Interactive value formation: drivers and outcomes from Airbnb guests’ perspectives

ABSTRACT

This study explores interactive value formation, particularly the underlying drivers of three value outcomes in the Airbnb context: co-creation, co-destruction and co-recovery. The study focuses on reviews posted online by Airbnb guests in English. These posts contained customers’ positive and negative experiences with Airbnb on Trustpilot. The data analysis uncovered two main themes that reflected the drivers of value co-creation, co-destruction and co-recovery (company’s customer service and hosts’ actions). First, after a service failure, many guests experienced value co-destruction because they felt that Airbnb’s customer service agents could not solve their problems in a timely and proper manner, while the use of successful recovery efforts by the service agents served as an antidote to value co-destruction, thereby contributing to value co-recovery. Second, host’s friendly behaviour, including prompt communication between the host and the guest, led to value co-creation. On the contrary, inadequate communication and unethical actions by the host generated value co-destruction among the guests and resulted in a decline in their well-being. The findings suggest that particular value dimensions can individually act as a source of either value co-creation or co-recovery, while their inadequate integration in the interactive value formation processes leads to value co-destruction.

KEYWORDS:

• Interactive value formation
• value co-creation
• value co-destruction
• value co-recovery
• service failure

Introduction

Airbnb, a peer-to-peer hosting service provider and a popular sharing economy platform, has a significant impact on the global tourism and hospitality industry (Casado-Diaz et al., 2020; Dogru et al., 2020) and in Nordic countries (Adamiak, 2018, 2020) including housing access, affordability issues and residential displacement (Garay et al., 2020). For example, Miami Beach, Florida surpassed all listings worldwide (Inman, 2020), while Copenhagen was among the top Nordic cities with the highest number of Airbnb listings (Adamiak, 2018) that generated an estimated 650 million dollars in 2018 alone (Airbnb, 2019). Furthermore, the active listings in Denmark and Iceland surpassed in recent years (Adamiak, 2019). In the same vein, in 2017, Visit Sweden, Sweden’s official tourism board, collaborated with Airbnb to turn the entire country into an Airbnb listing (Airbnb, 2017), while Airbnb represented 90% of the dwellings in Stockholm (Einefors, 2018). However, although many private rental owners have adopted a customer-centric view, with a focus on offering memorable experiences (Sthapit, 2018), others may have a firm centric view that is contrary to value co-creation (Binkhorst & Den Dekker, 2009).

Given that “value is uniquely, experientially and contextually perceived and determined by customers” (Grönroos & Gummerus, 2014) and, in the context of Airbnb, the service provider (host) predefines both the tangible and intangible aspects of various product/service bundles, a goods-dominant approach (Vargo & Lusch, 2004), an Airbnb experience may lead to multiple perceptions of value. The same service may result in diverse levels of value for different people, both service providers and customers, and making them better off (co-creation) or worse off (co-destruction) (Plé, 2017). Value co-creation refers to the resource integration process between the provider and the customer (Vargo & Lusch, 2008), and all the actors involved in the process act to benefit from the interaction (Grönroos, 2012). Contrarily, value co-destruction is a failed interaction process that has a negative outcome. It leads to a decline in well-being, which can transform into frustration or lost resources (Echeverri & Skålen, 2011; Prior & Marcos-Cuevas, 2016).

There have been calls for studies to concurrently examine both aspects of interactive value formation (IVF), a neutral and integrative term that includes both positive (value co-creation) and negative value outcomes (value co-destruction) and its sources in the context of the sharing economy (Nadeem et al., 2020; Sthapit & Björk, 2018) and adopt a customer-centric perspective (Camilleri & Neuhofer, 2017). In addition, the inherently inconsistent nature of these services (Sthapit, Björk, & Barreto, 2020) and the interactive nature of service encounters may lead to service failures and service recoveries (Sparks & McColl-Kennedy, 2001). However, to date, there has been limited research on the impact of service recovery, a resource integration process that is carried out after a service failure to reclaim value to the greatest extent possible (Vargo & Lusch, 2004, 2008), from the perspective of value co-creation, also referred to as value co-recovery (Skourtis, Decaudin, Assioura, & Karaosmanoglu, 2018). Echeverri and Skålen (2011) referred to this as recovery value co-formation because of the recovery-like nature of the IVF from co-destruction towards co-creation of value.

The overarching aim of this study was to explore the drivers of value co-creation, co-destruction and co-recovery in the context of Airbnb, also referred to as value dimension. A value dimension is a certain feature of a service that can potentially contribute to the overall value for the actors or beneficiaries involved (Woodruff, 1997). The research question this study aims to answer is the following: What are the drivers of co-creation, co-destruction and co-recovery in the context of Airbnb from a guest’s perspective?

Literature review

Sharing economy and Airbnb

The sharing economy denotes a new type of business model that is driven by sharing access to underused goods and services to satisfy individual demand in exchange for payment or benefit (Yang et al., 2019). The sharing economy can also be understood as service systems (companies or individuals), through short-term rentals, utilising and monetising underutilised assets (Kumar et al., 2018). The sharing economy consists of many online businesses that use internet technology as a platform for lending, borrowing, gifting, swapping or renting consumer products and services (Germann Molz, 2013).

In recent years, Airbnb, an accommodation sharing economy system, has become quite popular among travellers (Jiang et al., 2019). Today, Airbnb is the world’s largest alternative accommodations provider, with more than three million listings (Dogru et al., 2020). Airbnb is an online platform that enables private households to profit by monetising their idle rooms, apartments and/or houses as tourist accommodation (Guttentag, 2019). In addition, Airbnb fulfils travellers’ needs, such as low-cost accommodations, convenient locations and a variety of choices (Sthapit & Jiménez Barreto, 2018b).

Growth of Airbnb in Nordic countries and the dominance of single-unit hosts

In Nordic countries, individually owned Airbnb rental properties are growing significantly (Adamiak, 2018). Single homes (single-hosted entire homes/apartments) form the majority that Airbnb offers in these countries, such as Denmark (mainly), followed by Sweden, Norway and Finland, except Iceland (mixed, a high balance of single rooms, single homes, multi-rooms and multi-homes) (Adamiak, 2019). Some recent studies indicated that, contrary to single-unit hosts, multi-unit hosts devoted more time and attention into the operations of their room-sharing business and thus were more proficient in serving guests (Kwok & Xie, 2019). In addition, multi-unit hosts were more experienced than single-unit hosts as they quickly learned from their operations of multiple listings at a given time and could identify more solutions to the problems that occurred in the operations. Single-unit hosts were categorised as unprofessional hosts and amateurs (Gibbs et al., 2018).

The concept of value and its outcomes (value co-creation, value co-recovery, value co-reduction and value co-destruction)

Zeithaml (1988, p. 14) defined value as “the consumer’s overall assessment of the utility of a product based on perceptions of what is received and what is given”. Others have defined value as “an improvement in system well-being” (Vargo et al., 2008, p. 149) and an interactive consumption experience (Zhang et al., 2018). In the context of this study, value refers to “a function of interactions between subjects or a subject and an object; is contextual and personal; is a function of attitudes, affections, satisfaction, or behaviourally based judgement; and resides in a consumption experience” (Holbrook, 2006, p. 212).

Value cannot be measured on a universal level; it can only be created with and determined by the user in the consumption process and through use (Lusch & Vargo, 2014) . In the same vein, value resides in a customer’s interactions with a firm’s offering – such as employees, facilities, goods and services rather than the firm itself (Park & Ha, 2016). Because every customer is unique in their consumption experience, skill, preference and goal, value is subjective to a consumption situation (Vargo & Lusch, 2004). The value formation process is interactively co-created by operant resources acting on operand resources or by operant resources in collaboration (Echeverri & Skålen, 2011). Operant resources are individually possessed resources which are generally invisible and intangible and shaped out of skills and knowledge, while operand resources include tangible assets, such as economic resources (Alves et al., 2016). In addition, it is customers – in this context, Airbnb guests – who actively co-create with companies and other customers, incorporate different resources and extract value in use (Nadeem et al., 2020).

Value co-creation represents the activities of consumers involved in direct interactions with firms to create value in use (Grönroos, 2011). Value in use is realised by a customer during a service process as a function of the customer’s experiences (Grönroos & Gummerus, 2014) and reflects the degree to which the customer believes they are better off or worse off through consumption experiences (Grönroos & Voima, 2013). Value is always co-created through consumers’ adoption of firms’ resources to consume the services, ultimately improving their well-being (Grönroos, 2011), and companies can only offer platforms for value creation, not create value as such (Lusch & Vargo, 2014).

Echeverri and Skålen (2011) discussed mixed cases, where interactions are characterised by the presence of both dimensions. On one hand, incongruent practices causing value co-destruction could become congruent, where the practitioners collaboratively contribute towards the recovery of value, resulting in the possible outcome of value co-recovery. According to Sthapit and Björk (2018), these are distinguished by how the interactions start and end. For example, if a guest expressed negative feelings about their stay, but evaluated the breakfast experience as excellent, then the latter aspect was a value recovery as the experience is understood through the totality of the accumulated practices. Studies have suggested that value co-recovery is interpersonal (Skourtis et al., 2018) and the outcome of such recovery is based on the actors’ assessment of value in their respective contexts (Edvardsson et al., 2012). This involves a collaborative restoration of service following a service failure to alleviate the negative emotions experienced by the harmed partner in the value network (Mostafa, 2016). Contrarily, practices starting with congruency and ending in incongruence take on the subjection position of value co-reduction (Echeverri & Skålen, 2011). Echeverri and Skålen (2011) have described this interaction as “reductive value co-formation” because of the diminishing nature of the interaction value shifting away from co-creation towards co-destruction of value.

Actor-to-actor interactions may also result in negative outcomes, whereby at least one of the actors (e.g. a customer) experiences a decline in the value realised from an interaction with another (e.g. an organisation) (Plé & Cáceres, 2010). This latter outcome of an IVF process has been termed value co-destruction (Echeverri & Skålén, 2011). Value co-destruction refers to “an interactional process between service systems that results in the decline in at least one of the service systems’ well-being (given the nature of a service system, can be individual or organisational well-being)” (Plé & Cáceres, 2010, p. 431).

Methodology

Data collection

The data for this study included reviews of Airbnb guests published on Trustpilot. Some recent studies have indicated that analysis of qualitative data posted online by tourists can play an important role in improving the understanding of the tourist experience (Thanh & Kirova, 2018). In the context of Airbnb, most guests reflected their stay experience in the reviews posted publicly on completion of their stay, which contain valuable information about their experiences (Lee et al., 2019).

For the data collection method, the study employed nonparticipant observation in the form of netnography. In addition, this study used a passive, covert approach in which the researchers did not interfere with the naturally occurring discussions or influence the study subjects (Arsal et al., 2010). Netnography is a type of virtual ethnographic research based on fieldwork to examine digital spaces, artefacts and intra-actions within an online community (Kozinets, 2015). Netnographic data are rich, more naturalistic, objective and unobtrusive (Wu & Pearce, 2014). As a research method, netnography is faster, simpler and less expensive (Kozinets, 2002); excels at telling the story and understanding complex social phenomena; and assists the researcher in developing themes from the consumers’ points of view (Rageh, Melewar, & Woodside, 2013). Moreover, data can be gathered from many sources, in this context, online tourist reviews, and analysed thematically. Furthermore, this study also acknowledges that netnographic studies face issues related to the authenticity and quality of the data material (Xun & Reynolds, 2010) and faces limitations in regard to generalising its findings to customer groups outside the online platforms or online communities studied (Kozinets, 2002).

We adapted Kozinets’ (2002) netnography procedure to the sharing economy context. The first step was the search for forums, online communities and distribution lists related to the focus of the study. The selection criteria used were based on the number of users or their level of activity (Kozinets, 2002). The greater the number of interactions found among these components, the more they were considered important sources of data (Dwyer, 2011). We selected online reviews about Airbnb guests’ experiences posted on Trustpilot because it fulfilled the selection criteria, a high number of users, interactivity and posts about their recent Airbnb experiences.

The second step involved data collection. Internet forums are appropriate contexts for collecting netnographic data when they fulfil the following five criteria (Kozinets, 2002): matches the research question, have a high traffic of postings, have a large number of discrete message posters, can provide rich data and have a high degree of between-member interactions. The data were gathered from Trustpilot, given that it fulfilled all the above criteria. Data collection and analysis were conducted between August 2019 and January 2020. The search words “good”, “excellent”, “great”, “nice” and “amazing” were used to capture online visitors’ narratives that indicated value co-creation and value co-recovery, while the keywords “awful”, “bad”, “worst”, “terrible” and “poor” were linked to value co-destruction. Research has identified emotions as a key customer resource in the value creation process (Malone et al., 2018). Service encounters, particularly failed ones, often result in negative emotions (Hollebeek & Chen, 2014), further leading to negative outcomes of value co-destruction (Echeverri & Skålen, 2011). Value co-destruction is founded on negative emotional experiences and has a substantial effect on the customer’s well-being (Malone et al., 2018). The 10 keywords helped to avoid generating overwhelming amounts of data. In addition, to improve the reliability of the data collected, postings in languages other than English were rejected. Each review post consisted of one entry and approximately four sentences. Of 1264 online posts screened overall, the analysis focused on 371 reviews based on the 10 keywords: good (34), excellent (12), great (25), nice (10), amazing (21), awful (37), bad (53), worst (77), terrible (40) and poor (62).

The third step in netnography was the ethics of the researcher’s role. Studies have suggested that when accessing review sites as a nonparticipant observer, there is no compelling need to communicate research objectives or obtain consent, as data are public on the web and the posts have often been made months or years in the past (Mkono, 2012).

Data analysis

The collected data were analysed using a grounded theory research design (Glaser & Strauss, 1967). The grounded theory facilitates the development of concepts, theories or models through continuous data collection and analysis and allows theoretical concepts to emerge from the data without being influenced by predefined frames of previously existing theories (Strauss & Corbin, 1990). In other words, grounded theory is an inductive research approach that intends to inform and develop concepts, theories or models that are grounded in participants’ data (Charmaz, 2004) without prior theoretical assumptions from the data collected (Glaser & Strauss, 1967). A three-stage procedure of open coding, axial coding and selective coding was adopted to analyse the collected data (Strauss & Corbin, 1990).

First, open coding mainly involves breaking down the data into distinct units of meaning (Strauss & Corbin, 1998) and comparing incidents to each other in the data (Glaser, 1998). Charmaz (2006) recommended selecting the most suitable analytical codes. At this stage, every line of each review post was carefully read and separately analysed to identify emerging ideas and views that repeatedly appeared and extract the reviewers’ views (Nunkoo & Ramkissoon, 2016) and specific information as well as each participant’s views (Sthapit, 2018). At this stage, coding was done in every possible way and asking a set of questions of the data. For example, “What is this data a study of?”, “What category does this review post indicate?”, “What is actually happening in a particular review post (data)?”, “What is the main concern being faced by the reviewers/participants (based on the ten keywords)?” and “What accounts for the continual resolving of this concern?” (Glaser, 1998). These questions sustain the researcher’s theoretical sensitivity, transcend descriptive details and encourage a focus on patterns among incidents that yield codes. This coding process forces the researcher to verify and saturate categories, minimise missing an important category and ensures relevance by generating codes with emergent fit to the substantive area under study (Holton, 2007). For example, using the ten keywords, this particular review captures the visitor’s narratives linked to value co-creation, value co-recovery and value co-destruction,

An issue happened with the reservation (the credit card), and we did not pay attention to the emails received to warn us. We found ourselves without any flat reserved, and it was definitely our fault. Airbnb did all they could to resolve this problem. Great customer service.

This was coded as great customer service. Tables 1 and 2 illustrates how the open coding (line-by-line coding) worked in practice. The first column of the table contains the raw data extracted from the transcripts, and the second column details the initial codes extracted from the raw data via line-by-line coding. In addition, the constant comparative process continued through open coding and involved three types of comparison. First, each review post (incident) was compared to other posts (incidents) to establish the underlying uniformity of generated themes. Then, emerging themes were compared to more review posts (incidents) to generate new properties of the themes. The purpose here was theoretical elaboration, saturation and densification of themes. Finally, emergent themes were compared to each other to establish the best fit between potential themes (Holton, 2007).

Table 1. The coding process in practise (main theme1: customer service).

Participants views (extracted from transcripts)	Open coding (line-by-line coding)	Subthemesn (axial coding)	Main theme (selective coding)
An issue happened with the reservation (the credit card), and we did not pay attention to the emails received to warn us. We found ourselves without any flat reserved, and it was definitely our fault. Airbnb did all they could to resolve this problem. Great customer service. Airbnb customer service has been extremely good. They are very prompt and efficient. The woman who helped me … has been extremely informative [and] has kept me updated. Thank you … for resolving my issue. I will not recommend Airbnb to anyone. Bad customer service. I had booked a place via Airbnb successfully. On the day of my arrival, the host told me that they deregistered from Airbnb and cannot accommodate me. I had to find a new place to stay late in the afternoon. On top of that, there were issues with my refund from Airbnb. Customer service was not very helpful. Worst customer service. They will protect their host because that is how they make money and will not consider the distress they put guests in … After hours on the phone and on messages, they do not understand what to do. Never using them again. Waste of time, money and patience.	“great customer service”, “customer service has been extremely good”, “excellent customer service”, “honestly impressed”, “great customer service”, “honestly impressed”, “Airbnb customer services has been extremely good”, “good customer service”, “Airbnb customer service is amazing”, “customer service we received from Airbnb is amazing” and “great customer service in times of need”, “bad customer support”, “issues with refund from Airbnb”, “customer service was not very helpful”, “terrible customer service”, “worst customer service”, “waste of time, money and patience”, “poor customer service, money taken with no booking”, “poor treatment by all customer service we have used”, “worst customer service representative who would not respond”, “worst customer service I have ever experienced”, “Airbnb customer service is the worst ever”, “poor customer service”, “customer service is so bad”, “worst customer service ever”, “customer service is bad and unhelpful”, “the worst customer service I have ever experienced in my life”, “customer service has been simply awful”, “the worst customer service ever”, “absolutely the worst customer service”, “terrible experience with this company’s customer service”, “the worst customer service I have seen” “customer service is truly awful”, “very bad customer service”, “the worst customer service I have ever seen” and “terrible, untrained customer service staff”	customer service, prompt and efficient customer service, non-existent, no service	customer service contributed to both value co-recovery and co-destruction

Table 2. The coding process in practise (main theme2: hosts’ actions).

Participants views (extracted from transcripts)	Open coding (line-by-line coding)	Subthemes (axial coding)	Main theme (selective coding)
Stayed in one bedroom hosted by … [an] excellent host. Everything provided – clean towels, comfortable bed, shower gel [and] even toothpaste and brush! Bread and milk in fridge and cereals, butter, jam, tea and coffee. House was well maintained and felt very relaxed … Highly recommend this property and will come back again. Good hosts … made us feel welcome and right at home. Nothing was a bother or too much trouble. Our accommodations were clean, cosy and quiet. The very best review I can give. We will book with this host again. It is a great place to stay. Booked a holiday apartment in Italy. Listing had whole apartment for rent. Pictures were included for all the bedrooms. After check in, we were given access to only one bedroom. The hosts locked [the] remaining bedrooms. Even [the] living room was locked. Even after requesting to open other bedrooms, [the] host refused. Airbnb complaints to the host was useless … Do not book with Airbnb. It is a scam which will ruin your family holiday. We were [a] family of four on holiday. Airbnb and their awful host ruined our holiday. Beware of bad hosts. This appears to be a way of buffering liability between the owners and customers. The properties we experiences … appear to be ok but … had significant health risks. By all accounts, the host do [sic] not seem to care … Property 1 had raw sewage in the drinking water, and property 2 had multiple issues, including long-term, present black mould. Stay at a resort or hotel.	“each host has been nice, honest and upfront”, “everything provided”, “excellent host”, “good hosts”, “great host”, “amazing host”, “what an amazing host”, “host dealt with any issues very well”, “host was friendly from the moment we booked”, “great hosts … they were accommodating”, “great host and was helpful with anything we needed”, “the host was great where I stayed” and “made us feel welcome and right at home” and “house is great and the host is great” “awful host ruined our holiday”, “bad host”, “beware of bad hosts”, “host misrepresented the property”, “absolutely terrible host and a dirty, dusty and stinky room in reality”, “disgusting, abusive and bad host”, “beware, bad hosts cancel without any consequence”, “host does not carte”, “host do not seem to care”, “bad host … very rude, disrespectful and dishonest”, “no communication from the host”, “host did not contact me”, “host does not reply”, “host did not want to know”, “host cancelled without any notice”, “host did not pick up the phone at all”, “host refuse to speak”, “host visited house without any notice”, “was stranded in the door with no response from the host”, “no response from the host”, “last minute cancellation ruined the trip”, “host cancelled … very frustrating”, “host cancelled at the last minute”, “cancellation notice from the host with no explanation”, “very last minute cancellation”, “beware of cancellation … Never again will I use Airbnb”, “cancellation is a total rip off”, and “scam cancellation”	everything provided, well maintained, easy to contact the host, fulfilled all the needs hosts’ inappropriate personal behaviour, lack of communication, last minute cancellation	hosts’ actions contributed to both co-destruction and co-creation

Second, following intensive open coding, axial coding is the act of linking categories to their subcategories along the lines of their properties and dimensions to form a more precise and complete explanation of the phenomenon (Strauss & Corbin, 1990). At this stage, the researcher tries to understand answers to questions such as “How?” and “Why?” by locating the phenomenon in its conditional context and denoting response action/interaction over time to certain problems and issues (Shojaeia & Haeri, 2019). In practise, the emergence of patterns during the open coding marked the beginning of the selective coding. At this stage, open coding was ceased and delimited coding to only those variables related to the core variable in sufficiently significant ways (Holton, 2007). This process involved continually relating subcategories to a category, comparing categories with the collected data, detailing the properties of the categories and exploring the variations in the phenomenon. This process continued until the researcher elaborated and integrated the core variable and its connection to other relevant categories. This core variable/category unites all the stands to explain the behaviour under study (Strauss & Corbin, 1990), in this context, the drivers of value co-creation, co-destruction and co-recovery in the context of Airbnb.

Third, selective coding involves the integration of the categories that have been developed to form the initial theoretical framework. This involves “selecting the core category, systematically relating it to other categories, validating those relationships, and filling in categories that need further refinement and development” (Strauss & Corbin, 1990, p. 116). In addition, the concepts and relationships that emerge from the coding processes are compared with the extant literature (Daengbuppha et al., 2006). At this stage during the analysis, no new properties, dimensions or relationships emerge (Strauss & Corbin, 1990). The axial coding principle employed was to explicate a story by identifying a core category and linking the other categories around the core category. However, identifying and committing to a storyline is not an easy task, as Strauss and Corbin (1990, p. 119) observed, “one is so steeped in the data that everything seems important, or more than a single phenomenon seems salient”.

Findings and discussion

The coding process inductively identified two value dimensions as drivers of value co-destruction. One of the dimensions contributed to value co-recovery, and the other engendered value co-creation.

Drivers

Driver 1: company’s customer service

A major theme that elicited both value co-recovery and co-destruction is Airbnb’s customer service. Of the 371 reviews posted online (102 positive and 269 negative reviews), 253 were linked to Airbnb’s customer service. More specifically, 69 online posts were positive, while 184 were negative. Some guests experienced positive customer service, while others felt that it their problems could not be solved in a timely and proper manner. “Worst customer service” was frequently mentioned in the review posts, revealing its significance in this context. Table 1 shows the coding process in practice and interpretive codes indicating customer service contributing to both value co-recovery and value co-destruction.

The following one positive and one negative post about guests’ Airbnb experience further highlights this theme:

I received an [sic] excellent customer service … I have been using Airbnb for a few years but recently became a more frequent customer of theirs … I had an issue in which the host of a place that I had booked was not giving me the directions to the rental. I told her I was waiting in the street with my luggage at the check-in time, but she asked me to call her later. I booked a hostel and got in touch with Airbnb about the problem I had experienced. The Airbnb customer support person … has been so, so helpful. Not only are they refunding the cost of the Airbnb but she also had me send her the receipt for the hostel I had to book so they can refund me for that as well. I am honestly impressed

Terrible customer service. I was denied staying at an Airbnb. I called customer service. I emailed them. I e-chatted with them and have gotten zero result [sic]. I had heard such good reviews, and I am so disappointed in attempting to use them. I will never use them ever again. If I could give a negative 10, I would. I am so upset. I have friends who host, and they cannot believe the terrible customer service I am receiving

Our findings showed that the collaborative engagement between the two parties, the Airbnb guest and the customer service agent, enacting the recovery was both successful and unsuccessful after the service failure. Thus, the IVF process resulted in value co-recovery and value co-destruction. Service failures can be defined as situations in which value is lower than expected because of the ineffective use of operand or operant resources (Skourtis et al., 2018). The review posts illustrated that in some cases, the customer service was prompt and adequate and the interaction between the actors proceeded smoothly in the co-recovery process. This was seen as being pleasant and some described the customer service agents as “prompt”, “efficient”, “so, so helpful”, “amazing”, “extremely helpful”, “fantastic”, “polite”, “knowledgeable”, “extremely informative”, “readily available”, “respectful”, “understanding”, “sympathetic” and “genuinely interested”. Because of the positive customer service experience, some mentioned that they would continue to use Airbnb for booking accommodations irrespective of the impending harm caused by the service failure. This is emphasised by the following interpretive codes: “will use Airbnb in the future”, “would use again”, “will definitely book again”, “definitely use them again” and “for sure will be booking through Airbnb”. The findings showed that customers (Airbnb guests) recover from value co-destruction when the provider implements appropriate techniques during the service recovery process. Some service recovery techniques include apologising, offering compensation and being polite in the process (Karatepe, 2006).

Studies have suggested that the Airbnb platform provides answers to frequently asked questions and offers 24/7 customer service to help guests and hosts resolve problems (Ju, Back, Choi, & Lee, 2019). However, several review posts showed that guests felt that Airbnb’s customer support failed to properly or promptly resolve their problems, for example, offering compensation during the service failure, which resulted in a second service failure. Many of these service failures were quite severe, for instance, not having a place to stay because of last-minute cancellation by the host. Guests experienced value co-destruction because of the misalignment of operant (information) and operand (financial compensation) resources by the customer support agent (service system) during the service recovery processes. This gave rise to negative emotions and further led to a decline in guests’ well-being. Among the different service recovery techniques, our analysis showed that some guests judged the monetary compensation as inadequate, while many indicated that Airbnb did not compensate them in any way. This was stressed by the following interpretive codes: “they will not refund”, “were unable to help me with a refund of any sort”, “Airbnb never refunded me”, “refused my refund”, “did not even refund me”, “are not refunding it”, “no refund in sight” and “denied refund”. As evidenced by our findings, this showed reluctance on the part of Airbnb to take responsibility for the experience offered by their service providers and offer a solution.

Given that the recovery process of Airbnb’s customer service failed to address many customers’ complaints, this intensified the bad experience, and consequently, some guests mentioned that they would never use the service again and even warned others from using it. This was apparent in the following interpretive codes: “goodbye Airbnb”, “stay away”, “sadly our last”, “beware of Airbnb and go with Vrbo”, “do not book with Airbnb”, “avoid dealing with this company”, “will not be using Airbnb again”, “definitely not book over Airbnb again” and “will never use this site again”. These findings were in line with studies that showed poor customer service from Airbnb as one of the major complaints posted about Airbnb online (Phua, 2019) and that ineffective service recovery strategies could trigger customer-switching behaviour (Roos, 1999) and consumer discontinuance (Sthapit & Björk, 2019b). In addition, these findings emphasised the important role of service recovery staff in the IVF processes that lack of or insufficient customer support hindered value co-recovery. Some studies have suggested that empowering service representatives was a basic essential for a successful co-creation experience (Prahalad & Ramaswamy, 2004). Moreover, the findings in this paper stood in contrast to some studies, which indicated that hosts caused much of guest frustrations with home-sharing businesses such as Airbnb (Shuqair et al., 2019).

Driver 2: hosts’ actions

Another significant theme that appeared from our analysis was Airbnb hosts’ actions as the source of both positive and negative outcomes of the IVF process (value co-creation and co-destruction). Of the 371 reviews posted, 118 (31.80%) emphasised Airbnb hosts’ actions. Some guests perceived the host’s actions as positive, while others considered the actions as a predominant dissatisfier that did not fulfil Airbnb’s resource (value proposition) and their expectations. In fact, 33 online posts were positive, while 85 were negative. Table 2 demonstrates the interpretive codes around Airbnb hosts’ actions contributing to value co-destruction and co-creation.

The following one positive and one negative review describing Airbnb guests’ experiences can be linked to the incontinent nature of Airbnb hosts’ actions:

I have loved all the hosts that I have rented from. Each host has been honest and upfront with how to access the space, the rules of the house, and what parking has been available. They have made it extremely easy for me to contact them, and they have been great at responding to my message. Overall, it was a superb experience.

I booked with Airbnb well in advance in a college town for a homecoming football game weekend because I knew there would be no place to stay. The host cancelled my reservation three weeks before my stay, leaving me high and dry without a place to stay … This bad host. Do not ever book with this company

Hosts play an important role in Airbnb service experience because consumers value experiences with hosts who are friendly, conscientious, and responsive (Lyu, Li, & Law, 2019). However, an individual host’s actions are not homogenous. This core theme can be linked to service quality attributes, for example, behaviour (Grönroos, 1990), communication, courtesy, friendliness and responsiveness (Parasuraman et al., 1988). Service quality is the tourist’s subjective assessment of the interaction with the host and how well the tourist’s service needs have been met (Dabholkar et al., 2000). Customers’ perceptions of service experience are coloured by both the outcome of the process and, often even more important, how the process functions (Grönroos, 2008).

As shown in this study, the service quality of an Airbnb experience is dependent upon each host’s hospitality capability (Lalicic & Weismayer, 2018). In addition, Airbnb guests are more likely to experience varying service quality (Sthapit, 2018) and no service delivery guarantee (Huang, Coghlan, & Jin, 2020) because Airbnb hosts are not trained hospitality professionals (Birinci et al., 2018), leading to both good and bad experiences (Lee et al., 2019). Moreover, the host’s inappropriate actions are contrary to the positive service attributes linked with hosts in the peer-to-peer marketplace, namely, understanding and caring (Priporas, Stylos, Rahimi, & Vedanthachari, 2017) and hospitality hosting behaviour (Lalicic & Weismayer, 2018). The identified theme can be categorised as a functional and an important operant resource in the IVF processes. Contrary to this finding, studies indicated that Airbnb guests enjoyed a more personalised service quality (Mao & Lyu, 2017) and experienced remarkable customer satisfaction levels (Ert et al., 2016).

Apart from some other factors (accommodation transactions, the feeling of togetherness, feedback and social approval), value is co-created by guests and the Airbnb host through communication and interactions (Zhang et al., 2018). However, this was not the case for all the guests. When guests experienced problems, they typically attempted to resolve the situation by seeking support in the form of communicating with their host. In some cases, the host–guest interaction was successful, and guests experienced value because of the host’s promptness in responding to their requests. However, several review posts illustrated the incongruent application of resources (inadequate communication) by one service system (involving multiple Airbnb hosts). Consequently, the IVF process resulted in value co-destruction for several Airbnb guests, and Airbnb’s resource offer (value proposition) of “feel at home wherever you go in the world” did not hold true for them. Unpleasant actions, such as “contacted the host and got no reaction”, “no communication from the host”, “host did not contact me”, “host does not reply”, “host did not want to know”, “host cancelled without any notice”, “host did not pick up the phone at all”, “host refuse to speak”, “host visited house without any notice”, “was stranded in the door with no response from the host”, “no response from the host”, “host did not respond to messages” and “host have both been ignoring us and unwilling to help”, constitute evidence of multiple hosts’ lack of communication. These are contrary to value co-creation behaviours, for example, information sharing, responsible behaviour and personal interaction (participation behaviours) including feedback and helping (citizenship behaviours) (Yi & Gong, 2013).

The subsequent uncertainty and lack of clarity resulted in a decline in the guest’s well-being because of the unexpected resource loss (emotional, financial and temporal). This is highlighted by the following interpretive codes: “extremely disappointed and frustrated”, “hostile situation between a host and a guest”, “waiting for replies and getting stressed of not having an alternative”, “host charged us extra”, “host charged me $800 for the items missing”, “had to deal with the host and her husband for over 4 hours” and “booked and waited 48 hours for the host to answer. No answer”. Few reviews related to successful communication between the host and the guest were highlighted by the following interpretative codes: “host quickly replied”, “host’s communication is praiseworthy”, “host contacted me back and was quickly resolved”, “host was very attentive”, “host has been great in responding to my message” and “host stayed connected with me the whole time”. One of the causes of asymmetric quality of communication from the host can be attributed to the lack of site-wide hospitality standards (Sthapit & Björk, 2019a). This finding supported recent studies indicating that bad behaviour from the host was a common cause of negative Airbnb experiences (Sthapit & Jiménez Barreto, 2018a) and that value co-destruction emerged from inadequate communication (Vafeas et al., 2016).

Besides, a lack of communication, reports of unethical actions by the host, particularly Airbnb hosts cancelling bookings at the last minute, also generated value co-destruction among the guests and resulted in a decline in their well-being. The following interpretive codes provide relevant examples: “last minute cancellation ruined the trip”, “host cancelled … very frustrating”, “host cancelled at the last minute”, “cancellation notice from the host with no explanation”, “very last minute cancellation”, “beware of cancellation … Never again will I use Airbnb”, “cancellation is a total rip off” and “scam cancellation”. While Airbnb has been compared to lower-end hotels (Zervas et al., 2017), this type of behaviour rarely occurs in the conventional accommodation sector, for example, in a standardised hotel booking process (Thornton & Ocasio, 1999).

Evidently, the host’s friendly behaviour, including prompt communication between the host and the guest, led to value co-creation. However, incongruent resource applications by the host, such as inappropriate actions by the host (including a lack of communication and last-minute cancellation) contributed to value co-destruction. This suggested that maintaining a two-way communication channel throughout the service was important for reducing the incongruent application of resources. In addition, given the lack of adequate communication between the host and the guest following an incident, co-work opportunities do not arise for either actor; hence, there is no co-recovery of value. This finding contributed to the understanding of finer-grained value dimensions contributing to IVF outcomes in the context of Airbnb. Overall, this finding supported recent studies which indicated that the host’s friendly behaviour (Ju et al., 2019) and host–guest interaction was of utmost importance in Airbnb guest experience (Lalicic & Weismayer, 2018; Sthapit & Jiménez Barreto, 2018a).

Conclusion

This study provides three main contributions. First, the present study includes the drivers of value co-creation, co-destruction and co-recovery in the context of Airbnb: customer service and hosts’ actions. Previous studies have not examined all three aspects simultaneously, making this study more holistic. To the best of our knowledge, this study is the first of its kind to examine the drivers that contributed to the three aspects in the context of Airbnb, linking these concepts to service failure and service recovery. Thus, from a theoretical perspective, this research extends the literature on the sharing economy and IVF.

Second, the findings showed that not all service failures lead to value co-destruction and collaborative engagement involving a congruent application of resources, particularly operant resources, between actors following a service failure results in value co-recovery in the context of Airbnb. Some Airbnb guests recovered from the service failures and perceived the service recovery efforts of the customer service agents as positive in terms of achieving the outcome they desired. This highlights that operant resources are crucial in producing favourable experiences and solving customers’ problems during IVF processes. The two value dimensions (knowledge and skills) this study has pinpointed are important operant resources in the Airbnb IVF process.

Third, another important finding is that particular value dimensions can individually act as a source of value co-creation, value co-recovery and value co-destruction for the customer (Airbnb guest). For example, Airbnb hosts’ actions contributed to both positive and negative outcomes of the IVF process (value co-creation and co-destruction). In addition, successful recovery efforts by Airbnb customer service agents after service failure served as an antidote to value co-destruction, thereby contributing to value co-recovery. Conversely, the same value dimension generated value co-destruction when the customer service agents were unable to resolve customers’ (Airbnb guests) problems in a timely and proper manner and service recovery strategies did not satisfy their needs. Thus, in the context of this study, customer service is the source of value co-recovery and co-destruction.

Lastly, the findings supports recent studies indicating that the sharing economy model with fragmented supply chain, in this context, multiple Airbnb hosts, does not always result in improved service quality and customers (Airbnb guests) can experience difficulties in service failures and insufficient recovery strategies (Furunes & Mkono, 2019).

Managerial implications, limitations and future research

The results of this study have several managerial implications for Airbnb and its hosts, particularly in the context of Nordic countries that are largely dominated by single-unit hosts, who are not trained hospitality professionals, guests are likely to experience value co-destruction because of the inappropriate actions of the hosts as identified in this study. Thus, first, hosts who are frequently reported as unresponsive, unhelpful and unprofessional towards guests should be banned from hosting on Airbnb. Second, hosts must be held accountable in situations where things go wrong because of them, for example, if guests complain about a lack of communication and last-minute cancellations by the host. This involves clearly defining hosts’ tasks/responsibilities and educating them about service quality standards and effective service recovery strategies, including how they should be offered. When hosts are provided with clear responsibilities, customers can be served efficiently, and this reflects service presence.

Third, there remains a need for Airbnb to develop a strict policy on service failure and establish clear and efficient procedures for service recovery. In addition, both hosts and customer service agents should be provided service recovery training for providing responsive, caring and professional service. In the same vein, appropriate, efficient, well-enacted and timely monetary compensation – for example, travel credits, a discount towards future travel – could help to remedy guests’ compromised value perception of the purchase. Moreover, the presence of the host and Airbnb customer service personnel is important to respond to the guest during service failures. Fourth, given the lack of communication by the host during the service delivery process, the current policy on service failure for Airbnb, which requires guests to first contact the host/property owner/manager directly and only if the guest and host cannot reach an agreement on resolving the issue are guests asked to contact the platform, needs to be changed. The customer must be able to contact the platform during service failures.

This study has some limitations. First, this study employed netnography and was restricted to customers who shared their reviews online. Second, the data used in this study is biased as it focussed on customer reviews written in English only. Third, the number of keywords used in the data analysis and the number of screened review posts were limited. Although the findings of this study have a theoretical and practical salient explanation for the drivers of value co-creation, co-destruction and co-recovery, precautions should be taken in generalising the findings. Future studies should examine both hosts’ and guests’ perspectives to further augment the findings of this study and provide a broader understanding of the sources of IVF outcomes.

Disclosure statement

No potential conflict of interest was reported by the authors.