that is driving business innovation and creating new opportunities in communities globally. On account of widespread Internet， online accessibility makes owners of underutilized assets available. Apparently， sharing economy indeed undermines conventional market but creates values. Rather than being invisible behind the business， users in a sharing community are relatively transparent through profiles， messages and social platforms. The dominant value in a sharing economy is trust that is required in any interaction （Stephany， 2015）. In this case， we conducted a research on how do various trust mechanisms affect users participation in Airbnb.

In fact， the sharing economys impact on how we consume and its ability to fuel economic growth may be the most profound change of all that the Internet will bring. On one optional night in 2014， there were 785，000 people in 191 countries either staying in a stranger's home or welcoming ones into theirs only in Ainbnb's service （Gebbia， 2014）. Extra incomes of hosts and more affordable room rates are natural consequences. Moreover， traveling and economic activities are boosted by improvement in population mobility and flexibility. After breaking the stereotype that strangers equal to danger， local experiences replace efficient and consistent travel. Human connection benefits culture integration to some extent. Obviously， Airbnb's trust design has overcome stranger-danger bias successfully. How could they achieve this and what are users' reflections？

Discourse

In order to explain how Airbnb design trust， its co-founder Joe Gebbia did a second experiment during a TED Talk. By asking audiences to unclock their phones and pass it to the left-hand-side person， people are easily pushed out of the comfort zone and felt a little bit panic. Apart from the feeling of panic， most of people also felt sense of responsibility while holding a personal item just like how they would feel in private home （Gebbia， 2014）.

BOOKING

Unlike conventional hotels seeking third-party applications or creating homepages to offer information， Airbnb itself is a sharing platform containing all information of hosts， guests and home resources. Airbnb is created as a stage to help make introduction between strangers in a community （Aufmann， 2016）.

Identity authentication also differs between conventional hotels and Airbnb. Identity clarification is achieved by comparing jurisdictional information with guests' ID cards， passports or driving licenses. However， this traditional process is not enough for Airbnb users to know each other deeply. First impression is universally believed important， hence it is worth designing for trust building. Filing profiles fully is an opportunity rather than a task for users to demonstrate themselves in a community （Aufmann， 2016）.

Airbnb also operates a website called Airbnb Citizen to provide community and organizing opportunities contributed to improving the sharing economy. Third-party sites of Airbnb Citizen such as Facebook accounts are collected， so that uses are able to send invitation messages to friends via the third party site itself. Messages also reveal a bit of the character and personality of the writer， which helps hosts and guests start construct a trusting relationship. To some extent， trust takes effort. The more effort and care a guest show before booking， the more likely they will to be accepted by hosts （Aufmann， 2016）. Building the right amount of trust takes the right amount of disclosure. This means that either too little or too much sharing leads to a decrease in acceptance rate. In order to design for just the right amount of disclosure when a guest first messages a host， the size of box is used to suggest the right length. Guests are guided with prompts encourage sharing （Gebbia， 2014）.

PAYMENT

Handling payments is a sophisticated technical challenge， contributing to understanding who was making a booking better. Rules are designed to help remove some uncertainty around payments. Airbnb acts as a transfer station of payments between guests and hosts. Funds are kept by Airbnb for 24 hours after guests checked in. They are not released to hosts immediately as to give both parties some time to notify Airbnb if something is not right. A customer support organization has been built as a way to connect Airbnb when something goes wrong that now covers all timezones and numerous languages （Newman， and Antin， 2016）.

RATING SYSTEM

It is commonly believed that a well-designed reputation system is key for building trust. The situation on Airbnb platform is totally different from hotels because it is hard for users to leave bad reviews. Throughout the application process， hosts and guests can find reviews to build trust amongst users in the marketplace， hence bad reviews probably influence future renting and booking . Eventually， a 'double blind' process is used where guests and hosts reviews would be revealed only after both had been submitted or after a 14-day waiting period， whichever came first （Gupta， 2018）. A joint study with Standard demonstrated that it is human nature to trust similar people rather than dissimilar ones. But when reputation is added into the mix， in the form of reviews， and accumulates to 10 and above， preferences will change. High reputation beats high similarity （Gebbia ， 2014）. Therefore， Airbnb's robust reputational system comprised of authentic user-generated reviews and star ratings （Gupta， A. 2018）.

Survey

We conducted a survey initially to figure out how customers respond to ID systems.

According to the results， 70% of people prefer to choose hotels rather than Airbnb， considering the prices are same. Apparently， people are more willing to trust hotels， or otherwise， hotels' trust system is more well-known and successful.

One thing that both Airbnb and hotels have to conclude in their trust mechanisms is confirmation of customers' ID， the identity authentication. Customers' attitudes toward the provision of their ID to hotels and Airbnb platform are quite different. 57% of customers are "completely willing" to provide their ID to hotels， but only 30% of customers have the same willingness level to Airbnb. The difference illustrates customers do not trust Airbnbs privacy protection， but they trust the hotels. We also asked the reasons for their choices， most explainations are "I feel like I have to do that" in both hotels and Airbnb as they do not know how trust mechanisms work. Because Airbnb is just a private website or application while hotels connect directly to the police， 14% people chose "really resistant” to provide to Airbnb worrying about identity leakage.

The next question is about influences of eight factors on room selections. Customers of hotels pay most attention on the complete and transparent information given by suppliers because 73.64% respondents marked 5 points. The real-name system and services offered by numerous staffs are also important factors since more than 60% respondents marked 5 points. There is one thing that out of our expectation. Star rating is the least important factor that people concern， in which only 40% respondents marked 5 points. That is 18% less than the factor of “previous comments”. Surprisingly， the star rating system supported by government and professional department works less effectively than normal customers' comments.

Hosts regulations and protection of privacy are the major concerns of Airbnb customers （76.36% 5 points）. This can somehow response the previous questions about the ID system， and some explanations that customers are worry about identity leakage. Compared to hotel， the concern about privacy protection improved. The next important factor is feedbacks and comments from past customers with 64.63% respondents marked 5 points. There is no laws or complete official rating system for private renting houses， so it is in our expectation to see "feedbacks and comments" exists in the front row.

What surprises is hosts' photos and past experiences （50% 5 points） and how the houses looks like （63% 5 points） have large influence on the trust. Trustworthy photos do result with a price premium： hosts whose pictures are perceived as more trustworthy charge higher prices than their less trustworthy counterparts.

As many people cared about customers' comments more than official staring ，we did another survey on rating system in order to gain more specific information about the correlation between comments and room selections.

In the question about how often do customers make comments about a hotel or tenement， 63.64% of the respondents say they "sometimes" make comments. In the question about how much do customers trust the comment， 64% of respondents says they "trust but need to think". We see customers are relatively cautious when they make comments and choose to trust comments. We add the choices like "only trust negative comments" and "only make comments when you have bad experience". In the expectation， they should account for large proportion， however， there are only less than five percent of respondents tend to trust negative information. We asked another question， that is "Which kind of information have more influence on you？" The result finally met our expectation， 70% of people chose negative information and other 30% chose positive information. What we can conclude is trust and influence are two different sections， influence does not affect customers' trust.

Discussion

Referring to the questions in the introduction， indeed there is nothing new for a company to concentrate trust. The trust that makes Airbnb possible is when hosts and guests trust each other.

The flaw of the survey is mainly on the lack of respondents' personal information， such as their age groups， or cities. The lack of these directly causes we cannot know the changes if respondents' surrounding environment changes. Also， the limited number of respondents， which is only 221 people， constrains our persuasion.

Sharing houses provide a very good chance for people to create values with their extra spaces and willing to learn new culture. The future of shared cities tends to bring community and connection instead of isolation and separation. Trust is the key to build such a connected and peace community. According to the survey， not many people fully understand how trust mechanisms works in Airbnb. This paper can help them reassure the safety of Airbnb. Also， we tend to spread the ideas to more people， and promote the idea of sharing houses， offer more choices that are interesting and affordable for customers.

在完成這篇論文的過程中，查找閱讀整理文獻資料的能力獲得了顯著提升。同時也鍛煉了發起調查，處理數據的能力。對于共享經濟的理解更加深刻具體，對于信任機制的建立及反響有了更完整的認識。培養了自主調查研究的興趣，也認識到自己的不 足。

這是我個人第一次參與論文的書寫，感慨頗多。在我看來完成論文本身就是一次升華自己的過程，專業知識，寫作能力，思考問題的方式等都得到了長足的進步。印象最深刻的是有時明明知識和思路已經在腦海里，但還是很難把它們表達出來。努力本身是個看起來大義凌然，卻永遠虛無縹緲的東西，定要找到機會把努力承于某個載體上。從0到0.5很難，從0.5到1同樣也很難。