Resource integration, value creation and value destruction in collective consumption contexts

How consumers coordinate to create value in collective consumption contexts is of wide interest. While resource integration provides a promising theoretical lens, research has not yet examined collective resource integration. This article explores resource integration around subscription television use, through systemic interviews within households. A resulting framework relates five resource types to six resource integration activities: resource assembly, resource mastery, resource optimization, usage event planning, real-time usage design, and resource reflection. These interwoven activities enrich the prior view of resource integration as combining and applying resources. Furthermore, whereas literature assumes resource integration to be an individual process, each activity is observed to be undertaken collectively as well. Whether resource integration results in value creation or destruction depends upon consumers' agency over shared resources, their individual and collective mastery of those resources, resources' integrable quality, and the quality of collaboration between resource-integrating consumers. A new definition of resource integration is derived.


Introduction
Optimizing customer perceptions of value remains a strategic priority for firms seeking a competitive advantage. However, literature is in flux regarding the processes by which value arises (Grönroos, 2011). The traditional (albeit often implicit) view of value as delivered by firms through the act of product exchange or service delivery (Zeithaml, 1988) has been questioned by an evolving dialogue that emphasizes the customer's role in value creation (Heinonen & Strandvik, 2015;Vargo & Lusch, 2011). A number of recent empirical studies have explored the customer's role in business-to-business (e.g. Macdonald, Kleinaltenkamp, & Wilson, 2016) and business-to-consumer (e.g. McColl-Kennedy, Vargo, Dagger, Sweeney, & van Kasteren, 2012) contexts. A priority for scholars and practitioners alike, though, is to further understand how multiple actors combine to create value. A joint practitioner-academic consultation to identify service research priorities observed that "The most highly rated subtopics [for research] reveal a consistent theme …for tackling the interdependencies among actors in value co-creation" (Ostrom, Parasuraman, Bowen, Patricio, & Voss, 2015, p.147). A prevalent example is collective consumption contexts, where other consumers are present in the consumption environment (Albrecht, Walsh, & Beatty, 2017). To date, however, value creation in collective consumption contexts remains under-explored (Kelleher, Wilson, Macdonald, & Peppard, 2019).
Collective consumption contexts vary in a number of respects, including whether the presence of others is physical or virtual and whether it is synchronous or asynchronous (Kelleher et al., 2019). The emerging focus on these contexts (Carù & Cova, 2015) is due, in part, to their increasing profusion through online technologies, stimulating the formation of global brand communities (Schau, Muñiz Jr., & Arnould, 2009), national consumer networks (such as the UK's mumsnet.com) and local interest groups (for example, yelp.com). In many collective consumption settings, consumers are required to share limited resources. Sports club members, for example, must coordinate their use of sports pitches and equipment, expert coaches, and changing and leisure areas, while geocaching enthusiasts participating in GPS-enabled treasure hunts share globally distributed objects known as caches, which are accessed asynchronously (Boulaire & Cova, 2013). We apply a resource integration perspective to our study of such collective consumption contexts. Resource integration provides a promising lens for exploring how consumers use focal service provider offerings in conjunction with a diversity of other resources in such contexts, and how this creates or destroys value. Resource integration refers to the process by which actors, such as consumers, combine and apply resources in pursuit of value creation (Peters et al., 2014;Vargo & Lusch, 2004); we will refine this definition later.
The recent resource integration narrative is conceptually rich but predominantly theoretical. Of the sparse empirical works, business-tobusiness research has evidenced the collective nature of resource integration within that context, highlighting collaborative activities among managers in customer and supplier firms (Macdonald et al., 2016). By contrast, studies of consumers (e.g. Baron & Warnaby, 2011) have maintained a focus on individuals as resource integrators. This has enriched understanding of the resources individuals integrate -including resources provided by other individuals -but whether and how consumers undertake collective resource integration remains unclear. This leads to our research question: how do consumers in a collective consumption context integrate resources to create value ? We explore this question in the context of subscription television consumption. To reveal collective resource integration activities, we use a systemic research approach to semi-structured interviewing with households (Fleuridas, Nelson, & Rosenthal, 1986). We thereby contribute in several areas. First, we enhance understanding of the nature and role of the resources actors integrate. Notably, whereas prior literature describes consumers as individually owning or having access to integrable resources (Kleinaltenkamp et al., 2012), we find that resources may be shared, and that consumers' restricted agency over these shared resources may result in value destruction. Second, we substantially extend the literature on resource integration by unpacking what it actually involves. We find that consumers do not just combine and apply resources, as is commonly assumed (Peters et al., 2014); rather, they assemble, master and optimize them, plan and fine-tune usage events that apply them, and reflect on the extent to which these activities succeed in creating value. Third, the literature predominantly assumes that resource integration is undertaken by individual actors (Baron & Warnaby, 2011). While we observe resource integration activities indeed being undertaken by individuals, we find that resource integration can equally be a collective process. For example, whereas consumer learning about resources has been predominantly viewed as an individual issue, we find that resource mastery may be a joint learning activity. We also find that whether value is created or destroyed through collective resource integration is, in part, determined by what we term collaboration quality. We hence also contribute to literature on value creation versus value destruction, a neglected area of value research (Plé & Cáceres, 2010).
We next review the relevant literature on collective consumption, resource integration and value creation and destruction to clarify the research gaps. Subsequent sections detail the method, present findings, and discuss implications for theory and for managers.

Collective consumption contexts
Drawing on Henkel, Boegershausen, Rafaeli, and Lemmink (2017) and Albrecht et al. (2017), we define collective consumption contexts as settings within which multiple consumers are co-present with a resulting positive or negative impact on a focal consumer's experience. Collective consumption contexts vary in several dimensions: whether this presence of other consumers is offline, online or both (Quach & Thaichon, 2017), whether it is synchronous or asynchronous (Boulaire & Cova, 2013), whether other actors such as service personnel are also present (Kelleher et al., 2019), and whether the co-present consumers are known to each other (Argo, Dahl, & Manchanda, 2005). Another dimension is the depth of interaction between consumers. Interactions may be limited to observations of behaviors (Brocato, Voorhees, & Baker, 2012), or simple interactions such as asking another customer to pass a menu (Colm et al., 2017). Alternatively, consumers may co-ordinate their activities more extensively (Kelleher et al., 2019). One reason can be to make use of shared resources, such as physical spaces and facilities: gym users and concert-goers, for example (Kelleher et al., 2019). Deeper, more sustained forms of interaction can occur when consumers participate jointly in collective activities, characterized by 'we-intentions' and the pursuit of shared goals (Epp & Price, 2011). Examples include group trips to restaurants (Pavia & Mason, 2012), team-focused occasions such as action holidays (Arnould & Price, 1993), and in-home experiences such as multi-player gaming (Kelleher et al., 2019).
We explore the collective consumption of subscription television (TV). Subscription TV consumers have access to a digital or satellite service that provides a package of channels, some of which may be exclusive, and related benefits. In the UK, where our study is situated, roughly 60% of households subscribe to such a service (Broadcasters Audience Research Board, 2018). The average 3.5 h per day that British people watch TV (Broadcasters Audience Research Board, 2019) forms a rich slice of life which varies on many of the collective consumption dimensions we have identified. Consumer co-presence may be online where, for instance, multiple people stream broadcasts simultaneously, discussing it via social media. At other times fellow consumers are present offline within the consumption environment, which is frequently their home. Consumption may be synchronous (for example, where people watch together) or asynchronous (when people watch individually, but coordinate over such resources as who is occupying which room at what time). Service personnel are mostly absent, though they may enter the context remotely when technical help is needed. Within a household, consumers are mostly known to each other. However, their depth of interaction varies: consumption can be a solitary activity with minimal interactions with others, while at other times consumption is collective as multiple household members consume jointly in pursuit of shared goals (Epp & Price, 2008;Lull, 2014). Even when watching individually, though, household members may need to co-ordinate their usage of shared resources, such as television sets, recording capacities, viewing spaces within the home, comfortable seating, broadband access, and so on. Although some consumption occasions are primarily individual, then, the diverse range of other, collective consumption events exhibited by this context render it a fruitful one for study of collective consumption. In our empirical work we focus primarily on collective consumption activities within this context as they occur offline among co-habiting household members.
We next discuss the literature on value creation before introducing our focal analytical lens: resource integration.

Value creation in collective consumption contexts
Traditional conceptions of value assume it is embedded in firms' offerings in the form of quality attributes, transferring value to the consumer at the point of product or service delivery (Zeithaml, 1988). More recently, separate streams of literature have highlighted the active role of the consumer, rather than just the provider, in value creation (Alves, Fernandes, & Raposo, 2016). This theme is central to contemporary narratives such as customer-dominant logic, which conceives value as formed through consumers' embedding of services within their wider activities (Heinonen & Strandvik, 2015), and servicedominant logic, which describes value creation as occurring within networks of actors, including consumers, service providers and other businesses (Vargo & Lusch, 2011). Common to these narratives is an assertion that to understand value creation requires an understanding of the consumer's role. Grönroos and Voima (2013) illuminate conceptually how the roles of customers and firms in value creation may vary in different consumption contexts. Where a customer and firm are co-present, their direct interactions within this "joint sphere" (Grönroos & Voima, 2013, p. 141) provide the basis for value co-creation. Where firm personnel are not present, as is typically the case in our empirical context, customers' interactions with firm-provided resources (rather than their personnel or processes) underlie a process of independent value creation by consumers within a "customer sphere" (Grönroos & Voima, 2013, p. 142). Grönroos and Voima's (2013) conceptualization of the customer sphere incorporates collective consumption, as they propose that value creation within this sphere may be informed by consumers' interactions with each other, as well as by their interactions with firmprovided resources. In contrast with the joint sphere, however, value creation 1 and consumers' associated roles within the customer sphere in general, and in collective consumption contexts specifically, remain under-explored (Kelleher et al., 2019).

Resource integration in collective consumption contexts
Investigating value creation within collective consumption contexts, then, requires an understanding of consumers' roles. A promising lens for exploring these roles is resource integration, whereby actors with agency, such as consumers, deploy a combination of resources in a process of value creation (Vargo & Lusch, 2004). The integrated resources are owned by or accessible to the consumer (Edvardsson, Kleinaltenkamp, Tronvoll, McHugh, & Windahl, 2014), and include those provided by other actors, such as firms or other individuals (Kleinaltenkamp et al., 2012). Drawing upon Payne, Storbacka, and Frow (2008) and Peters et al. (2014) we view a process as a series of activities, and the resource integration process as comprising activities through which consumers seek to create value, for which consumers may need to develop skills and knowledge (Hibbert, Winklhofer, & Temerak, 2012). Subsequent value appraisal is focused on the properties emerging from resource integration; that is, the structure, mechanism or capacities arising from resource integration that "exceed the sum of its parts" (Peters, 2016, p. 3000). Accordingly, we define resource integration as a process whereby actors combine and apply resources in pursuit of value creation, though we refine this in light of our findings.
Resource integration offers a distinct perspective on product/service use by shifting the focus from using a single offering from one firm to how such offerings are used in combination with a diversity of other resources (Vargo & Lusch, 2004). Within the resource integration narrative, the role of 'user' is one of many played by consumers in value creation; others include 'chooser' (of multiple resources, not just a single product/service), 'payer', 'combiner' and 'disposer' (Hibbert et al., 2012). The role of the firm is to provide resources in the form of products or services, which consumers combine with other resources and then apply (Vargo & Lusch, 2004). Consequently, while a consumer or indeed a researcher might frame a specific period of activity around a product/service use event -for instance, the same hour at home may be variously described as 'watching TV' or 'relaxing on the couch' -a resource integration lens explores the diversity of resources that consumers combine and apply during that period (Grönroos, 2011). Scholars have proposed that resource integration is collaborative in nature (Kleinaltenkamp et al., 2012); interactions between diverse parties are expected to be key antecedents to resource integration (Gummesson & Mele, 2010), and to be shaped by institutional logics and associated shared understandings of norms (Edvardsson et al., 2014).
Despite this well-developed theorizing, empirical studies of resource integration are sparse. Notably, studies within business-to-consumer contexts are limited by their predominant focus on resource integration by individual consumers. Baron and Harris (2008) and Baron and Warnaby (2011) evidence physical, social and cultural resources, which individual consumers integrate when executing a campaign to save a local cinema and when using a library. In doing so, they provide partial empirical support for Arnould, Price, and Malshe's (2006) theoretical model of consumers' integrable resources, which identifies physical, social and cultural operant resources, alongside economic and material operand resources. Operand resources are typically tangible and require an action to be performed upon them during value creation, while operant resources are typically intangible and are used in actions upon operand resources (Arnould et al., 2006;Vargo & Lusch, 2004). Overall, there is a lack of empirical insight regarding consumers' integrable operand and operant resources, particularly in contexts outside the firm-managed servicescape (Arnould et al., 2006). Furthermore, Baron and Harris (2008) and Baron and Warnaby (2011) do not examine what the resource integration process itself consists of. Although Baron and Harris (2008) allude to joint activities where individuals have resources in common (e.g. cultural values), the nature of these activities has not been given explicit focus. A further limitation of these prior works is their emphasis on value creation, largely ignoring any value destruction outcomes -a prevalent weakness within the wider value literature (Plé & Cáceres, 2010).

Other lenses on collective consumption
Work on resource integration in collective consumption contexts, then, has a number of limitations. Related perspectives or lenses do, however, shed some light on collective consumption. We next briefly review what is known from three of these lenses: customer-to-customer interactions, social practice theory, and customer engagement.
First, a number of studies explore customer-to-customer interactions as drivers of firm performance (Heinonen, Jaakkola, & Neganova, 2018). Reflecting Grönroos and Voima's (2013) customer sphere, these studies show that in addition to customers' interactions with firms, customer-to-customer interactions inform customer satisfaction and perceived value (Gruen, Osmonbekov, & Czaplewski, 2007;Martin & Pranter, 1989). Exploring customer-to-customer interactions within firm-managed service environments, Martin and Pranter (1989) and Wu (2007) highlight that these interactions are driven, in part, by a need for consumers to share service spaces and facilities. In contrast, Heinonen et al. (2018) contend that many such interactions occur outside the service environment. Heinonen et al. (2018) and Gruen et al. (2007) identify drivers of interaction such as a desire for knowledge exchange and for social interaction, and interaction activities such as problem solving and community building. As Heinonen et al. (2018) and Gruen et al. (2007) observe, customer-to-customer interactions can result in value creation or destruction.
This consumer-to-consumer lens, then, is helpful in understanding how consumers relate in general. Some of this work also touches on what resources they draw on in these value-creating interactions, such as skills and shared spaces. Our resource integration perspective expands on this work by giving such individual and shared resources explicit focus, examining a wider diversity of resources consumers apply, exploring their agency in accessing these resources, and eliciting in more detail how consumers combine and apply them.
A closely-related second lens, social practice theory, also relates in large part to how consumers interact, exploring people's activities and outcomes (Feldman & Orlikowski, 2011). A practice is defined as "a routinized type of behavior" (Reckwitz, 2002, p. 249). A social practice lens has been beneficially applied to some collective consumption contexts. For example, Schau et al. (2009) delineate twelve common value-creating practices across a diversity of brand communities, while 1 Note that our research objective refers to 'value creation' rather than 'value co-creation'. Different authors vary in their definition of co-creation. For example, Vargo and Lusch (2004) argue that customer's use activities are necessary for value to emerge, so the customer and not just the firm creates value; in this sense, value is always co-created. Others in the service logic stream (e.g. Grönroos & Gummerus, 2014) prefer to reserve use of the term co-creation to describe value creation processes within Gröonroos and Voima's (2013, p. 140) "joint" and "customer spheres", which we have discussed. It is not within the scope of this paper to resolve this terminological debate. Instead, we focus on the wide agreement that the further study of how consumers interact to create value is needed. Carù and Cova (2015) observe a series of practices within collective contexts that may result in value creation or destruction. McColl-Kennedy et al. (2012), contrastingly, maintain an individual customer focus and identify eight co-creation activities which health-service customers carry out. In much of this work, resources are under-emphasized or absent. However, the conceptualization of practices includes use of artefacts (Warde, 2005), and Shove and Pantzar (2005) assert that products only have value when integrated into practices. Empirical work here is limited, but in deriving a typology of consumption practices Holt (1995) explores individual consumers' valuecreating activities centered around specific firm offerings (e.g. a sporting event), concluding that such offerings support practices and stimulate interactions between consumers. We extend this work by focusing on how these activities draw on diverse resources, and by investigating collective activities to combine resources, particularly when resources are shared.
Customer engagement, a third complementary lens, is described by Storbacka, Brodie, Böhmann, Maglio, and Nenonen (2016) as an antecedent of resource integration and subsequent value creation. It therefore has the potential to illuminate customers' role in value creation by articulating an underlying factor in their participation. A typical definition is a "customer's particular psychological state induced by the individual's specific interactive experiences with a focal engagement object (e.g. a brand)" (Brodie, Hollebeek, Jurić, & Ilić, 2011, p. 257). This psychological state has cognitive and affective dimensions; some also add a social, or social identity, dimension (Mollen & Wilson, 2010). Customer engagement manifests in a variety of engagement behaviors, which include the provision of resources during interactions with the focal firm or other stakeholders (Jaakkola & Alexander, 2014) in activities such as co-operating with firm employees and helping other customers (Verleye, Gemmel, & Rangarajan, 2014).
An engagement lens is useful in exploring consumers' dispositions to undertake resource integration and in understanding their psychological states that accompany their interactions with service organizations. As with the other lenses we have reviewed, resources have tended as yet to be in the background rather than the foreground, with some partial exceptions such as Jaakkola and Alexander (2014). Another limitation of work to date is that it has primarily adopted an individual theoretical and empirical perspective, engagement having initially been conceived as a property of a single person. Recent commentary proposes that the construct can be extended to collective actors such as managers in a customer firm (Kleinaltenkamp, Karpen, Plewa, Jaakkola, & Conduit, 2019), but conceptual and empirical work taking this perspective in a consumer context is awaited.
While these three lenses, then, shed light on how consumers interact in collective consumption context, a resource integration perspective promises to complement these with explicit focus on the diverse objects, skills and other resources that consumers draw on, the process by which these resources are combined and applied in pursuit of value creation, and the extent to which this process is a collective rather than just an individual one. Value creation may not always prove the outcome of such resource integration; we therefore turn next to the literature on value destruction.

Value destruction
Consumers perceive value created through resource integration as a sense of enhanced wellbeing or being "better off" (Grönroos, 2011, p. 282). In referring specifically to the 'pursuit of value creation', our definition of resource integration allows for value destruction, where resource integration results in perceptions of diminished wellbeing or being worse off (Smith, 2013).
The resource integration narrative theorizes several ways in which consumers' behaviors might result in value destruction. Focusing on individual consumers, Hibbert et al. (2012, p. 252) suggest that a failure to develop the skills required for resource integration may be "costly in terms of the overall value generation". Acknowledging the interactions inherent within resource integration, Plé and Cáceres (2010) conjecture that value is destroyed where one party misuses resources -that is, integrates resources in a manner deemed inappropriate or unexpected by another. Building on this, Smith's (2013) empirical study evidences value destruction where firms misuse customer resources; for instance, by accepting payment for a service they later fail to deliver, resulting in the loss of funds and reduced self-efficacy among consumers. Some studies have also examined interactions between actors within the consumption context as the cause of value destruction. For instance, adopting a social resource theory lens, Quach and Thaichon (2017) highlight consumers' limited online interactions with brands as destroying perceptions of value. Within a face-to-face service setting, Echeverri and Skålén (2011) evidence value destruction from social practices when, for example, consumers lack an understanding of service delivery processes. Similarly, in their study of business-tobusiness solutions Macdonald et al. (2016) suggest that poor-quality joint processes can result in value destruction.
The literature, then, evidences some respects in which resource integration can destroy value, such as resource misuse, and individuals lacking the relevant skills and understandings. Our empirical study, to which we now turn, extends this prior work, identifying further respects in which resource integration can destroy, rather than create, value.

Data collection
We conducted group interviews within the homes of 29 UK households that had used a subscription TV service for a minimum of 12 months. Household types included couples, families with children of varying ages, and young professionals sharing a home; single-person households were excluded due to the focus on collective consumption. 2 Each interview comprised two or more members of the household. 71 individuals took part, ranging from eight to 57 years old. Households were located throughout the UK and were of varying socioeconomic status. Households were paid £75 for each group interview. Interview length averaged 65 min.
The semi-structured interviews borrowed from family psychotherapy in adopting a systemic approach (Clarke, 2012;Tomm, 1988), combining two forms of questioning: circular and lineal. Circular questioning explores the interconnectedness of members of a collective, enabling interviewees to reflect upon themselves as a system (Fleuridas et al., 1986). This facilitates the study of collective consumption by exploring how household members undertake joint activities. Two different types of circular questioning were used: pattern questions and difference questions (Tomm, 1988). Pattern questions explore recurrent patterns of behavior and perception. We used pattern questions to elicit recurrent resource integration activities relating to TV consumption (for example, 'as a household, do you have a routine in the evenings?' and 'what role does subscription TV play in this routine?'). Subsequent probing (for example, 'can you tell me more about those occasions?' and 'who takes part in that?') ensured detailed descriptions of activities and associated resources. Difference questions identify distinctions between individuals, exploring variations in routines and value perceptions among household members (for example, 'does it always happen that way? Why/why not?' and 'who in the house is most committed to/ enjoys that activity the most/least?') (Tomm, 1988). Lineal questioning is used to identify cause and effect within systems (Ryan & Carr, 2001) and was used alongside circular questioning to elicit value creation or destruction resulting from the various activities (for example, 'what do/ don't you like about these routines?' and 'what effect does that have on you/the family?').
Applying a systemic approach in this manner illuminated the social reality of collective consumption within a household. One of the authors is a practicing systemic family psychotherapist and attended each interview with a second member of the research team. This facilitated the effective use of systemic questioning, while the presence of two interviewers ensured responses from all participants were probed. As far as possible, interviews resembled a conversation with discussion between household members encouraged (Peräkylä & Ruusuvuori, 2011). Siting the interviews within the home enabled participants to draw upon contextual cues such as their television equipment and room layout, enriching the discussion and enhancing the researchers' understanding (Rosenthal & Capper, 2006).

Data analysis and interpretation
We collated 1143 pages of transcripts from the 29 interviews. Following Spiggle (1994), we applied a systematic, multi-stage approach to data analysis and interpretation which involved multiple passes through the data. To reduce researcher-induced bias the research team worked as an interpretive group (Thompson, Locander, & Pollio, 1989), critically reviewing each other's analysis after each stage. The stages were as follows: 1. Members of the research team read each transcript, and isolated sections of text that described resource integration. That is, we identified participants' descriptions of combining and applying resources, including any neighboring material on the consequences of these activities. This served to identify meaningful portions of text and exclude sections of text containing no relevant information (Spiggle, 1994). 2. Next, we coded the reduced volume of text to inductively produce a set of resources and resource integration activities (Bryman, 2016;Spiggle, 1994). This stage proceeded iteratively: where data yielded a new resource or resource integration activity, a new code was created. This resulted in the categories of resources and resource integration activities shown in Fig. 1 (further details of resource integration activities are presented in Table 1). In coding the resource integration activities we also distinguished between collective and individual activities. Activities were coded as collective where multiple household members participated jointly. Conversely, activities where single household members integrated resources independently were coded as individual. 3. For each resource integration code derived in stage 2, we examined participant descriptions of activities under that code to explore similarities and differences in their value outcomes. First, we distinguished instances where resource integration resulted in value creation (enhanced wellbeing or feeling 'better off') from those resulting in value destruction (reduced wellbeing or feeling 'worse off'). We then examined instances of value creation and destruction to establish how or why an activity might result in value destruction rather than value creation. This analysis informed the commentary about each resource integration activity in Sections 4.1 to 4.6 below. 4. Finally, two scholars not involved in the study reviewed a portion of the data against the emergent codes. The researcher team and independent coders then discussed any areas of difference and agreed a revised interpretation.

Findings
From our findings we develop a picture of resource integration in a collective consumption context. We observe six activities through which household members integrate resources in the creation, or destruction, of value: resource assembly, resource mastery, resource optimization, usage event planning, real-time usage design, and resource reflection. We find that each may be undertaken individually and collectively. Table 1 presents these activities with illustrative quotations.
We observe that these activities are typically interwoven, and may evolve through their continued enactment. Moreover, whether resource integration results in value creation or destruction is influenced by consumers' varying agency when integrating shared resources, by the extent of consumers' collective skills and knowledge, by human error and resource failure, by resources that lack the characteristics necessary for effective integration, and by the behaviors of those involved in resource integration.
We explore these observations in the following sections, in which we discuss each of the emergent resource integration activities in further detail, highlighting instances of value creation and value destruction. While each activity may be carried out individually, we focus solely on collective activities within this narrative. Data analysis also yielded five categories of integrable resources, which we discuss as we present the first activity: resource assembly (Section 4.1). Fig. 1 summarizes the study findings.

Resource assembly and integrable resources
Resource assembly involves the acquisition and disposal of resources in order to establish a desirable set for integration. In identifying resource assembly we highlight that, rather than simply having access to stocks of integrable resources as is typically assumed within the literature (Arnould et al., 2006), consumers' agency in value creation extends to assembling resources for subsequent integration.
As Fig. 1 illustrates, households assemble a diversity of resources. We categorize these as personal resources (such as the skills, tastes and preferences of household members); socio-cultural resources (such as relationships and norms); focal service provider resources (including goods such as hard-drive recording equipment and core offerings such as program transmissions); complementary service provider resources (such as internet facilities and electricity to the household); and service environment resources (such as rooms, furniture and TV sets). These categories include both operant (e.g. personal skills and knowledge) and operand (e.g. physical goods from the focal service provider) resources.
Whereas the literature dominantly explores resource integration undertaken by individual actors (e.g. Baron & Warnaby, 2011), interviews provided instances of collective, as well as individual, resource assembly. Resource assembly occurs collectively when acquisition and disposal tasks are agreed upon or executed by multiple consumers as a combined effort. Married couple Joanna and Pete demonstrated collective resource assembly when subscribing to their TV service: Joanna: "My dad bought us the basic package as a wedding present. We then added the bits we wanted: the lifestyle channels for me and the sports for Pete." Pete: "Before the wedding things were a bit tight money-wise. Afterwards we could afford to do that." Joanna and Pete's usage of 'us' and 'we' highlights the collective nature of this resource integration activity: Joanna and Pete jointly designed their ideal package, integrating (among others) focal service provider resources (channel packages) and personal resources (their disposable income) in the creation of value. Joanna and Pete share some of the resources integrated: their disposable income, for example. This contrasts with the resource integration literature's dominant view that resources come from different actors, each provided by a single actor (Kleinaltenkamp et al., 2012), and then integrated by one focal actor. Value destruction can occur where a consumer has limited agency over these shared resources. For example, married couple Nick and Holly's collective resource assembly activities result in Nick feeling worse off: Nick: "I used to watch it [subscription TV] with my dad. We had the sports channels, so we'd have the football on. We haven't got that here. She won't let me have it!" Holly: "No, because it would cost a lot of money each month and it would literally be on the telly all day, every day." Nick: "I would love to have it. I'm a massive football fan." The need for Nick and Holly to agree how much of their joint income should be spent on subscription TV, and how frequently their shared space should be dedicated to football broadcasts, results in value destruction for Nick, at least. So, despite collectively integrating (among others) focal service provider resources (channel packages) and shared personal resources (disposable household income), value destruction has occurred from Nick's perspective. Here, a need to collaborate over the integration of shared resources (in this case, the TV set, the room, and the household budget) limits Nick's agency as a resource integrator, resulting in uneven value perceptions between the couple.

Resource mastery
Consumers undertaking resource mastery integrate resources in the development of skills and knowledge required for future resource integration activities, such as usage event planning and real-time usage design. Resources integrated in resource mastery include pre-existing skills and knowledge (e.g. prior experience of equipment), supplier resources (e.g. user information and guidance), and relationships with other consumers who can facilitate learning. Hibbert et al. (2012) theorize that consumers undergo a process of self-directed learning in order to become effective resource integrators, often triggered by a perceived need to learn. Our findings provide empirical support for this, as resource mastery encompasses both self-directed and firm-managed learning, and often takes place in response to receiving a new or updated resource from the focal service provider. We add to Hibbert et al. (2012) by noting that, as well as involving individual learning, resource mastery can be carried out collectively. For instance, Bob and Vivian, a retired couple who spend much time watching TV together, discussed their collective resource mastery: Bob: "We've just mastered it after five years. There's still bits we can't..." Vivian: "Yeah, there's a lot of features we don't use." Bob: "Well there's a lot we haven't even tried to use." Vivian: "We switch it on and as long as there's a picture, we're OK." Collective resource mastery is evident in Bob's use of 'we've', suggesting that skills development has been a combined endeavor. Vivian indicates that she and Bob have developed a mutually agreeable level of skills and knowledge ("... as long as there's a picture, we're OK") which does not require further exploration of their service's features. Resource mastery, then, can be a collective process, and furthermore may support joint goals. This resource integration activity also echoes Hughes, Vafeas, and Hilton (2018) in observing the enhancement of operant resources (in this case, skills and knowledge) as a result of resource integration.
Conversely, participants described value destruction when their resource mastery efforts had proved insufficient thus far. For instance, despite being able to use many aspects of their TV subscription well, Sara and Richard find some of their collective skills inadequate: Sara: "You miss [previous provider's] 'on demand' service, don't you?" Richard: "Yeah. We've got it on this package but we don't know how to use it." In noting that 'we don't know how to use it', Richard is confirming their combined inability to effectively draw upon their 'on demand' service. Value destruction is evident in Richard's 'missing' his former service. Our findings therefore provide empirical support for Hibbert et al.'s (2012) conjecture that a failure to learn about resources can destroy value. We add that this failure to learn can be collective as well as individual: if one of the couple could use the 'on demand' service, that might suffice, and furthermore they might have taught the other.

Resource optimization
Consumers do not just master resources: in a resource integration activity previously unreported in the resource integration literature, they can also optimize them. Consumers undertake resource optimization to ensure that their resources remain integrable and to maximize their effectiveness. Collective resource optimization occurs when such activities are carried out or agreed upon by multiple consumers. A commonly-described example is the repairing of faulty supplier-provided technology, which entails the integration of diverse resources from the focal service provider and complementary providers, as well as shared personal resources such as disposable income. Other collective resource optimization activities relate to the ongoing management of recorded TV content. Nick and Holly describe how their recording facility often reaches full capacity, requiring coordinated attention: Nick: "We end up with stuff on there that we just don't want." Holly: "Stuff that we don't watch, or we've watched but not deleted." Nick: "We have to go through it every now and again. I'll go: 'do you want this, do you want that?'" Interviewer: "Do you take it in turns to do that?" Holly: "We do it together, really." Holly's description of 'doing it together' captures the collaborative nature of their resource optimization. By integrating focal service provider resources (recording equipment) and socio-cultural resources (an ability to make joint decisions, born of their relationship), among others, Nick and Holly enable future value creation.
For others, resource optimization was not as collaborative, destroying value. Failing such collaboration, actors' agency to optimize shared resources may be limited, with resultant value destruction for at least some household members. Teenage twins Mark and Dean describe such a scenario: Mark: "Mum's the worst..." Dean: "Me and Mark will save things to watch at a specific time, say when our friend comes round, and she just deletes it!" Mark: "She doesn't give you any option; she just wants to save her stuff." Mark and Dean thus highlight the potential challenges of sharing resources (in this case, the recording capacity) and the potential for resultant value destruction.

Usage event planning
Usage event planning involves the integration of resources to facilitate a future service usage episode, which in the case of subscription TV generally involves viewing broadcast material. Time-related resources are prominent in this planning activity. Relevant resources include household members' personal time, as well as technology that can help with its management such as automated reminders and, of course, Table 1 Resource integration activities.

Resource integration activity Description
Collective activity examples Quotations from consumers in a subscription TV context

Resource assembly
Acquisition and disposal of integrable resources Subscribing to subscription TV. "We subscribed because friends were saying: 'you can't live without it.'" Enhancing or reducing subscription. "We add the sports channel during the football season, then take it off again."

Resource mastery
Developing the skills required for effective future resource integration Learning how to use technology. "We can watch through the games console but we haven't got used to that yet" Resource optimization Ensuring resources are in a condition that enables their effective combination and application.
Deleting recorded content to allow new items to be recorded. "We delete things we've viewed to free up space" Restoring damaged/non-functioning resources. "We had an incident with the kit. They put in new cabling" Usage event planning Preparing for usage (of focal service provider offering) Recording subscription TV content. "We record programs and series and watch them when we want to" Using provider technology to prompt viewing. recording. Collective usage event planning involves multiple consumers in collaboration; for example, the recording of a show on behalf of a fellow household member, or the synchronizing of diaries to watch TV with (or without) others. Ken describes how he, his wife, and their two sons plan communal viewing amid their hectic schedules:

"You hit a button and it reminds us when it's on"
"We look at the listings and when the big movies come through, the family ones, we'll book that time in to watch it together as a family." Here, the family integrates focal service provider resources (the movie package and schedule information), service environment resources (their TV set) and their personal resources (notably time and attention), enabling them to watch a movie together.
Evident within our findings relating to usage event planning is the subsequent enhancement of operand resources. For example, program recording functions (a focal service provider resource) often become personalized repositories of favorite TV shows or momentous sporting events, allowing them to be viewed repeatedly and indefinitely. Whereas resource optimization activities ensure that resources maintain their integrable characteristics, our findings indicate that resources can also be enhanced through usage event planning. Previous empirical studies of resource integration have highlighted subsequent enhancements in customer operant resources (e.g. Hughes et al., 2018). We add that operand resources may also be enhanced through resource integration. We represent this relationship between resources and resource integration activities through the use of double-headed arrows in Fig. 1.
Conversely, we observe value destruction from collective usage event planning where this activity is in some way ineffective. Primarily, this occurred due to human error -for instance, where a household member forgot to record a transmission -or where resources failed to function as required, as married couple Kate and Fred describe: Kate: "When we went away on holiday it [the recording function] basically blew up." Fred: "The thing just goes pop if you're not careful and you end up with nothing." This echoes Macdonald et al. (2016) in evidencing the importance for value creation of high quality, reliable resources as well as effective resource integration activities. In addition, we further extend the work of Hibbert et al. (2012) by observing that, even where consumers possess the skills and knowledge for resource integration, a simple human error might nonetheless result in value destruction.

Real-time usage design
Even after any usage event planning, consumers are often far from passive during the service usage episode of watching TV. Rather, they conduct real-time usage design activities to fine-tune these viewing experiences and enhance the value created. Thus, as we anticipated earlier (in Section 2.3), rather than creating value by simply executing a 'use process' (Vargo & Lusch, 2004), consumers combine diverse resources in addition to those provided by the firm, designing activities to create value from the usage event (Grönroos, 2011;Vargo & Lusch, 2004). Real-time usage design involves choosing what show to view, selecting a viewing platform (such as a TV set or a tablet) and location within the home, and controlling the experience by such means as fastforwarding through advertisements or pausing the transmission. In collective real-time usage design activities, consumers work together to manage their viewing experiences; for example, when household members choose and watch a show together, as Kate and Fred describe: Kate: "We start from the top of the listings and work our way down." Fred: "We just go: 'Nothing on that page, nothing on that page, oh that looks vaguely interesting'." Kate: "Certain shows are a combined focus for us, so we'll usually go for something like that." Fred: "We like the 'whodunits?', that sort of thing." Kate: "Yeah, they're good!" Fred: "We try to work out who did it." Kate: "You'll go, 'it was him' and I'll say, 'no it can't have been'. Then the daft conversation starts about what might have happened. < laughter > ".
Within this scenario, Kate and Fred collaborate in selecting and viewing the broadcast material and in their related discussions. Through integrating socio-cultural resources (the friendship born of their relationship), personal resources (their tastes and sense of humor) and focal service provider resources (the program transmission), value is created. In contrast, Joseph describes value destruction from his realtime usage design activities with his housemate, Loretta: Joseph: "Loretta gets angry with me when I talk to her during the soaps. She's very high maintenance." Loretta: "I do! In the end I just pause it, let him finish what he wants to say, then go back to watching." Like Kate and Fred, Joseph and Loretta have integrated their personal resources (a shared enjoyment of the relevant transmission), focal service provider resources (the transmission) and service environment resources (a shared space within their home). However, Joseph's annoyance (as evidenced by his description of Loretta as 'high maintenance') and Loretta's anger indicate some value destruction. It seems that, even where resources are combined and applied effectively, the conduct of those involved in collective resource integration may result in value destruction. Another common example of value destruction from collective real-time usage design related to difficulties in selecting a show to view, as Joanna and Pete describe: Joanna: "It's good to have that choice but sometimes it's hard to choose what to watch, isn't it?" Pete: "Some evenings, when we do sit down, we think: 'God, what do we even watch?'" Despite integrating focal service provider resources (the selection of shows), service environment resources (space within their home) and personal resources (time), Joanna and Pete experience choice overload (Scheibehenne, Greifeneder, & Todd, 2010) and struggle to decide what to watch. In this instance, excessive or inadequately guided choice represents a resource that is less than optimal in supporting Joanna and Pete's habits for selecting programs.
These findings therefore suggest that whether value is created or destroyed depends, in part, upon the extent to which resources possess the characteristics necessary to support the resource integration activity -a feature we term integrable quality. Adopting the traditional view of quality as referring to product or service excellence or superiority (Parasuraman, Zeithaml, & Berry, 1988), an extensive choice of program options would seem to constitute high quality and therefore support resource integration and value creation (Macdonald et al., 2016). However, as Joanna and Pete indicate, the integration of resources may nonetheless result in value destruction, where resources are of low integrable quality.

Resource reflection
Finally, consumers undertaking resource reflection discuss their resource integration activities with other consumers, both within and outside the household. Discussions may focus on mutually viewed broadcasts or on recommendations of shows to watch. Alternatively, consumers may discuss resources, such as equipment, or their appraisals of the service itself. Trevor describes resource reflection as follows: "At work we talk about what we've watched. Usually that involves bragging rights over a colleague that doesn't have subscription TV: 'Oh, you can't get that, can you, Ted?'" Within this scenario, Trevor and his colleagues are integrating socio-cultural resources (e.g. social conventions that allow good-natured teasing) with service provider resources (a TV subscription).
Participants also described value destruction within some of these resource reflection activities. A commonly-cited example was 'spoilers', when a fellow consumer unwittingly revealed details of a dramatic plot twist or sporting result before others had seen the program. Such incidences of value-destroying resource reflection were reported in both face-to-face and digitally-mediated communications, such as on social media. Another source of value destruction in resource reflection was contradictory opinions or even conflict. Claire, a 53-year old wife, mother, and high school teacher, describes such an experience: "It's seen as trendy or intelligent to say, 'Oh, we don't watch that, it's such rubbish'. We were with friends in a wine bar a year or so ago and we had this very conversation. We got into a terrible row." Here, conflict with Claire's friends due to their disparagement of her reported viewing results in value destruction for Claire, at least. As with Loretta and Joseph's value destruction in the previous section, these findings highlight the conduct of actors undertaking resource integration as a potential contributor to value destruction. We propose that collaboration quality is instrumental to value creation. We define collaboration quality as the extent to which consumers coordinate during resource integration to create value for all parties in the consumption context.

Resource integration: interwoven activities
Our data highlighted a dynamic interweaving of these six resource integration activities: specifically, the undertaking of a resource integration activity may stimulate and inform subsequent resource integration. For example, Fiona describes how resource reflection informed her real-time usage design and usage event planning activities with her partner, Leon: "We'd heard about the show but didn't plan to watch it. Then a friend came to spend an evening with Leon and me, and she kept rambling on and on about how good it was. In the end she made us sit and watch it while she was there. After that we wanted to see the rest, so we recorded them." As another example, participants described resource assembly as triggering resource mastery, for instance where the addition of a new service to the household required a renewal of skills required for resource integration. Equally, participants noted how real-time usage design might stimulate and inform resource assembly; for example, where a household lacked sufficient seating to enable comfortable viewing together. Consequently, just as we have observed that resources evolve, we find that resource integration activities evolve and are adapted through their continued enactment.

Theoretical implications
This study investigated how consumers in a collective consumption context integrate resources in pursuit of value creation. In particular, we explore a context where the service provider is mostly present only in the resources it provides, where individuals may consume separately or jointly, and in which consumers may need to integrate shared resources to create value. Fig. 1 presents the emergent framework. We find that, within this context, consumers integrate a diversity of resources through six activities. These may be undertaken individually or collectively. Whether these activities result in value creation or destruction is influenced by consumers' agency when integrating shared resources (described in relation to resource assembly in Section 4.1 and resource optimization in Section 4.3), consumers' collective skills and knowledge (discussed in relation to resource mastery in Section 4.2), human error or resource failure (highlighted in our discussion of usage event planning in Section 4.4), resources' integrable quality (introduced with real-time usage design in Section 4.5), and consumers' collaboration quality (discussed alongside real-time usage design in Section 4.5 In addition, in collective consumption contexts, some resources may be shared by multiple consumers, creating challenges around their access and integration.
Proposition 1: Consumers' restricted agency over shared resources within collective consumption contexts may result in value destruction.
Operant resources may be enhanced through resource integration.
Both operand and operant resources may be enhanced through resource integration.
Proposition 2: Both operand and operant resources should be viewed as evolving through resource integration. Resource quality (perceived excellence or superiority) supports value creation.
Resources are also judged on their integrable quality.
Proposition 3: Value creation and destruction depend on the integrable quality of resources, and not just their perceived excellence or superiority. How?
Resource integration activities Consumers combine and apply resources in pursuit of value creation.
In addition, consumers master and optimize resources, plan and fine-tune service usage events applying them in real-time, and reflect upon their resource integration activities. These activities are interwoven and evolve over time.
Proposition 4: Consumers' agency as resource integrators includes mastering and optimizing resources, and planning and fine-tuning their application. These activities are interwoven with, and may be inseparable from, usage events. Proposition 5. Resource integration activities should be viewed as dynamic. Individual consumers may undergo selfdirected learning to become effective resource integrators and avoid value destruction.
Resource mastery can be a collective activity as well as an individual one. Despite effective resources mastery, human error may limit the efficacy of resource integration activities.
Proposition 6: Collective resource mastery can enable value creation, but value destruction may still occur where consumers with the required skills experience human error or resource failure.

By whom?
Interactions between consumers Individual consumers carry out resource integration within the social context.
In addition, multiple consumers may carry out resource integration collectively in pursuit of shared consumption goals.
Proposition 7: Beyond being socially contextualized, value creation in collective consumption contexts may require collective resource integration activities. Actor interactions are key antecedents to resource integration, and are governed by shared rules and norms.
Additionally, consumers' collective behavior during resource integration may result in value destruction for some or all of those involved.
Proposition 8: In collective consumption contexts, consumers' collaboration quality is instrumental to value creation. and resource reflection in Section 4.6).
Our findings make several contributions to literature on resource integration and value creation and destruction within collective consumption contexts. We summarize these contributions and eight resultant propositions in Table 2, which is used to structure the discussion that follows. The table covers three topics: the resources that are integrated; the activities comprising the resource integration process; and how multiple consumers interact within this process.
The first three propositions, then, concern the resources that are integrated to create value. Arnould et al. (2006) theorize a typology of operant and operand resources which individual consumers might integrate, receiving empirical support for the operant portion of their model from prior studies in firm-managed servicescapes (Baron & Warnaby, 2011). We extend this by evidencing a diversity of operant and operand resources that are integrated by consumers in a context outside the firm-managed service environment, broadly confirming Arnould et al. (2006). Furthermore, whereas individual actors have been assumed to be the providers of resources, whether through ownership or other access (Kleinaltenkamp et al., 2012), we find that in a collective context, the requirement to share certain resources may restrict consumers' access to them and limit consumers' agency as resource integrators, which can lead to value destruction. Hence our first proposition: P1. Consumers' restricted agency over shared resources within collective consumption contexts may result in value destruction.
Secondly, we evidence the dynamic and evolving nature of resources as a result of their integration. Prior studies have observed the enhancement of actors' operant resources (for example, their skills and knowledge) through resource integration (Hughes et al., 2018). In contrast, we find that both operant and operand resources may be enhanced through resource integration, with the potential to improve the resource integration activity and enrich the value created: P2. Both operand and operant resources should be viewed as evolving through resource integration.
Our third contribution regarding resources relates to the criteria against which resources are judged by consumers. We complement the notion of product or service quality, defined as the perceived excellence or superiority of the product/service (Macdonald et al., 2016;Parasuraman et al., 1988), by proposing that integrable quality -the extent to which resources possess the characteristics necessary to support resource integration activities -influences whether value is created or destroyed. We subsequently offer our third proposition: P3. Value creation and destruction depend on the integrable quality of resources and not just their perceived excellence or superiority.
Propositions 1 to 3 echo and also extend the work on customer-tocustomer interactions we reviewed earlier. For instance, Martin and Pranter (1989) and Wu (2007) highlight that value creating or destroying interactions can occur due to a need to share space or facilities within the service environment, suggesting (in our terminology) some kind of collective resource integration. We add that agency over these shared facilities may vary, and when it is restricted, value destruction may result. We also add that consumer-to-consumer interactions around resources may optimize the resources and not just apply them; this and other insights into the specific activities comprising resource integration are expanded on in the sections that follow.
Our observation of the collective nature of some resource assembly activities is not always reported in previous work taking a social practice lens; Holt (1995), for instance, focuses on the individual as (in our terms) the resource integrator. This finding is, however, clearly conceptually consistent with a social practice lens, as is evident from more recent empirical work (Carù & Cova, 2015). Again, we add clarity on the importance of access to resources, and how consumers' agency in that regard can constrain value creation. This mirrors the recent findings of Kelleher et al. (2019) about access to knowledge and skills, and the service provider's role in facilitating that.
The next three propositions in Table 2 concern resource integration activities. While Macdonald et al. (2016) examined the resource integration process in a B2B context, we have seen that prior B2C work (Baron & Harris, 2008;Baron & Warnaby, 2011) has not elicited what consumers actually do when integrating resources. We address this gap by identifying six resource integration activities that consumers undertake. This leads us to revise our earlier definition of resource integration (Section 2.3) as follows: Resource integration is a process consisting of activities to assemble, master, and optimize resources, to plan and fine-tune usage events in real time, and to reflect upon previous activities.
The picture that emerges of this complex process of integrating diverse resources echoes a social practice perspective, for which artefacts of varying sorts are central to many practices (Warde, 2005). However, we extend the existing empirical work applying that lens in the holism with which resources are viewed -as opposed to a primary focus on resources from a focal firm -and in the extended nature of these resource integration activities. For instance, Holt's (1995) study of consumption practices illuminates activities directed towards a specific 'consumption object' -that is, the focal firm's offering. 3 We add that integration of resources from multiple sources takes place before (through resource assembly or usage event planning) and after (through resource reflection) use, as well as during it (through real-time usage design). From the perspective of the customer engagement literature, the six activities we identify might be viewed as individual and collective customer engagement behaviors. In addition, we note the dynamic interweaving of resource integration activities, whereby a resource integration activity both stimulates and informs subsequent resource integration. Hence our next two propositions: P4. Consumers' agency as resource integrators includes mastering and optimizing resources, and planning and fine-tuning their application. These activities are interwoven and may be inseparable from usage events.

P5.
Resource integration activities should be viewed as dynamic. Hibbert et al. (2012) propose that, in order to become effective resource integrators, consumers must develop the required skills and knowledge, potentially through a process of self-directed learning. We support this proposition through empirical evidence of resource mastery activity which encompasses self-directed learning. In addition, we observe that in collective consumption contexts, self-directed learning may be undertaken collectively and that this activity is instrumental in value creation. We also find that even where consumers have undertaken effective self-directed learning, value destruction may occur as a result of human error or resource failure. Subsequently, we offer our sixth proposition: P6. Collective resource mastery can enable value creation, but value destruction may still occur where consumers with the required skills experience human error or resource failure.
Our final theoretical contributions relate to who is involved in resource integration and how they interact in this process. While the wider narrative around value creation emphasizes the social context and consumer interaction (Grönroos & Voima, 2013), conceptual and empirical work on resource integration specifically has focused predominantly on individuals as resource integrators (Baron & Harris, 2008;Baron & Warnaby, 2011). Similarly, some studies applying customer engagement (e.g. Verleye et al., 2014) and practice (e.g. Holt, 1995) lenses have adopted an individual customer focus. In contrast, we observe that in a collective consumption context, consumers may integrate resources collectively as a joint endeavor. This echoes the emphasis on collective value-creating activities in Heinonen et al. (2018) and Kelleher et al. (2019), while adding more granularity on the role of resources within these activities. This leads to our seventh proposition: P7. Beyond being socially contextualized, value creation in collective consumption contexts may require collective resource integration activities.
The potentially collective nature of resource integration also adds to the value destruction dialogue. We have noted Smith's (2013) finding that value destruction can occur when actors misuse resources. Some instances of value destruction within our findings support this. For example, when Dean and Mark's mother deletes their prerecorded shows (Section 4.3) she is misusing resources in the eyes of Dean and Mark. We thereby add that in collective consumption contexts where resources are shared, value destruction may arise indirectly from resource misuse, due to the subsequent limiting of another consumers' agency. We captured this contribution in Proposition 1. Furthermore, whereas Kleinaltenkamp et al. (2012) and Gummesson and Mele (2010) have suggested that interactions between consumers form key antecedents to resource integration, we add that interacting consumers' conduct during resource integration may also influence whether value is created or destroyed. We have thus derived the term 'collaboration quality' as the extent to which consumers coordinate during resource integration to create value for all parties in the consumption context. This relates to Edvardsson et al.'s (2014) theorization of resource integration as shaped by shared understandings of rules, norms and acceptable conduct that ultimately govern actors' interactions. Where collaboration quality is low it would seem that this governance is weak, leading to value destruction for some or all involved; this requires further focused investigation. Consistent with our data, then, we offer the following proposition as a basis for further research: P8. In collective consumption contexts, consumers' collaboration quality is instrumental to value creation.

Managerial implications
For firms seeking to optimize consumer perceptions of value within collective consumption contexts, an understanding of resource integration is crucial, as we have found many respects in which the process can create value or, alternatively, destroy it. Insight into resource integration activities, and the resources they draw on, might be gained through an adapted service blueprinting process (Bitner, Ostrom, & Morgan, 2008), capturing consumers' points of contact with resources from the provider but equally also from other sources before, during and after usage events. Additional probing would be required to relate resource integration activities to value outcomes. Due to the observed dynamic, evolving nature of both resources and resource integration activities, blueprinting should be undertaken regularly.
The role of the firm is to provide resources for consumers to integrate (Vargo & Lusch, 2004). Our findings highlight that firms should endeavor to provide resources that offer integrable quality alongside more traditional forms of quality (such as reliability), since an absence of either may lead to value destruction. The blueprinting we have suggested could check for this. Firms must also ensure that resources made available to consumers offer scope for evolution. This entails a consumer-oriented focus on functionality, tailorability and compatibility with other resources. Additionally, firms might facilitate resource integration by enabling consumers' access to resources from complementary providers: for example, a subscription TV company might partner with a takeaway food provider to facilitate 'movie night'.
Managers may find that they can limit value destruction through resource design. For instance, firms might use technology to develop individual and collective profiling in order to alert consumers to human error (e.g. failure to record) or resource misuse (e.g. by advising someone that their actions may destroy value for others). Such profiling might also identify consumers' failure to learn the required skills for resource integration; for example, where a feature is underused or used ineffectively. Firms might respond to such instances with support and guidance.
Moreover, service providers might seek ways to facilitate collaboration quality and, in so doing, limit value-destroying consumer interactions. Where services are consumed within the provider's servicescape, this may be straightforward. Within contexts where service personnel are not present, firms might apply their knowledge of common incidences of value destruction to service communications that depict such scenarios (perhaps in a humorous manner) and suggest alternative patterns of behavior. Another option might be to deploy inhome technologies such as Alexa, to identify scenarios in which resource integration is leading to value destruction and to intervene. For example, in the case of housemates Joseph and Loretta, an Alexa-type device might detect one-sided dialogue during a specific period (e.g. when a soap opera is aired on TV) and begin recording the relevant show to ensure nothing is missed.

Research directions
Scholars building upon our findings might further knowledge of value creation by examining how resource integration activities first evolve. Our investigation discerned established resource integration activities, but not how individuals and groups initiate these activities. Furthermore, how and when resource integration is discontinued was not part of this study. Scholars examining interactions between consumers undertaking resource integration might focus on consumers' underlying goals and examine whether the level of congruence between individual and collective goals (Epp & Price, 2011) is important. Relatedly, the potential interplay between individual and collective resource integration activities requires further investigation.
Our single consumption context -albeit one accounting for a significant proportion of people's lives -clearly limits this study: further work might beneficially check whether our propositions hold within other contexts. Intuitively, of the propositions in Table 2, all but P1, P7 and P8 may prove relevant in both individual and collective contexts. That is, one might expect that in settings not characterized by the presence of other consumers, operand and operant resources and resource integration activities may nonetheless evolve (P2 and P5), resource integrable quality may prove instrumental in whether value is created or destroyed (P3), and consumers may undertake the activities described in P4. Moreover, instances of collective learning (P6) might facilitate resource integration in relatively individual contexts; for instance, where a consumer develops knowledge of running techniques through training with others, then runs alone. Resource integration studies in alternative collective consumption contexts would also be beneficial. Collective consumption contexts vary along a number of dimensions, as we have discussed in Section 2.1; the prominence of technology in this TV context as a mechanism for service delivery might form a further variable. Exploration is needed into how resource integration varies across these different dimensions.
Some other limitations of our study are evident. First, we studied consumers well after service commencement; studying different relationship stages might shed light on how resource integration evolves. Second, by focusing primarily on household members' offline interactions we overlooked the potential for online interactions which may occur where, for instance, an individual is away from home but uses internet services to view a specific show at the same time as other household members. Third, alternative consumer groups may prove different: single-person households, for example, may engage in collective resource integration when viewing with visitors or via social media. Fourth, given the often-negative stereotypes associated with TV consumption, research in this context may be subject to social desirability bias, such that interviewees might attempt to disguise their consumption. This is difficult to rule out, and one can only attempt to be aware of this possibility when interpreting the interview data. Interviews held within the household, which is essentially a collective of familiars, may reduce this danger as consumers are less able to disguise their patterns of consumption from the researcher or each other.

Declaration of Competing Interest
None.