On the same boat but singing a different tune: Coopetition between hotels and platforms close to customers

Coopetition — consisting of concurrent cooperation and competition — mainly focuses on business activities far from the customer, such as research and development. However, coopetition close to the customer, comprised of marketing and sales, remains an under-researched area. Applying grounded theory, we investigate how hotels and platforms (e.g., Expedia.com and Booking.com) pursue coopetition in close customer proximity. Our findings suggest that, contrary to insights from the literature, coopetition is not only competition dominant close to the customer. Rather, coopetition patterns range between cooperation- and competition-dominant approaches, depending on tension levels. Further, in contrast to current views, separating cooperation and competition over time is impossible when episodes of cooperation and competition are short. Instead, hotels utilize multiple spaces (platforms, physical hotels, and direct channels) to isolate the two forces.


Introduction
Platforms are hubs that connect businesses and customers who would otherwise find it challenging to find each other (Hänninen & Smedlund, 2021;McIntyre & Srinivasan, 2017). For instance, the Airbnb platform connects hosts and guests who would not be able to connect otherwise (Huang, Coghlan, & Jin, 2020;Volgger, Taplin, & Pforr, 2019). Such platforms have disrupted many industries. For example, in the hospitality sector, Expedia.com and Booking.com have displaced traditional travel agents (Verhoef & Bijmolt, 2019;Verhoef, Kannan, & Inman, 2015). In fact, hotels generate more than 50% of their revenue via platforms (Rossini, 2015). Further, Holidayhomes.com has shifted the public preference from staying at hotels to booking holiday homes. Platforms are also transforming other industries including entertainment (e.g., Netflix, Amazon Prime Video), taxi services (e.g., Uber, Ola), and food services (e.g., Opentable, Foodpanda). They are therefore a vital interface, enabling businesses to reach out to customers and earn revenue (Gao & Bi, 2021;Ling, Guo, & Yang, 2014). However, businesses need to be mindful of the bargaining power of platforms, as they charge high fees, which reduces profits (Gawer & Cusumano, 2014;Sharma & Nicolau, 2019). Thus, effectively managing relationships with platforms is fundamental for business success.
However, most coopetition research focuses heavily on business activities far from the customer interface; for instance, in research and development (Chen, Tang, Wu and Wang, 2021;Gnyawali & Park, 2009, 2011Jakobsen, 2020;Park, Srivastava, & Gnyawali, 2014;Yang, 2020). As a result, there is a limited empirical literature on coopetition close to the customer in areas such as distribution, pricing, promotions, branding, and customer service (e.g., Lindström & Polsa, 2016). In addition, although the coopetition literature explains the business activities near the customer (Bengtsson and Kock, 2000;Lindström and Polsa, 2016), it does not explicate whether and how the coopetition approach varies in different situations.
Further, research insights reveal that coopetition far from the customer is cooperation focused, whereas coopetition near the customer is an arena for competition (Basterretxea, Charterina, & Landeta, 2019;Bengtsson & Kock, 2000;Walley, 2007). For example, competitors may collaborate in a joint research project to develop new products (far from the customer) but compete in marketing products and building a loyal customer base (close to the customer). However, coopetitive relationships may also include collaboration near the customer (Lindström & Polsa, 2016;Ritala & Tidström, 2014). For instance, in hotel-platform relationships, collaboration occurs entirely near the customer because platforms are a marketing interface for hotels to connect with customers. The literature does not provide sufficient knowledge on coopetition close to the customer that also consists of collaborative actions.
Based on the above background, and in response to calls for more research on coopetition near the customer (Lindström & Polsa, 2016) and coopetition between hotels and platforms (Chang et al., 2019), we developed the following research question: How do hotels pursue coopetition in close customer proximity when working with platforms? We used grounded theory to address our research question (Corbin & Strauss, 2008;Strauss & Corbin, 1990), gathering primary data from the hospitality sector. Twenty-three hotels and two platform firms took part in our study, in which we interviewed 46 managers.
We contribute to the coopetition literature by showing that, contrary to the current understanding, coopetition is not always competition dominant in close customer proximity. Rather, coopetition ranges between cooperation and competition-dominant approaches depending on tension levels. In turn, tension levels depend on (i) the intensity of cooperation and competition, and (ii) whether firms play complementary and/or incompatible roles in business relationships. Further, although the literature suggests tension is managed by separating cooperation and competition over time, we find that such an approach is impossible when the time frames between cooperation and competition are short. Instead, hotels separate the collaborative and competitive forces into different spaces or areas: platforms, physical hotels, and direct hotel channels. Overall, we offer novel insights into coopetition approaches that enable hotels to work successfully with platforms.
The rest of the paper is organized as follows. First, we provide the background to coopetition, tensions, and tension management, and argue the need for more research on coopetition between hotels and platforms close to the customer. Second, we describe our research method based on grounded theory. We explain our sampling, data collection, and data analysis approaches, and how we ensured the trustworthiness of our findings. Third, we present our findings and illustrate our conceptual framework. Fourth, we discuss the theoretical and managerial contributions of our study. Finally, we describe the limitations of our work and identify areas for future research.

Coopetition: Current perspectives and schools of thought
Coopetition is defined as the ability of a firm to concurrently enact cooperation and competition with another firm (Bengtsson & Kock, 2003;Fong et al., 2021). It represents a "dyadic and paradoxical relationship that emerges when two firms cooperate in some activities, […] and at the same time compete with each other in other activities" (Bengtsson & Kock, 2000, p. 412). Drawing on paradox thinking, coopetition can be best understood based on the existence of two separate continua: one focusing on cooperation, and the other on competition (Akpinar & Vincze, 2016;Bengtsson, Raza-Ullah, & Vanyushyn, 2016). This suggests three key aspects of coopetition. First, coopetition includes the simultaneous involvement of cooperation and competition in business relationships (Bengtsson and Kock, 2000;Gnyawali & Park, 2011). In other words, both contradictory forces need to be present at the same time (Chiambaretto, Fernandez, & Le Roy, 2022). Second, coopetition can incorporate different degrees cooperation and competition, e.g., low cooperation and low competition, high cooperation and low competition, or high cooperation and high competition (Bengtsson, Eriksson, & Wincent, 2010;Lado, Boyd, & Hanlon, 1997). Third, coopetition is thus not a trade-off between cooperation and competition , where if the degree of one force increases, the other decreases, because that neglects the paradoxical nature of coopetition (Bengtsson et al., 2010;. There are two distinct schools of thought relevant to exploring coopetition: the (i) actor and (ii) activity schools of thought Raza-ullah & Kostis, 2020). The actor school of thought takes a broad perspective, considering all the actors in the value net, such as competitors, suppliers, complementors, and distributors, and examining their interrelationships (Brandenburger & Nalebuff, 1996;Damayanti, Scott, & Ruhanen, 2017;Rusko, 2014). In this case, firms develop cooperative relationships with one set of actors, and competitive relationships with another set of actors (Hoffmann, 2007;Sytch & Tatarynowicz, 2014). However, this stream of work does not take into consideration the simultaneous cooperation and competition occurring within the same business relationship (Bengtsson et al., 2010;. Rather, its focus is on how firms develop either cooperative or competitive relationships with other actors (Hannah & Eisenhardt, 2018;Pattinson, Nicholson, & Lindgreen, 2018). In addition, this school of thought suggests that actors first pursue collaboration to create value and later compete to capture value in return (Damayanti et al., 2017). The analogy of first baking the cake together (cooperation) and afterwards taking the bigger slice of the cake (competition) is often used to explain the phenomenon (Chen, Luo, & Wang, 2019). As a result, the two contradictory forces of cooperation and competition are divided over time rather than simultaneous in nature, i.e., they do not occur at the same time.
In contrast, the activity school of thought takes a narrower perspective to focus on one-to-one business relationships (Bengtsson and Kock, 2000;Gnyawali & Park, 2011;Chang et al., 2019;Chen, Yao, Zan, & Carayannis, 2020). In this case, firms simultaneously pursue one set of business activities that are cooperative in nature, and another set of activities that are competitive (Le Roy et al., 2021;Tidström, Ritala, & Lainema, 2018). In this perspective, business relations are not considered coopetitive when firms (i) cooperate with one set of actors and compete with another set of actors, and/or (ii) separate collaboration and competition over time . Instead, for a relationship to be considered as coopetition, firms must enact cooperation and competition activities with the same business partner at the same time Gnyawali & Charleton, 2018). In the present paper, we adopt the activity school of thought to explore hotel coopetition with platforms. However, because hotels pursue coopetition with a portfolio of platforms, we do not restrict our analysis to one-on-one relationships (e.g., Hilton hotel and Expedia. com). Instead, we explore concurrent cooperation and competition business activities between many hotels and platforms. Our perspective echoes Bengtsson and Kock's (2014) approach of examining coopetition between two or more actors because there are multiple firms involved in this process.
The coopetition phenomenon can also be examined from a strategic (Chen et al., 2019) or relational perspective (Le Roy et al., 2021;Razaullah, 2021). These perspectives are distinct yet complementary in nature. The strategic approach focuses on decisions about when to enter into coopetition, with which partners, and for how long (Fernandez, Chiambaretto, Chauvet, & Engsig, 2021;Kraus, Meier, Niemand, Bouncken, & Ritala, 2018). In contrast, the relational approach considers how firms maintain concurrent cooperation and competition, and the tension management approaches needed to sustain such business relationships (Gnyawali, Madhavan, He, & Bengtsson, 2016;Le Roy & Czakon, 2016). The current research primarily focuses on relational coopetition to explore how hotels manage an ongoing relationship with multiple platforms.
In vertical coopetition, however, firms exchange resources because they are involved in supplier-seller relationships (Lacoste, 2012;Lechner, Soppe, & Dowling, 2016;Rajala & Tidström, 2021). Such firms do not compete directly in the same market but indirectly through other avenues Dowling et al., 1996). For example, while car rental firms cooperate with Expedia.com to sell their service, they also compete by bypassing the platform using their direct channels. In this case, car rental firms do not compete with Expedia.com in the car rental market directly, but indirectly in the distribution avenue.
Therefore, horizontal coopetition is different from vertical coopetition, as it depends on with whom a firm pursues coopetition (e.g., a direct competitor, supplier, or seller); in which market/avenue coopetition occurs (e.g., same market or distribution avenue); and how firms use their resources (e.g., resource combination or resource exchange).
However, it is sometimes challenging to define a business relationship as exclusively horizontal or vertical . This is because in today's business environment, firms play multiple roles, such as supplier, manufacturer, buyer, competitor, or business partner Johansson, 2012). As a result, the roles played by firms are constantly changing and evolving (Dowling et al., 1996). For instance, a supplier of computer processers for Asus or Dell can decide to produce computers themselves and enter the desktop and laptop market. In this case, the supplier holds the dual role of both supplier and also competitor for Asus or Dell, and hence the relationship becomes vertical as well as horizontal in nature. Bengtsson and Kock (2014) provide an updated definition that solves the issue of restricting coopetition to only horizontal or vertical relationships. They define coopetition as "a paradoxical relationship between two or more actors simultaneously involved in cooperative and competitive interactions, regardless of whether their relationship is horizontal or vertical" (p. 182). As a result, our study explores the coopetitive relationship between hotels and platforms based on the roles played by each actor. Here, hotels play the role of a supplier and distributor of room inventories, whereas platforms perform the role of a distributor and data partner.

Coopetition in close customer proximity
Firms can effectively pursue coopetition by separating the forces of cooperation and competition depending on the closeness of their business activities to customers (Gast, Gundolf, Harms, & Matos Collado, 2019;Hoffmann et al., 2018;Rusko, 2011;Tidström et al., 2018;Walley, 2007). Business activities far from the customer include research and development, procurement, inbound logistics, and manufacturing. Business activities near the customer include marketing, distribution, sales, and customer service. Research shows that competitors cooperate in activities far from the customer but compete in activities near the customer (Basterretxea et al., 2019;Bengtsson & Kock, 2000;Tidström, 2009). For instance, in the study conducted by Erzurumlu (2010), competing suppliers collaborate to develop innovative products but compete with each other in marketing the products. Nalebuff and Brandenburger (1997) cite a real-life example from the media industry where two firms cooperated in printing newspapers to make full use of their production capacity but competed for advertisers and customers.
Evidence from empirical studies confirms that competitive pressures increase as coopetition moves closer to the customer. For instance, Fernandez, Le Roy, and Gnyawali (2014) explore coopetition between two manufacturers of telecommunication satellites, who pooled their resources to develop a new range of satellites. However, during the commercialization stage, competition between the manufacturers increased as the newly developed satellites were competing in the market with the existing range of satellites sold independently by the manufacturers. In other research, Tidström and Hagberg-Andersson (2012) explain how a relationship becomes competitive once suppliers begin to bypass the manufacturer to sell finished goods directly to customers. Therefore, coopetition close to the customer is considered a field of competition (Basterretxea et al., 2019;Bengtsson & Kock, 2000).
However, some empirical work has challenged the view that coopetition shifts towards competition near the customer (Lindström & Polsa, 2016;Ritala & Tidström, 2014;Tidström et al., 2018;Tidström & Hagberg-Andersson, 2012). For instance, Tidström and Hagberg-Andersson (2012) find that coopetition changes over time from collaboration, to coopetition, and later to competition. In another study, Ritala and Tidström (2014) show that coopetition in the area of sales varies depending on the geographical market: competitors collaborate in markets where their product offerings are diverse, but pursue a mix of collaborative and competitive actions in markets where there exists a degree of overlap in their product range. Yet in another study by Tidström et al. (2018), a company collaborates with a business partner to utilize their sales channel to sell their products in a new market, but the collaborative relationship also becomes competitive when the business partner starts producing and selling similar products. In our case, hotels cooperate with platforms to gain access to a broad customer base and sell their room inventory (Ling et al., 2014;Yang & Leung, 2018). However, hotels also compete with platforms because they charge high fees, which considerably reduces hotel profits (Bahar, Nenonen, & Starr Jr, 2021;Verhoef & Bijmolt, 2019). Platforms also attract customers away from hotel websites, reducing the chance of hotels generating direct bookings. Therefore, there is conflicting evidence regarding coopetition near the customer. One plausible explanation for this is that coopetition may vary depending on the context. We now move on to reviewing the literatures on paradox thinking and actor roles to understand how coopetition might vary in different circumstances.
Drawing on the paradox literature, the level of coopetitive tension depends on the degree of cooperation and competition in (i) cooperation-dominant, (ii) competition-dominant, or (iii) balanced (strong, moderate or weak) coopetitive situations (Bengtsson & Kock, 2000;Bengtsson & Raza-Ullah, 2017;Lado et al., 1997;Park et al., 2014;Smith & Lewis, 2011). Cooperation-dominant situations create low tensions (Bengtsson et al., 2010) because the pull for cooperation is stronger than the pull for competition (strong cooperation and weak competition) (Wilhelm & Sydow, 2018). Managers therefore do not feel stressed by having to handle two strongly opposing forces (Raza-Ullah, . Instead, they are pulled towards cooperation. For example, smaller hotels and motels rely more on platforms such as Booking.com and Expedia.com to attract visitors (cooperation) rather than aggressively promoting their website through search engine advertisements to generate direct bookings and rival these platforms (competition).
Competition-dominant situations also foster low tensions (Bengtsson & Raza-Ullah, 2017). In this case, the pull for competition is greater than the pull for cooperation (high competition, low cooperation) (Akpinar & Vincze, 2016). As a result, managers are drawn towards competition and do not need to manage two contradictory forces with high intensity (Raza-Ullah et al., 2014). In the hospitality sector, larger hotels that are part of globally known chains invest heavily in generating direct bookings via their direct channels (website and mobile app) and focus on building a loyal customer base to compete with platforms like Hotels. com and Agoda.com (competition). As a result, these hotels are less reliant on platforms to generate bookings (cooperation).
Finally, balanced coopetitive situations occur when neither cooperation nor competition is the dominant force (Bahar et al., 2022;Gnyawali & Charleton, 2018). Balanced coopetition comprises three categories: strong, weak and moderate (Lindström & Polsa, 2016). Balanced-strong coopetitive situations lead to high tensions because the paradoxical condition encompasses high cooperation and high competition (Park et al., 2014). Under such circumstances, managers are challenged by two opposing, high-intensity forces (Raza-Ullah, Bengtsson, & Vanyushyn, 2018). This results in managerial difficulty in making effective decisions, thus requiring additional support within the firm to deal with the paradoxical situation (Bengtsson & Raza-Ullah, 2017). For example, extremely high tensions arose when two strong rivals -Sony and Samsungworked together on innovation projects (Gnyawali & Park, 2011). This paradoxical situation meant that although there were high levels of rivalry, the two giants had to find a common ground to work together and innovate.
In contrast, balanced-weak coopetitive situations result in low tensions because the intensity of cooperation and competition is low, thus creating less stress (Bengtsson & Raza-Ullah, 2017;Santos, 2021). Managers in these situations spend less time and effort exploring possible opportunities (cooperation) and threats (competition) in business relations (Bengtsson & Raza-Ullah, 2017). For instance, tourism firms located in Australia or New Zealand do not primarily target tourists from Nepal or Bangladesh, as there are comparatively low numbers of visitors from these countries. As a result, while these firms may work with a few travel agents who mainly cater to these markets, they will maintain a distant relationship and devote minimal time, effort, and resources to pursuing coopetition with those agents.
Balanced-moderate coopetitive situations create moderate tensions because the intensity level of the cooperation and competition is optimal: neither too strong nor too weak (Chen, 2008;Das & Teng, 2000). Such a manageable degree of tension motivates managers to make appropriate decisions that blend cooperation and competition to "get the best of out of both worlds". For example, many hotels have realized the benefits of focusing on platform sales (cooperation) and direct channel sales (competition) to attract customers and attain optimal profits.
While the paradox literature offers crucial insights, considering the concept of role adds a complementary perspective. Drawing on role theory (Biddle, 1986;Homburg, Müller, & Klarmann, 2011;Solomon et al., 1985), we suggest that tensions depend on the functional roles played by firms during coopetition. Role theory proposes that the social interaction between firms in a business relationship is based on the role each actor plays within a certain context (Heide & Wathne, 2006;Michaels, Day, & Joachimsthaler, 1987). We apply Herrmann, Jahnke, and Loser's (2004) definition of role as "the sum of all behaviour expectations of a social system towards a concrete role actor. The role actor is in a certain position, which is linked to tasks and functions" (p. 168). In our case, hotels are expected to perform the functional roles of (i) supplier and (ii) distributor. As suppliers, hotels enact specific tasks such as providing accommodation services, managing room inventory, and setting rates (i.e., pricing). As distributors, hotels, for instance, sell their room inventory via direct channels (e.g., hotel website and mobile app).
In contrast, platform actors (e.g., Expedia.com) are also expected to perform the functional roles of (i) distributor and (ii) data partner. As distributors, platforms are expected to sell room inventory to customers and offer additional support to make the process as efficient as possible (e.g., providing a chat function for hotels and customers to communicate with each other). However, lately, these platform actors have become more than just distributorsthey are data partners offering analytics dashboards for hotels to make informed decisions.
We propose that coopetitive tensions arise when firms must simultaneous play (i) complementary and (ii) incompatible roles. Complementary roles between firms (e.g., supplier-distributor) focus on cooperation, whereas incompatible roles (e.g., distributor-distributor) emphasize competition. Therefore, performing complementary and incompatible roles concurrently may become challenging to handle, creating friction between the roles and fostering moderate to high levels of coopetitive tension. In contrast, when firms lean towards either complementary or incompatible roles, the tension levels are low. This is because the roles either focus heavily on cooperation and less on competition (complementary roles), or less on cooperation but more on competition (incompatible roles). As a result, firms find it comparatively less difficult to perform their roles.
The paradox thinking and actor role approaches also augment our understanding of the differences between coopetitive and competitive tension Raza-ullah, 2020). The former occurs when firms face the contradictory influence of cooperation and competition (Le Roy & Fernandez, 2015; because they have to play both complementary and incompatible roles. However, competitive tension is related to threat, conflict, and risk in business relationships (Dahl, 2014;Raza-ullah, 2020) based predominantly on incompatible roles. In this case, managers are not dealing with the paradox of cooperation and competition. Instead they are pulled towards competition as their respective firms are focused more on incompatible roles. The research focus in the current study is coopetitive rather than competitive tensions. We next review the tension management literature to gain insight into how coopetitive tension can be managed close to customers.

Tension management: Approaches and capabilities
Tension management approaches offer insight into how to handle the paradox of cooperation and competition, whereas capabilities relate to what skill sets are required by firms to manage the resulting tensions. Table 1 provides definitions of both tension management approaches and capabilities.
Separation and integration of cooperation and competition underpin approaches to managing tensions. Separation is pursued by staff members who work in lower levels of the organizational hierarchy, whereas integration is implemented by their managers in the upper levels Das & Kumar, 2010).
Spatial separation relates to isolating cooperation and competition in different domains or locations to reconcile these two opposing forces (Joseph et al., 2020;Poole & de Ven, 1989). Firms may separate the two forces into different value chain functions (Le Roy & Czakon, 2016;Oinonen et al., 2018). For example, firms tend to cooperate in upstream activities (e.g., new product development) but compete in downstream activities (e.g., sales) (Kylänen & Rusko, 2011;Ritala & Tidström, 2014). Alternatively, the two forces may be divided based on product/service areas or markets (Fernandez et al., 2014). For example, two tourism destinations may collaborate to jointly organize an event but compete in other areas such as accommodation. In the dairy industry, firm A may supply yogurt to firm B if firm B does not produce yogurt (Bengtsson & Kock, 2000), but the two firms may compete in other related product areas such as milk and cheese. Further, cooperation and competition can be split based on geographical proximity and location (Fernandez & Chiambaretto, 2016;Hoffmann et al., 2018). For instance, a tourism destination may collaborate with a nearby destination but compete with a faraway destination, or two firms may collaborate in Asia but compete in Europe.
Temporal separation refers to dividing cooperation and competition over time, which creates a sequential pattern (e.g., first cooperation, later competition) (Damayanti et al., 2017;Smith & Lewis, 2011). In other words, temporal separation is achieved by eliminating overlap between cooperation and competition (Lavie, Kang, & Rosenkopf, 2011;Wilhelm & Sydow, 2018). For instance, Navis and Glynn (2010) found in the satellite radio market that firms first pursued cooperation to build the legitimacy of the new market, but then competed against each other to attract more customers once the market had been established. Further, tourism destinations may separate cooperation and competition according to seasonality, for example by cooperating during the winter (when customer demand is low) and competing during the summer (when customer demand is high). Research in this area has focused mostly on longer periods of cooperation and competition (e.g., de Rond & Bouchikhi, 2004; Navis & Glynn, 2010), with little consideration of temporal separation when the time frames for cooperation and competition are short.
Integration suggests that individuals pursue simultaneous cooperation and competition in the same space or location and at the same time (Das & Teng, 2000;Gernsheimer, Kanbach, & Gast, 2021;Lado et al., 1997). In other words, no means exist to separate the two opposing forces. Instead, individuals synthesize cooperation and competition (Chen, 2008;Fernandez, Le Roy, & Chiambaretto, 2018;Luo, Slotegraaf, & Pan, 2006). For example, hotel managers are careful to share information with platforms that is relevant to their business relationship (cooperation), while withholding any information that may reveal their intellectual property (competition).
In order to effectively integrate cooperation and competition, managers need to apply analytical and executional capabilities (Gnyawali et al., 2016). Analytical capability is the ability to think paradoxically, appraise paradoxical situations effectively, and understand the opportunities and threats these situations present Raza-Ullah et al., 2018). On the other hand, executional capability is the ability to identify and implement solutions to handle paradoxical situations and the resultant tensions (Gnyawali et al., 2016;Lundgren-Henriksson & Kock, 2016). It involves firms coming up with routines to integrate cooperation and competition (Lundgren-Henriksson & Tidström, 2021). There is evidence that both analytical and executional capability are needed for successful coopetition (Gnyawali et al., 2016;Rai et al., 2022).
While managerial actions are key to handling the tensions in coopetition, recent empirical work suggests the significance of emotions in tension management (Pradies, 2022;Raza-ullah, 2020). Paradoxical situations in coopetition often lead to positive and negative emotions arising concurrently . Examples of positive emotions include happiness, confidence, and enthusiasm, with negative emotions including sadness, resentment, and anger. These mixed feelings foster tensions as managers find it difficult to deal with them (Razaullah et al., 2020). To handle such tensions, managers need to possess emotional capability, a skill set that enables them to embrace mixed emotions, understand their implications, and internally regulate and control such feelings (Raza-Ullah et al., 2018). Possessing emotional capability has positive implications for firm performance when combined with other tension management skills (e.g., balancing capability) (Raza-ullah, 2020).
In addition, relational capability is the ability to develop mechanisms to manage the paradox of trust and distrust (Raza-ullah, 2021;Raza-ullah & Kostis, 2020). Trust relates to beliefs that a business partner is dependable, honest, and has the competence to get the job done Lascaux, 2020). Trust creates peace of mind due to the perception that a business partner has the partnership's best interests at heart and will work reliably (Raza-ullah, 2021). In Table 1 Tension management approaches and capabilitiesdefinitions and key references. Raza-ullah (2020); Raza-ullah et al. (2020) Relational capability Developing mechanisms to manage trust and distrust.
Czernek and Czakon (2016); Raza-ullah (2021); Razaullah and Kostis (2020) Balancing capability Embracing the opposite influence of cooperation and competition without letting either of the forces become the dominant focus. Bahar et al. (2022); Gnyawali and Charleton (2018); Razaullah (2020) contrast, distrust refers to beliefs that a partner firm has a hidden agenda, acts in a selfish manner, or lacks the skillset to deliver results (Raza-ullah & Kostis, 2020). It brings a wariness that safeguards firms from being over-optimistic about business relationships and prepares them from any undesirable behaviors or surprises (Lumineau, 2017). Therefore, effectively managing the trust-distrust paradox is meaningful in tension management. In fact, empirical work has demonstrated that firms who manage simultaneous trust and distrust in a balanced way are able to yield superior performance compared to firms that focus more on either trust or distrust (Raza-ullah, 2021). Finally, balancing capability is the ability to embrace the opposite influence of cooperation and competition without letting either take over or become the dominant focus (Gnyawali & Charleton, 2018;Hannah & Eisenhardt, 2018;Raza-ullah, 2020). Balancing capability incorporates the separation and integration of cooperation and competition to develop a range of alternative business activities to manage tensions (Bahar et al., 2022;Raza-ullah, 2020;Raza-Ullah et al., 2018). It also requires a work environment that supports employees to deal with opposing forces (Raza-ullah, 2020).
Overall, applying appropriate tension management approaches is vital to successfully handling concurrent cooperation and competition. These approaches offer insightful perspectives on how to manage tensions through separation and integration of cooperation and competition. In addition, capabilities (e.g., balancing and relational capabilities) are the types of expertise firms need to handle those tensions. However, studies have mostly focused on coopetition and tension management far from the customer, with few studies of coopetition close to the customer. Our research addresses this knowledge gap by focusing on hotel-platform relationships, which are always in close customer proximity. More specifically, we explore how this coopetition unfolds with different degrees of tension. We now describe our empirical work towards providing in-depth understanding on coopetition between hotels and platforms.

Research methods
We used a qualitative, grounded theory approach to discover how coopetition occurs between hotels and platforms (Corbin & Strauss, 2008;Strauss & Corbin, 1990). Our choice of grounded theory was based on three distinct rationales. First, there is limited knowledge of (i) coopetition in platform environments and (ii) coopetition in close customer proximity. Grounded theory allow researchers to draw on the experiences of knowledgeable participants to generate new explanations when existing scholarly knowledge is limited or lacking sufficient detail. Second, grounded theory is frequently used in marketing when examining complex and nuanced topics (cf., Chase & Murtha, 2019;Homburg, Jozić, & Kuehnl, 2017;Johnson & Sohi, 2016;Nenonen, Storbacka, & Windahl, 2019) such as coopetition. Third, grounded theory is a well-developed qualitative approach, providing researchers the tools needed for meticulous analysis as well as offering readers the guidelines for assessing the analytical procedures.

Sampling
Theoretical sampling underpinned our research method because our aim was to build theory (Mello & Flint, 2009;Strauss & Corbin, 1990). Theoretical sampling is an ongoing procedure for recruiting informants to "maximize opportunities to develop concepts in terms of their properties and dimensions, uncover variations, and identify relationships between concepts" (Corbin & Strauss, 2008, p. 143). We recruited informants based on their expertise on coopetition with platforms. In the beginning, we sampled employees who worked in the reservations departments of hotels. However, based on our ongoing findings, we then sampled informants in other roles, including customer service, hotel operations, revenue management, marketing, and sales. Thus, our sampling approach was a continuous process; based on the evolving findings from the initial informants, we recruited other types of informants to gain the full range of perspectives (Flint & Woodruff, 2001;Strauss & Corbin, 1998). This ongoing sampling procedure was fundamental to developing fresh theoretical insights. It also helped us develop both depth and breadth of knowledge on coopetition. In practice, the data collection and data analysis stages were intertwined until theoretical saturation was attained (Matteucci & Gnoth, 2017;Suddaby, 2006). In other words, there was no clear division between the data collection and analysis stages. We achieved theoretical saturation after interviewing 46 informants, at which point no new respondents were sampled. We relied on two separate indicators to make an informed decision on theoretical saturation. First, we determined that further sampling of informants would result in redundant information; and second, we had developed comprehensive explanations for each of the identified themes (Flint, Woodruff, & Gardial, 2002;Glaser & Strauss, 1967).
Overall, we recruited 41 respondents from 23 hotels for our study. These hotels worked with multiple platforms simultaneously. We also recruited five employees from two platform firms. These platform firms worked with our sampled hotels. Although we were investigating the hotels' perspective of coopetition, we sampled employees from platform firms to further enhance the credibility of our findings. Table 2 provides an overview of our respondents and the hotels/platform firms that participated in the study. As shown, we attained a high level of diversity at both the respondent level (organizational hierarchy, role, job experience) and firm level (e.g., size, ownership structure). From the perspective of our research question, such diversity helped us capture rich information on how hotels manage coopetition with platforms close to the customer interface. In Appendix A we provide a detailed list of our respondents.

Data collection
Consistent with the guidelines for conducting grounded theory research, we collected qualitative data in multiple waves (Matteucci & Gnoth, 2017). Data collection was conducted in New Zealand and ran for eight months from August 2018 to March 2019. Data collection primarily involved face-to-face interviews. The interviews began with broad questions. Based on responses to those questions, we then articulated specific questions to gather in-depth accounts of coopetition. In Appendix B, we share a sample of our interview questions. The length of the interviews ranged from 26 to 76 minutes, averaging 45 minutes. The interviews were held in meeting rooms, offices, or cafes. We used an audio recorder to record the interviews, and the audio files were transcribed for the purpose of coding. We also gathered secondary documents (e.g., reports on the type of analytics information hotels are able to access on platforms). These documents provided background information on how the hotels worked with platforms.

Table 3
Data structurecoopetition approaches to manage tensions.

Concepts Themes Dimensions
Receiving positive reviews.
Optimizing visibility Optimization Providing up-to-date and complete content information (e.g., pictures, service details). Replying to customer queries promptly. Participating in promotional programs. Developing rates based on rate information of strategic groups.
Optimizing rates Developing rates based on rate information of individual competitors. Disconnecting from platforms when there is a sporting event in the city.
Disconnecting during high demand Temporal disintermediation Disconnecting from platforms when there is a musical or cultural event in the city. Disconnecting from platforms when the forecasted occupancy levels are high.
Disconnecting during high occupancy Disconnecting from platforms when there are only a few rooms left to sell. Regulating inventory based on market demand.
Regulating inventory Integration Regulating inventory based on occupancy levels. Regulating inventory to maintain the right balance.
Developing key performance indicators (KPIs) to manage costs.
Controlling distribution costs Prioritizing direct sales over platform sales. Evaluating the returns of distributing via platforms. Targeting a portfolio of customer groups (e.g., business, leisure).
Broadening the target market Not relying too heavily on platforms or direct hotel channels. Accessing customer data (e.g., booking data).
Exchanging customer data Protecting customer data (e.g., customer likes and preferences). Understanding the reasoning behind positive and negative emotions.
Regulating positive and negative emotions Thinking about constructive action plans based on mixed emotions. Demonstrating confidence that platforms have the competency to market hotels to a wide audience across the world.
Regulating confidence and caution Demonstrating confidence that the content of conversations with platforms will be kept confidential. Demonstrating caution by discovering and evaluating the self-interest of platforms. Adhering to rate parity in the platform space.
Executing rate dis(parity) Spatial separation Not adhering to rate parity in the hotel channel space. Cooperating in the platform space to attract new customers by connecting to a large customer base.
Acquiring new customers Competing in the hotel channel space to attract new customers (e.g., by engaging customers with loyalty programs and providing more attractive offers). Communicating with customers pre-stay in the platform space.
Communicating with customers Communicating with customers during the stay phase in the hotel space. Replying to online customer reviews post-stay in the platform space. Communicating with customers post-stay in the hotel channel space. Cooperating in the platform space to provide customer freedom to book on the platform of their choice.
Influencing customer service Cooperating in the platform space to alter customer bookings. Competing in the hotel space by providing personalized customer service.

Data analysis
To examine the interview transcripts and secondary documents, we applied open, axial, and selective coding procedures recursively (Strauss & Corbin, 1997, 1998. Open coding generated the initial concepts, axial coding assembled these concepts into themes based on their common properties, and selective coding synthesized the themes into higherorder dimensions. We applied the constant comparison method throughout these three coding rounds (Charmaz & Thornberg, 2021;Urquhart, Lehmann, & Myers, 2010). In other words, we compared the data from different participants, secondary documents, and the coopetition literature to develop common patterns. We used NVivo computer software to efficiently carry out the data analysis.
To ensure rigor and transparency in our data analysis procedure, we utilized the Gioia methodology (Gioia, Corley, & Hamilton, 2012) to create a data structure (see Table 3). The resulting data structure demonstrates how we first came up with concepts from the informants' reflections, and subsequently assembled them into themes and dimensions (Gehman et al., 2018).

Trustworthiness of the research
To ensure the trustworthiness of our inquiry, we satisfied four core criteria: credibility, generalizability, dependability, and conformability (Denk, Kaufmann, & Carter, 2012;Flint et al., 2002;Lincoln & Guba, 1985;McGinley, Wei, Zhang, & Zheng, 2021;Miles & Huberman, 1994;Strauss & Corbin, 1990). We assessed the credibility of findings via respondent check-in. In other words, the respondents were asked to offer their comments on our findings and conceptual framework. Further, three researchers were involved in the research process. We established generalizability through the use of theoretical sampling. The insights from this study are applicable for hotels pursuing coopetition with multiple platforms. To ensure dependability, we encouraged the respondents to share their thoughts and experiences covering both the past and present. We regard our findings as dependable because of similarities across the respondents' reflections. The findings are also dependable because of the rigorous application of the constant comparison method to derive consistent patterns from the rich data. Finally, we achieved conformability by gathering detailed descriptions from many respondents (n = 46) and sharing a summary of the findings with those respondents for feedback. These procedures ensured that the results reflect the respondents' perspectives and not the researchers' biases.

Findings
We identified 36 concepts that coalesce into 14 themes, which integrate to form four dimensions of coopetition approaches (see Table 3). These dimensions are optimization, temporal disintermediation, integration and spatial separation. We offer definitions and quotes for the themes and dimensions identified for coopetition approaches in Appendix C.
These respective coopetition approaches are relevant depending on the degree of tension between hotels and platforms. To understand the degree of tension, we drew on paradox thinking (Smith & Lewis, 2011) and actor roles (Solomon et al., 1985), finding that tension levels depend on (i) the degree of cooperation and competition (see Fig. 1) and (ii) the roles that hotels and platforms plays in different situations (see Table 4). In summary, when tensions are low, hotels apply either optimization or temporal disintermediation. However, under high tensions, hotels pursue integration, while under moderate tensions they enact spatial separation.
We next explain each coopetition approach in a more in-depth manner as follows. First, we explain when the coopetition approach is pursued (i.e., at which tension level). Second, based on paradox thinking and actor roles, we delve deeper into why the tension level is at a given level in a specific situation. Third, we explain how the tension is managed (i.e., the coopetition approaches). Finally, we discuss what capabilities are required to handle those tensions (see Table 5).

Optimization
Optimization in this context is the ability of hotels to boost their ranking in search queries and establish an ideal room rate to increase sales in platform environments. Optimization is pursued under low levels of tension. The indicators for low tension are (i) high cooperation and low competition, and (ii) hotels and platforms play complementary roles. In practice, hotels rely heavily on platforms and seek their support to attract more customers than other hotels in the market (high cooperation). As a result, the competitive aspects of the hotel-platform relationship are less salient during optimization activities (low competition). This is because competition is detrimental for optimization as it reduces hotel sales (e.g., through lower hotel visibility on Booking.com). This results in low tension because the pull for cooperation is greater than for competition. In this situation, hotels do not face the challenge of handling contradictory forces that have high intensity.
In terms of actor roles, hotels play the role of supplier by, for example, providing satisfactory accommodation services to receive positive reviews and boost their ranking in search results on Expedia. com. Platforms, on the other hand, work as distributors (e.g., by managing hotel ranking in search results) and data partners (by providing room rate intelligence reports to hotels). These complementary roles focus heavily on cooperation and less on competition. As a result, hotels find it comparatively less difficult to perform their roles. To be precise, they do not face the challenge of performing both complementary and incompatible roles, thus leading to low tensions. Two themes relate to optimization: optimizing visibility and optimizing rates.
Regarding optimizing visibility, hotels cooperate with platforms to attract customer attention in platform contexts. Hotels work hard to receive positive reviews, ensure that their content (e.g., service details, pictures) is current and complete, reply to customer questions in a timely manner, and join promotional programs. These actions enhance hotels' ranking on platforms and improve the likelihood of attracting customers. Platform B offered insights into the crucial role of reviews in hotels being prioritized in search results: "To ensure our customer experience, hotels with review scores under 4 out of 5 cannot be chosen to be our [Platform B] first priority to recommend V.S. Bahar et al. to our customers. But we will help and guide potential properties to increase their service quality based on the reviews." [ID: 46] With regard to optimizing rates, hotels collaborate with platforms to learn about the rates of competing hotels. Platforms provide detailed rate information for hotels to develop an ideal rate, as explained by Hotel K: "Every morning, we get an email where they [platforms] tell us that for the next week-these are our rates, but our competitors' average rate is this. They say my rates are slightly lower than the average rates of the competitor hotels. So, I may want to increase my rate a little bit, or if my rates are too high, then I want to decrease it a little bit." [ID: 22] To pursue optimization, hotels need to possess executional capabilities (Gnyawali et al., 2016), and develop streamlined operational routines that complement and match their business activities with the resources available on platforms. For example, some reservations departments have established routines of replying to customer queries in the chat function of a platform (e.g., Booking.com) within 12 hours.

Temporal disintermediation
Temporal disintermediation refers to hotels disconnecting from platforms and deciding not to sell room inventory via them for a limited time. Temporal disintermediation occurs under low tension because (i) competition is high and cooperation is low, and (ii) hotels and platforms play incompatible roles. In practice, hotels focus temporarily on generating direct channel sales only (high competition), avoiding the high commissions involved in selling their inventory via platforms (low cooperation). This usually occurs when market demand and hotel occupancy levels are high (e.g., when a city is host to a sporting or cultural event). This results in low tension because the pull for competition is greater than for cooperation. Hotels are thus not faced with the challenge of handling simultaneous cooperation and competition occurring at high intensity.
In this case, hotels and platforms both perform distributor roles. These incompatible roles focus more on competition and less on cooperation. As a result, hotels feel comfortable in enacting the distributor role as it leans towards one dimension of coopetition (i.e., competition). This causes low tension because hotels do not face the difficulty of playing both complementary and incompatible roles.
Despite this temporary stop on sales via platforms at an operational level (competition), hotels continue to have a business relationship with platforms at a strategic level (cooperation). The operational and strategic aspects of temporal disintermediation are not separated between departments or spaces. Rather, the same department (revenue management) conducts this activity. Further, temporal disintermediation is different from temporal separation. In the former, both cooperation and competition exist at the same time, albeit with competition being the dominant force. In the latter case, however, the two forces do not exist at the same time but rather are separated over time (Lavie et al., 2011;Wilhelm & Sydow, 2018). Two themes relate to temporal disintermediation: disconnecting during high demand and disconnecting during high occupancy. Hotel W reflected on this activity: "On a particular day, week, or month, if we know that all our bookings are going to come through our direct channels, we will close them [platforms]. So, you've also got to have that relationship to be able to do that as well." [ID: 41] To pursue temporal disintermediation, hotels require executional capabilities (Gnyawali et al., 2016) that allow them to create flexible routines to opt-in and opt-out of working with platforms in a timely manner, as the situation demands (e.g., during peak seasons).

Integration
In our framework, integration relates to synthesizing cooperation and competition at the same place and time to attain coopetition balance. Integration is pursued at high tension levels. The indicators for high tension are (i) high cooperation and high competition and (ii) complementary and incompatible roles performed in the same place and time.
In practice, hotels rely heavily on platforms as their business partners (high cooperation) but also compete with them (high competition) to enhance economic profits. As a result, two contrasting forces with high intensity levels create high tension, as managers are challenged to deal with the opposing influences of cooperation and competition. In terms of roles, hotels feel overwhelmed by the demands of being a supplier (cooperation) and a distributor (competition) simultaneously in the same place and time, resulting in high tension.
Integration approaches includes regulating inventory, controlling distribution costs, broadening the target market, exchanging customer information, regulating positive and negative emotions and regulating confidence and caution.
With regard to regulating inventory, hotels change the volume of rooms sold on platforms; that is, they either do not offer all their Table 4 Coopetition approaches based on actor roles.   Bahar et al. inventory (cooperation) or provide zero inventory (competition). Instead, they perform a balancing act and identify appropriate levels of inventory to be sold based on market demand data and occupancy levels. This activity requires managers to maintain simultaneous cooperation and competition in their decision making. In terms of controlling distribution costs, hotels pursue concurrent cooperation and competition to optimize the cost of working with platforms. In practice, hotels identify key performance indicators (KPIs) to control expenses when dealing with platforms. These KPIs ensure that hotels do not allocate a budget that is too high to sell via platforms (cooperation), or a budget that is too low (competition), which may result in low occupancy levels. Further, hotels may prioritize direct sales (competition) over platform sales (cooperation) and assess the returns from distributing via platforms. Such activities ensure that costs are kept in check to enhance profits. Hotel H described how they handled costs:

Role of platforms Complementary and/or incompatible roles Level of tension
"Every room that you can sell through your own direct channel, you shouldn't sell it on Platform C. So, they [platforms] should only allow you to bring additional business, not anything of your own business and that's how you will benefit from them. When you start letting them [platforms] take over your own booking engine, that's when you lose money." [ID: 15] Towards broadening the target market, hotels attract a wide range of customer segments, such as leisure and business customers. Leisure customers predominantly book on platforms (e.g., Hotels.com), whereas business customers mainly book via direct channels (e.g., through a hotel sales team). Therefore, broadening the target market scope protects hotels from having to depend solely on either platform sales (cooperation with platforms) or direct channel sales (competition with platforms). Instead, hotels pursue simultaneous cooperation and competition by targeting multiple customer groups to generate revenue.
In terms of exchanging customer data, hotels access and protect customer data from platforms. To be precise, hotels receive customer booking data (cooperation) from platforms, but do not share customer likes and preferences (competition). To support this process of simultaneous cooperation and competition, hotels use IT systems to manage customer data from one central location.
However, hotel managers experience the paradox of positive (e.g., love) and negative (e.g., hate) emotions when dealing with platforms, which causes high tension (Raza-ullah, 2020;. As a result, managers find these extreme contradictory emotions very challenging. To deal with such emotional tensions, hotel managers regulate positive and negative emotions by learning to understand the underlying causes of these mixed emotions and thinking about constructive action plans based on these sentiments. The Director of Revenue at Hotel W had this to say on how to regulate positive and negative emotions: In addition, hotel managers experience the paradox of trust (comprising competency and confidentiality) and distrust (hidden agenda and self-interest) beliefs towards platforms, which engenders high tension (Raza-ullah, 2021). To handle these extremely contradictory beliefs, managers regulate confidence and caution in their business relationships with platforms. In terms of regulating confidence, managers exhibit a strong belief that platforms have the competency to market hotels to a wide audience across the world. In addition, they show confidence in the people working for platforms in terms of keeping the content of their conversations confidential and not sharing information with other hotels. This encourages hotels to work with platforms, attract international guests and share confidential information. However, managers simultaneously exert caution by seeking to discover and evaluate their counterpart's self-interest and ulterior motives. Demonstrating caution safeguards hotels from undesirable platform behaviors. For instance, the Reservations Manager at Hotel G discovered how platforms encourage hotels to engage in campaigns, concluding that it helps platforms make more money at the expense of hotel profitability: Overall, we find that hotels possess a range of capabilities for pursuing the integration of cooperation and competition, including analytical capability, executional capability, emotional capability, relational capability and balancing capability (Gnyawali et al., 2016;Raza-ullah, 2020Raza-ullah, , 2021. For instance, when it comes to broadening their target market, managers require analytical capability to think paradoxically and evaluate the opportunities and threats of targeting (i) either one customer group (e.g., leisure or business) or (ii) multiple customer segments. Targeting one customer group will tip the business relationship with platforms towards either cooperation or competition. Conversely, targeting multiple customer segments will ensure both cooperation and competition are at play in a balanced way. Subsequently, executional capability is used to implement the latter strategy and come up with appropriate management processes or routines.
In addition, managers possess the emotional capability to effectively regulate positive and negative sentiments towards platforms and embrace these emotions constructively. Moreover, relational capability offers managers the expertise to simultaneously synthesize trust and distrust beliefs. This helps managers develop mechanisms to concurrently demonstrate confidence and caution when dealing with platforms.
Finally, hotels require balancing capability to integrate cooperation and competition without allowing either of the forces to become dominant. For instance, to control distribution costs, hotel allocate optimal budgets to work with platforms. Instead of assigning high (cooperation) or low (competition) budgets, hotels perform a balancing act, devoting the right amount of money to generate optimal sales and profits.

Spatial separation
Hotels use three distinct spaces to separate cooperation and competition and achieve balance between the two forces: hotel direct channels (e.g., website, mobile app), the physical hotel, and the platform interface. Hotels use spatial separation under moderate tension levels. The indicators for moderate tension are: (i) optimal degree of cooperation and competition (not too high, not too low) and (ii) complementary and incompatible roles pursued in different places. In practice, hotels face the pull of cooperation and competition in different areas or places, keeping tension levels manageable and optimal. In other words, they are not overburdened by having to handle the effects of cooperation and competition in the same place.
In terms of roles, hotels for instance, are expected to play the role of supplier in the platform space (cooperation) and that of distributor in the direct channel space (competition). Hence, although hotels experience stress in enacting these two roles, they perform these roles in separate, dedicated places, making the tension levels optimal rather than high. We identify four mechanisms used by hotels to achieve spatial separation: executing rate (dis)parity, acquiring new customers, communicating with customers, and influencing customer service.
For executing rate dis(parity), hotels cooperate in one location (platform interfaces) where they follow rate parity, but compete in another location (direct channels) by not observing rate parity. As a result, hotels provide similar rates on all platforms but decide to offer a lower rate via direct channels. Hotel G shared their view on this: "You need to have the strategy wherein you are priced lower than them [platforms] just because you want more people to book through your own website. And I had heard this from one of the account managers about a study that 37% of people will go to the hotel's website and see if they can find a cheaper room." [ID: 12] In terms of acquiring new customers, hotels separate cooperation and competition between two spaces: the platform interface and direct channels. Cooperation occurs on the platform interface as hotels connect with a vast number of customers across the globe, which for most hotels is impossible without the support of platforms. Conversely, competition occurs via direct channels (e.g., hotel websites), as hotels involve customers in loyalty programs and provide them with additional incentives to book direct (e.g., free internet or complimentary breakfast).
For communicating with customers, hotels use different spaces during the customer journey. They use the platform space at the pre-stay stage to answer customer questions or send pre-arrival messages, therefore collaborating with platforms. However, during the stay phase when staff members in the hotel space interact with customers face to face, they can promote the value of booking direct, hence competing with platforms. Finally, post-stay, hotels use the platform space to reply to customer reviews (cooperation) but use direct channel spaces such as telephone and email to entice customers to make a reservation directly with them for their next trip (competition). Hotel T described how in the stay phase, they convinced customers to book direct next time: "So basically, when the guest checks in, if they have booked through them [platforms], we ask them, 'why are you here? Have you booked yourself, or did someone else book for you? Do you come often? Are you coming next time? Here is a card-call us next time, book direct, we will give you a discount. ' We get them to start booking with us directly." [ID: 36] Regarding influencing customer service, hotels cooperate with platforms in the platform space but compete in the hotel space to shape the customer experience. Specifically, in the platform space, hotels work with multiple platformsa practice commonly known as multihomingso that customers have the freedom to make a reservation on the platform of their choice. Hotels also collaborate with platforms to support customers when they want to make changes to their reservations. However, hotels compete with platforms in the hotel space by offering personalized service, which pleases customers and motivates them to book directly next time.
Overall, to pursue spatial separation, hotels possess executional capability and balancing capability (Gnyawali et al., 2016;Gnyawali & Charleton, 2018). For instance, hotels come up with established routines to follow rate parity in the platform space (cooperation) but not on the hotel channel space (competition). As a result, hotels strike a balance of cooperation and competition, rather than focusing on one of the forces excessively.

Discussion
Research on coopetition has focused mainly on business activities that occur far from the customer. As a result, empirical insight into coopetition between firms close to the customer is limited. Our study therefore investigated how hotels and platforms pursue coopetition in close customer proximity. Using grounded theory, we developed novel insights illustrating differences in coopetition approaches depending on the degree of tension in business relationships. We identified four distinct coopetition approaches: optimization, temporal disintermediation, integration, and spatial separation. Optimization is pursued under low tension when the influence of cooperation is greater than for competition (cooperation-dominant situation) and firms play complementary roles. Temporal disintermediation is also applied under low tension, but in this case the influence of competition is greater than for cooperation (i.e., competition-dominant situation) and firms have incompatible roles to perform. However, integration is pursued under high tension levels because the degree of cooperation and competition is high (i.e., balanced-strong coopetitive situations) and firms are required to play complementary and incompatible roles simultaneously, in the same place and time. Finally, spatial separation is conducted under moderate levels of tension as the degree of cooperation and competition is neither too high, nor too low (i.e., balanced-moderate coopetitive situations) and firms perform complementary and incompatible roles, but in different places. Based on our findings, we now describe the theoretical and practical implications of our work, explain the research limitations, and suggest avenues for future empirical work.

Theoretical implications
Our research contributes to the coopetition literature in four ways. First, although the literature offers insight into how paradoxical situations relate to tensions in business relationships (Bengtsson & Raza-Ullah, 2017;Gnyawali et al., 2016;Raza-ullah, 2020;Raza-Ullah et al., 2014), there is limited understanding of how the roles played by different actors influence tension levels. Drawing on role theory (Biddle, 1986;Homburg et al., 2011;Solomon et al., 1985), we complement and contribute to this stream of work by suggesting the following connection between actor roles and the degree of tension. When hotels play complementary (e.g., supplier) or incompatible (e.g., distributor) roles to that of platforms, tension levels are low as the expectations of those roles lean towards cooperation and competition, respectively. In this case, hotels do not feel burdened to enact both complementary and incompatible roles simultaneously.
In contrast, when hotels perform complementary (e.g., supplier) and incompatible (e.g., distributor) roles in the same place and time, tension levels become high. This is because it becomes extremely challenging for hotels to handle two roles simultaneously: one focusing on cooperation (supplier), and the other on competition (distributor).
However, when hotels have to perform complementary (supplier) and incompatible (distributor) roles, but in different places, then the degree of tension is moderate (neither too high, nor too low). In this case, although hotels must play two roles, they enact those roles in different areas (e.g., supplier role on Expedia.com and distributor role on hotel website). Therefore, due to the availability of dedicated spaces to perform distinct roles, tension levels remain more manageable and tolerable.
Second, although the literature indicates that coopetition is competition focused close to the customer (Basterretxea et al., 2019;Bengtsson & Kock, 2000;Hoffmann et al., 2018;Tidström, 2009;Walley, 2007), our findings show that this is not always the case. We find that coopetition ranges between cooperation-and competition-dominant approaches depending on the degree of tension.
Under low levels of tension (high cooperation, low competition), hotels collaborate with platforms and reduce competition to optimize their visibility in platform contexts and develop ideal rate (price) plans. In practice, hotels come up with routines (Gnyawali et al., 2016) to align their activities with the platform environment and attain strategic fit. For example, hotels maintain high-quality content (e.g., high-resolution pictures) and high review scores in order to be prioritized in search results.
Further, when tensions are low because of low cooperation but high competition, hotels decide to disintermediate from platforms temporarily, although they continue to have an ongoing business relationship with them. In practice, hotels develop opt-in/opt-out routines that enable them to be agile and temporarily halt sales on platforms when market demand and occupancy levels are high. We find temporal disintermediation to be different from temporal separation. In the case of temporal disintermediation, competition is the dominant force (at the operational level), but cooperation exists (at the strategic level) at the same time. As a result, cooperation and competition occur simultaneously. However, with temporal separation, cooperation and competition do not take place simultaneously, but are separated over time (Damayanti et al., 2017;Smith & Lewis, 2011).
In contrast, when tensions are high (high cooperation and high competition), hotels manage coopetition by synthesizing the forces of cooperation and competition at the same place and time (Das & Teng, 2000;Gernsheimer et al., 2021;Lado et al., 1997). In practice, managers develop a coopetitive mindset, appraising paradoxical situations and making decisions that are both cooperative and competitive Raza-Ullah et al., 2018). For example, a revenue manager at a large hotel (that has 500 rooms) will analyze market demand data and occupancy levels to evaluate whether to sell 350, 319, or 300 room inventories on platforms. Subsequently, the manager implements procedures and routines (Gnyawali et al., 2016) for allocating appropriate inventory levels using IT technologies (e.g., distribution management tools). Hence, hotels maintain a balance of cooperation and competition (Gnyawali & Charleton, 2018;Hannah & Eisenhardt, 2018) by managing inventory in a way that maximizes profits. Integration of cooperation and competition also requires managers to be comfortable with experiencing positive and negative sentiments , and able to come up with constructive action plans based on these mixed emotions. It also challenges managers build business relationships on the basis of trust and distrust by demonstrating confidence and caution (Raza-ullah, 2021;Raza-ullah & Kostis, 2020).
Moreover, under moderate tensions (optimal cooperation and competition; not too high, not too low), hotels separate the two forces into different spaces (platforms, physical hotel, direct channels). Hotels develop mastery of the spatial environment by designing appropriate routines and actions for each space so that the benefit of working in all spaces is greater than working in one space alone. For example, hotels that use both platforms (cooperation) and direct channels (competition) to acquire new customers perform better than those that focus on either platforms or direct channels. This enables hotels to balance cooperation and competition, without allowing either of the forces to take precedence (Bahar et al., 2022;Raza-ullah, 2020).
Third, the tension management literature identifies locations at which cooperation and competition can be separated (Joseph et al., 2020;Poole & de Ven, 1989). These locations are related to different value chain functions, product areas, and geographical regions (Fernandez et al., 2014;Fernandez & Chiambaretto, 2016;Kylänen & Rusko, 2011;Le Roy & Czakon, 2016;Oinonen et al., 2018;Ritala & Tidström, 2014). We identify additional spaces to isolate cooperation and competition: platforms, physical hotel, and direct channels. These spaces are relevant when coopetition is close to the customer but not applicable for coopetition far from the customer. The identified spaces relate to channels for serving customers (Ailawadi, 2021;Cui et al., 2021) and we suggest firms can use different channels to pursue either cooperation or competition. Direct channels (e.g., hotel website, telephone) can be used for competition, whereas indirect channels (e.g., Ex pedia.com, Booking.com) are suited to cooperation.
Further, the literature suggests the importance of temporal separation for dividing cooperation and competition (Damayanti et al., 2017;Lavie et al., 2011;Smith & Lewis, 2011;Wilhelm & Sydow, 2018). However, we argue that temporal separation is possible mostly when the episodes of cooperation and competition are of a longer duration. For example, such separation is relevant when firms first cooperate to develop a new market and later compete to attract more customers (Navis & Glynn, 2010). However, in the context of this research, brief episodes of cooperation and competition last for only weeks, days, or even hours. Under such circumstances, temporal separation is not sufficient to manage tensions during short episodes of cooperation and competition in the same location. Instead, hotels must separate the two forces into different spaces: platform, physical hotel, and direct channel. For example, hotels communicate with customers about their policies three days before arrival using the chat function on platforms (cooperation), motivate customers to book direct when they are staying at the hotel (competition), reply to reviews on platforms within a few hours after the guest has left a review (cooperation), and send emails directly to guests one month after check out to offer incentives to book on the hotel website (competition). Hence, temporal separation of cooperation and competition is impossible when hotels work with platforms. Instead, spatial separation plays the dominant role in dividing the two forces.
Finally, the literature provides information on the organizational hierarchy levels better suited to handling separation or integration of cooperation and competition Das & Kumar, 2010). However, there is limited insight regarding whether the ability to utilize specific coopetition approaches depends on firm size. Our analysis shows that hotels of all sizes can optimize visibility and price in platform environments. This is because platform firms inform all hotels about ways to improve their visibility. All hotels also have access to rate information on platforms as a basis for better decision making on rates. Moreover, all hotels, irrespective of their size, are capable of separating cooperation and competition in different locations because the focus is only cooperation in the platform space, and only competition in the physical hotel and direct channel space.
However, larger hotels are more capable of integrating cooperation and competition because they have revenue management professionals to synthesize cooperation and competition effectively. For instance, larger hotels have dedicated managers (e.g., Revenue Manager, Revenue and Yield Manager, Director of Revenue) who have the right experience and skill sets to think paradoxically, regulate positive and negative emotions and maintain ongoing business relationships on the basis of trust and distrust. These managerial positions, which are fundamental to effectively integrating cooperation and competition, are not usually present in smaller hotels. Additionally, larger hotels typically decide to stop working with platforms temporarily when market demand is high because they have a vast and loyal customer base and the capacity to heavily promote through direct channels to sustain their business. Such temporary actions are less salient for smaller hotels, as they rely heavily on platforms to attract guests even during periods of high demand.

Practical implications
We offer valuable practical insights for hotels and platform firms. First, we encourage hotels to implement a range of coopetition approaches to manage their relationships with platforms depending on the contextual situation. Specifically, (i) when the pull for cooperation is greater than for competition and hotels are in the supplier role, they must attain a strategic fit to achieve a high ranking on platforms and optimize rates. However, (ii) when the pull for competition is greater than for cooperation and hotels are in the distributor role, they can create an opt-in/opt-out tactic to remain flexible and temporarily stop selling on platforms when they deem it necessary (e.g., when there is a large influx of international guests) and focus on direct channel sales only.
Further, (iii) when the pull of cooperation and competition is high and hotels are responsible for simultaneously enacting the role of supplier and distributor, hotel managers need to think analytically, evaluate external (e.g., market demand) and internal (e.g., finances, occupancy level) data and execute decisions that balance cooperation and competition. Under these circumstances, managers must possess the ability to handle mixed emotions (positive and negative) to come up with constructive action plans. They also need to demonstrate both confidence and caution to develop ongoing business relationships with platforms based on trust and distrust.
In addition, (iv) when the influence of cooperation and competition is moderate (neither too high nor too low) and hotels are required to pursue the role of supplier and distributor in separate places, they must become expert in managing distinct spaces to leverage the benefits of cooperation (in the platform space) and competition (in the direct channel and physical hotel space) in a balanced way. Second, we observed that some hotels, especially smaller ones, are less capable of synthesizing cooperation and competition to enhance profits. We recommend that these hotels (i) set up a revenue management team and recruit managers with the right skill sets (e.g., revenue managers) to handle the contradictory influence of cooperation and competition and/or (ii) focus on upskilling their existing employees through training and development to optimally manage coopetition.
Finally, we encourage platform firms to develop plans to reduce the number of hotels that disintermediate from them on a temporary basis. The primary reason for temporal disintermediation is that hotels can attain high occupancy levels without selling via platforms when market demand is high. At these times, we encourage platform firms to communicate with hotels about billboard effects, which occur when customers first gather information about prospective hotels on platforms and later go to the hotel website to make a booking (Anderson, 2009). Therefore, if hotels sell a minimum level of inventory on platforms during high-demand periods, they will benefit from billboard effects, while platform firms will also continue to generate sales and so avoid temporal disintermediation.

Limitations and future research avenues
Our study has some limitations that have wider implications for future research on coopetition. Our findings are specific to the relationship between hotels and platforms (e.g., Hotels.com, Trip.com). Therefore, future research could explore coopetition in close proximity to customers between hotels and partners that are not platforms. In particular, studies could focus on whether any differences in coopetition exist when hotels work with partners that are platforms (e.g., Expedia. com) and those that are not (e.g., traditional travel agents, wholesalers).
Further, our research offers insights into how the ability to pursue coopetition differs depending on hotel size. We also point out how coopetition approaches change based on tension levels. Future research could thus take the field forward by exploring other contextual factors influencing coopetition. For instance, we encourage researchers to conduct longitudinal studies and investigate whether and how the spread of the novel coronavirus is affecting tension levels and changing coopetition patterns between hotels and platforms over time.
Finally, this research focuses on hospitality platforms such as Boo king.com and Expedia.com. However, new and emerging blockchain enabled platforms have different characteristics to hospitality platforms. Blockchain is a ledger system that facilitates a transparent record of irreversible transactions (Iansiti & Lakhani, 2017). Blockchain removes the need for intermediaries and thus decentralizes interactions between different stakeholders (Weking et al., 2020). Future researchers could therefore explore coopetition in blockchain enabled platform contexts.

Declaration of Competing Interest
None.

Appendix B. Sample interview questions
Broad questions: • "Can you explain how Hotel X works with various booking platforms? What are the key activities? What resources are required?" • "What are the main benefits and challenges of working with platforms?" Specific questions: • "You mentioned the importance of hotel ranking on these platforms. How does Hotel X maintain a good ranking?" • "You mentioned that you have access to business analytics data. How and in what ways do you use such data?" • "You stated the importance of maintaining a balance in inventory levels. How do you decide on the right inventory levels?" • "You touched upon the importance of getting direct bookings. What does Hotel X do to receive direct bookings?" • "What are some of the other ways hotels compete with these booking platforms?" • "You mentioned the rate parity clause. How does the rate parity clause affect your decision making?" • "You stated that sometimes you stop selling on these platforms. When and why do you make this decision?" Final question: • "Are there any other ways Hotel X deals with various booking platforms that we have not touched on?"

Appendix C. Definitions and quotes Dimension Themes Quotes
Optimization is the ability to enhance hotel visibility in platform environments and develop ideal room rates.
Optimizing visibility is the ability of hotels to attract customer attention on platforms by prioritizing their position in search results.
"The most important thing is when your customers are happy, your review scores are getting better and better. The platform algorithms work in a way that they push your rankings higher." [ID: 12] Optimizing rates is the ability to come up with ideal rates (price) on platforms based on rate information reports provided by platforms. Temporal disintermediation is the ability to develop an opt-in/opt-out routine to remain agile and stop sales on platforms for a short time.
Disconnecting during high demand occurs when hotels stop selling their inventory on platforms when market demand is high (e.g., when there is a sporting event in the city). Integration is the ability to synthesize cooperation and competition at the same time and place.
Regulating inventory relates to changing the volume of rooms made available on platforms to enhance profits.

Dimension Themes Quotes
Broadening the target market is the ability to focus on multiple customer segments to generate sales.
"Part of our marketing strategy is to contact the corporates and schools. We let them know two years in advance what our rates are-just to maintain the relationship. So we do have long-term customers, and that is based on their contracts with us. We also target tourists and travelers." [ID: 13] Exchanging customer data relates to accessing and protecting customer data from platforms.
"The [customer] booking details from them automatically drop into our system." [ID: 28] "We maintain a profile of our customers, which is secured in our property management system." [ID: 25] Regulating positive and negative emotions relate to understanding the reasoning behind mixed (positive and negative) emotions, and thinking about constructive action plans based on these emotions. Spatial separation is the ability to develop mastery in managing different spaces (platforms, direct channels, and physical hotels) and divide cooperation and competition in separate areas.
Executing rate (dis)parity refers to following rate parity in the platform space but not the direct channel space.
"Usually we push one rate out to all platforms-this is called rate parity." [ID: 36] "Our channel is always 10, 5 dollars less, and we advertise that." [ID: 6] Acquiring new customers is the ability to cooperate with platforms in the platform space but compete with them in the direct channel space to entice new customers.
"They [platforms] are a bridge connecting you to the larger audiences that you may not have access to." [ID: 34] "Sometimes we will provide a better deal on our website. We will say, 'we will give you a discounted rate plus breakfast for one dollar.' So that's the benefit in itself. If you book direct, we'll give you breakfast at one dollar rather than paying 37 dollars." [ID: 26] Communicating with customers is the ability to cooperate with platforms in the platform space but compete with them in the hotel and direct channel space through customer communication. Influencing customer service refers to cooperating with platforms in the platform space but competing with them in the hotel space to shape the customer experience.