Green entrepreneurship in the sharing economy: utilising multiplicity of institutional logics

The green potential of the sharing economy to exploit underutilised or redundant resources has generated a considerable interest and expectations on the part of government institutions, investors and consumers. Alongside the emerging green logic, more established economic and social logics appear to be critical for growth of sharing platforms. Applying an institutional logics approach, this paper investigates how entrepreneurial teams in the sharing economy deal with this complexity of expectations of various constituents and institutions. Based on 30 semi-structured interviews with founders and executives of UK sharing platforms, we examine the strategies used by entrepreneurial teams to utilise and combine the green logic with other institutional logics present. The results demonstrate that sharing platforms are able to grow via utilising the green logic together with the economic and social logics in a flexible manner, applying complexity reducing and complexity absorbing strategies as well as temporal adjustments in the use of logics.

, it may also give rise to more sustainable business and consumption models based on engagement with underutilised and redundant resources (Cohen and Kietzmann 2014;Demailly and Novel 2014;Hamari et al. 2015).
Although the actual green credentials of the sharing economy are subject to an increasing debate, the green dimension emerges as a distinct institutional logic which can be enacted by sharing platforms in order to attract environmentally conscious platform users and to achieve legitimacy in the eyes of sustainabilityoriented local authorities and investors (see, for instance, sustainability reports by Cleantech (2014) and Aecom (2015) prepared for Airbnb). The emerging green logic is likely to be an addition to a set of more established institutional logics (e.g. economic and social logics) which should be taken into consideration by sharing platforms to achieve their growth objectives (Hamari et al. 2015). This points out to the condition of institutional complexity facing sharing platforms, where different logics may coexist, compete or blend (Greenwood et al. 2011;Battilana and Dorado 2010;Binder 2007;McPherson and Sauder 2013;Zilber 2002). The literature, however, provides no or little insights into how these complexities are dealt with by sharing platforms and their entrepreneurial teams to succeed and grow.
Hence, the aim of this paper is to examine the challenges of handling the diversity of expectations and demands facing sharing platforms, with particular focus on strategies used by entrepreneurial teams to utilise and combine the logic of green with other salient logics. Based on the insights from the institutional logics literature (Thornton et al. 2012;Smets et al. 2015;Besharov and Smith 2014;etc) and the materials of 30 semistructured interviews with the founders and top executives of UK sharing platforms, the paper explores the origins of multiple logics configurations in the sharing economy, their dynamic interrelationships, and practices pursued by entrepreneurial teams to accommodate the green logic in the context of the institutional logics complexity.
The remainder of the paper is organised as follows. Section 2 develops a conceptual discussion on the green aspects of the sharing economy, the green logic in the context of the logics multiplicity and the potential strategies which can be deployed by entrepreneurial teams to navigate across institutional logics complexities. Section 3 details the methodology. Section 4 presents the analysis of the empirical research, including the typology of different configurations to utilise the green logic in the sharing economy, and the respective strategies shaping these configurations, with Sect. 5 providing further discussion before concluding.
2 The sharing economy and the green logic 2.1 Green promises of the sharing economy The 'sharing economy' is a contested term, and there is an ongoing debate about which mechanisms and models should be covered by it. We apply an inclusive perspective on the sharing economy using sharing economy as an umbrella term for various types of sharing models. In a broad understanding, the sharing economy is about business models aiming to provide convenient and costeffective access to underutilised or redundant resources facilitated by digital platforms (Bardhi and Eckhardt 2012;Belk 2014;Eckhardt and Bardhi 2015;Matzler et al. 2014). The focus on underutilised and redundant resources represents the basis for the green discourse around the sharing economy. It has been often argued that the sharing economy holds the potential to facilitate restructuring to more sustainable business and consumption patterns (e.g. Cohen and Kietzmann 2014;Hamari et al. 2015).
Although the actual green performance of sharing platforms has to be critically reflected upon and can often be questioned, three types of potential environmental benefits can be identified (Demailly and Novel 2014). First, by enabling the use of underutilised and redundant resources, the sharing economy can lead to more efficient utilisation of resources and reduced need for buying more physical goods. Consequently, this can reduce physical production, the respective use of natural resources and negative impacts on the environment. For instance, with most privately owned cars or household items being unused for most of the time, there is a considerable potential to share those resources with peers. Car sharing clubs can reduce the number of privately owned cars (Shaheen and Cohen 2013). Another illustration comes from the statistic indicating that one third of household waste represents shareable goods (Demailly and Novel 2014). Second, the sharing economy could potentially lead to reduction of transportation of physical goods as it may facilitate shorter-distance interactions within cities and regions (Cohen and Kietzmann 2014). Third, the sharing economy may potentially increase cultural awareness about environmental dimensions of consumption and therefore facilitate more sustainable consumption patterns. However, there is considerable empirical uncertainty as to what extent those widely held 'green' promises of the sharing economy are really the case (Demailly and Novel 2014;Schor 2014a, b).
Given the diversity of the sharing economy, it is critical to analytically distinguish between different types of sharing as their operation and environmental impacts may vary (Fig. 1). Based on the ongoing discussion of sharing economy models (e.g. Frenken et al. 2015;Belk 2014;Bardhi and Eckhardt 2012;Mont 2002;Schor 2015;Frei 2005;Nesta and TSR 2015;Dervojeda et al. 2013;PWC 2015;Matzler et al. 2014;Grinevich et al. 2015), we separate between the following types of sharing: & 'Pure' sharing economy as peer-to-peer (P2P) platforms facilitating temporary, and possibly paid for, access to underutilised physical assets only (Frenken et al. 2015;Grinevich et al. 2015) & Second-hand economy as P2P marketplaces facilitating the transfer of ownership (rather than accessbased exchanges) of underutilised or redundant assets, including resale of second hand goods, gifting, swapping and bartering activities & On-demand economy including marketplaces facilitating access to intangible assets of individuals such as time and different types of skills (manual, specialised, professional); these may represent a combination of P2P and consumer-to-business (C2B) models & Product service systems (PSS) which are businessto-consumer (B2C) models where a platform owns and rents out assets, which may be considered by consumers too expensive or impractical to own on a long-term basis, with consumers normally performing certain asset maintenance functions when renting and handing over the assets to other consumers & Business-to-business (B2B) sharing, renting, lending and re-selling of assets, which can be based on any of the above four types Each type of sharing is characterised by a substantive degree of empirical uncertainty related to its ecological performance (Demailly and Novel 2014;OCU 2016). In the case of the pure sharing economy, there is a potential that convenience may trigger higher numbers of use, when, for instance, more cars or household items are available in the neighbourhood. This can lead to higher rates of consumption or, in the case of transport, may potentially reduce rates of cycling, walking and use of public transport. Furthermore, lower costs of the offerings may make it possible to consume goods and services, which were previously unaffordable. For instance, people may now be able to travel to expensive destinations because of the sharing economy, contributing to an increase of CO2 emissions. According to this well-known rebound effect (Greening et al. 2000), innovations which increase energy or resource efficiency lead to a reduced per-unit price, which subsequently can lead to an increase in the number of units consumed, potentially offsetting the efficiency gains.
Similarly, PSS may, for instance, enable the availability of cars, which can potentially lead to an increase in car driving and congestion. Again, the on-demand economy can undermine the very idea of underutilised or unused resources when, for instance, people use their car for taxi services on request instead of offering an unused seat or their unused car for others. Whilst the second-hand economy involves considerable potential for increasing the use of products to minimise waste, such redistribution markets may enable access to many Fig. 1 Overview of sharing activities types. Source: Adapted from Frenken et al. (2015) more items which may facilitate hyper-consumption, for instance in the clothing. Furthermore, users may sell goods to buy new goods, and the product life may be shortened through frequent use of goods with low durability. Finally, there can be global rebound effects when individuals save money via using sharing services instead of ownership in several domains, which can mean that those individuals have more money to spend elsewhere with unknown effects on the environment.
Despite the above-mentioned academic debate and empirical uncertainties regarding the actual ecological sustainability of the sharing economy (Demailly and Novel 2014;OCU 2016), the public, business and media discourse tends to strongly link the sharing economy with a more sustainable way of organising a market place (Botsman and Rogers 2010;Rinne 2016). This can potentially make the green logic one of the key strategies for entrepreneurship in the sharing economy to achieve legitimacy in the eyes of government, legislators, customers and investors. As indicated in the sustainable entrepreneurship literature (Schaltegger and Wagner 2011;Gibbs and O'Neill 2012), one can distinguish between different forms of sustainability management and entrepreneurship, depending on the centrality of sustainability issues to a given business model. In this context, we are not examining the ecological or sustainability performance of sharing platforms themselves but are investigating the way in which they utilise green agendas, pressures and expectations originating from different institutions, alongside other institutional logics, to achieve business success.
2.2 Institutional complexity and the green logic in the sharing economy Most sharing economy organisations are operating as two-sided platforms bringing together and matchmaking providers (suppliers; sellers) of goods and services with consumers (buyers). In order to succeed, sharing economy platforms have to engage in 'several games' at the same time to accommodate stakeholders at both the demand and supply side. As such, they are facing a substantive institutional complexity (Kraatz and Block 2008;Smets et al. 2015), which is further exacerbated by uncertain regulatory environment and fragmented market infrastructure support (such as access to capital, insurance services, etc.) provided for sharing economy organisations (Grinevich et al. 2015). This may lead to extra challenges for them in terms of defining strategies to deal with different external constituents (e.g. investors, regulators, legislators, etc.).
A more detailed look into the platform-centric organisation of sharing economy businesses helps further explain why performing a seemingly straightforward economic function of an intermediary can be associated with a substantive institutional complexity. In the case of a two-sided platform (Fig. 2), its value depends not only on the total number of platform users, but also on the extent to which the number of users on the supply side is balanced with that on the demand side. Twosided platforms (e.g. Airbnb, Peerby, Blablacar, Uber among others) provide the benefit of bringing together and matching two broad user groups, in particular, service providers (individuals who offer their underutilised resources) and service consumers (individuals who would like to use these underutilised resources). This type of platforms may suffer because of an unbalanced increase in the total number of users (Salminen 2014;Shih 2011). For instance, an increase in the number of service providers (e.g. Uber drivers) may discourage other potential service providers to join the platform if there are very few consumers (e.g. Uber passengers) on the platform. Conversely, an increase in the number of consumers (e.g. Airbnb travellers) on the respective platform, which is not accompanied by the matching increase in the number of suppliers (e.g. Airbnb hosts), can reduce the attractiveness of the platform. Because of this vulnerability of two-sided platforms, they can be extremely sensitive to the issues of understanding and dealing with different institutional logics which make suppliers and consumers join the platform. The strategic choices made by a given platform to accommodate the logics which are perceived critical for its survival affect 862 V. Grinevich et al.

Fig. 2
Stakeholders in a two-sided sharing economy organisation specification of rights and activities allowed for different platform users, user-specific pricing and remuneration structure and levels, and the order of dealing with and getting on the platform different user groups (Shih 2011). These can also have implications when seeking support from external stakeholders.
There is a small but growing literature indicating that sharing economy platforms have to manipulate with three key logics in order to attract platform users and enjoy support from external stakeholders. First, it is an economic logic of convenience and cost, referring to the advantages that sharing economy platform users can enjoy compared to mainstream offerings. Individuals can get access to goods or services they otherwise may not be able to afford, or they can get this access in a more convenient fashion (e.g. on demand, when they need it and where they need it) (Hamari et al. 2015). Second, it is a social logic. Sharing platforms can engage with it by facilitating sharing among strangers (Schor 2014a(Schor , b, 2015. This may be appealing to those potential platform users, who enjoy new socialising experiences. Platforms can also facilitate local community building, which could be especially attractive to socially conscious platform users and other stakeholders. Finally, it is a green logic. As discussed in Sect. 2.1, it is widely held that sharing economy platforms can potentially improve ecological sustainability, which can be appealing to certain categories of potential platform users, regulators and environmentally conscious investors (Cohen and Kietzmann 2014;Hamari et al. 2015;Demailly and Novel 2014;Shaheen and Cohen 2013).
These three logics appear to be critical for sharing economy platforms regardless of whether they are purely for-profit or socially oriented. Growth of users is a key success factor for all platform-based business models, and the above three logics are important drivers for growth of platform users. We, therefore, abstract from the question of whether sharing economy platforms deal with for-profit versus not-for-profit logics but focus on the logics that drive platform growth.
Hence, we conceptualise the green logic as part of a complex institutional environment, facing a sharing platform, alongside the social and economic logics. These logics are reproduced in the motivations and demands of the key constituents of the sharing economy such as providers/sellers and consumers/buyers. Due to the platform-centric design of sharing economy organisations, these logics have to be addressed by their entrepreneurial and management teams in order for the platforms to succeed and grow. This should also be done in a way which is acceptable for external stakeholders.
Before we proceed to discuss in a more detail the notion of the multiple logics complexity and potential strategies which may be deployed by sharing platforms to deal with such complexity, it is worth noting that in the case of product service type of sharing (see Fig. 1), we deal with one-sided platforms, where the supply side is fully controlled by the platform. This may reduce the complexity of decision-making on the one side of the marketplace but does not affect the relevance of the key logics above as they remain important for the rest of the constituency groups involved (buyers and external stakeholders).

Institutional logics and complexities
Institutional logics are defined as 'symbolic systems, way of ordering reality, and thereby rendering experience of time and space meaningful' (Friedland and Alford 1991: 243). Putting forward the value of material and symbolic aspects of institutions, an institutional logics perspective helps us to recognise how material practices and symbolic interactions influence organisational practices (Thornton et al. 2012). Logics influence organisations both as prescriptions and demands, with most organisations facing institutional complexity as a result of their interactions with a range of stakeholders spanning multiple domains (Schildt and Perkmann 2016).
Co-existence of institutional logics, rather than dominance of particular logic within the field, has been documented well by institutional scholars (e.g. Dunn and Jones 2010;Jarzabkowski et al. 2013;Reay and Hinings 2009;Schneiberg and Clemens 2006). Greenwood et al. (2010) initially used the 'institutional complexities' term when articulating the incompatible demands of different stakeholders. Recent studies recognise that multiplicity of logics within the field does not always cause institutional tensions and a challenge for the organisations, yet they may shape organisational structures and practices in constellation (e.g. Goodrick and Reay 2011;Waldorf et al. 2013;Schildt and Perkmann 2016;Pallas et al. 2016). This, for instance, could take the form of conforming to dominant logics or dissenting, by resisting identification with a single institutional logic (Mair et al. 2015).
Aligned with the evidence on co-existence and constellation of institutional logics, the current literature debate has shifted towards developing a better understanding of how organisations experience and navigate within the complex institutional environments (e.g. Raynard 2016, Besharov and Smith 2014). In this context, we specifically investigate institutional complexities facing sharing platforms by looking at the interrelationships among institutional logics pursued by their key constituency groups and responses of entrepreneurial teams to accommodate them in order achieve platform growth. We now proceed to present a literature discussion on organisational responses to institutional complexities, with implications for theorising the strategies to utilise the green logic in the context of multilogic complexity.

Theorising strategies to deal with institutional complexity
We conceptualise the green logic as an emerging institutional logic which further adds to the complexity of institutional environment for the sharing economy, alongside more established economic and social logics, with their respective prescriptions and demands (i.e. values, expectations, identities). Thus, important challenge for the organisations can be to navigate through these complexities in order to survive and succeed. The literature indicates that organisations may choose 'complexity reducing' or 'complexity absorbing' strategies based on the organisational experience of these complexities (Jay 2013;Raynard 2016). Organisational characteristics as well as social interactions and sense making process within the organisations can be relevant too.
Complexity reducing strategies aim to 'resolve complexity through separating alternative logics and the practices of people that enact them' (Smets et al. 2015: 934). For example, allocation of a new organisational unit pertinent to meeting demands of new institutional logic, referred as 'compartmentalisation' or 'structural differentiation', is a complexity resolving strategy (Boxenbaum and Jonsson 2008;Tushman and O'Reilly 1996;Perkmann and Schildt 2015). In addition to compartmentalisation as an organisation-level differentiation strategy, there can also be individual-level strategies. Within this context, the 'segmenting' term has been introduced and defined as 'practices that use given organisational structures to allow individuals to enact coexisting logics separately, where and when appropriate, to protect them from scrutiny by, and loss of legitimacy with, referent audiences of competing logics' (Smets et al. 2015, p. 960).
By contrast, complexity absorbing strategies may be chosen by organisations with the aim of benefiting from a 'broader practice repertories, access to additional resource pools and enhanced innovation…' (Smets et al. 2015: 934). Hybridisation of practices and structures through integrating institutional logics from different domains is well-demonstrated as a viable strategy for organisational accomplishments (Scott 2001;Battilana and Dorado 2010;Bjerregaard and Jonasson 2013;Jay 2013;Pache and Santos 2013;Tracey et al. 2011;McPherson and Sauder 2013). By strategically employing practices drawn from alternative logics, organisations can gain legitimacy and acceptance (Pache and Santos 2013). Known as 'selective coupling', this strategy can allow hybrid organisations (e.g. social enterprises) to selectively couple and implement genuine practices from different logics and therefore achieve legitimacy from stakeholders without faking compliance. Related to this, Smets et al. (2015) introduced the term 'bridging' to refer to 'the situated and judicious combination of practices governed by competing logics in order to reap their complementarities' (p. 961).
In the case of the sharing economy, it is an open question how the platforms can deal with the logic multiplicity involving an emerging green logic and more established and salient economic and social logics. It remains unclear to what extent and how the platforms can utilise these logics via complexity reducing or complexity absorbing strategies in one business model in order to grow successfully. We aim to explore these issues by examining the manifestation and interrelationship of the green logic vis-a-vis other salient logics and how decision makers of sharing platforms develop viable strategies.

Methodology
This research focuses on sharing economy platforms which are operating, or have operated, in the UK. These platforms include UK home-grown companies and overseas-based platforms which established their operations in the UK or attempted to expand in the UK but later had to withdraw. Based on our conceptual discussion in Sect. 2.1, to capture the wide public understanding of the phenomenon, our sample centres on sharing economy platforms in a broader sense, including 'pure sharing' platforms, second-hand markets, on-demand services, product service platforms and B2B sharing. The sample covers accommodation, transportation and personal/professional services. These three sectors are widely regarded as the main fast-growing segments of the sharing economy (EPRS 2016;PWC 2015). The target population of sharing platforms was identified using publicly available directories, such as meshing.it, collaborativeconsumption.com and compareandshare. com, and was manually verified by analysing the content of web-sites of the respective companies. The target list includes 75 sharing economy platforms. Of these, 30 agreed to take part in research interviews scheduled and conducted in 2015. Table 1 outlines the distribution of the target and interviewed platforms by sector and sharing economy type.
The sharing platforms analysed and interviewed vary in terms of their age and scale of their popularity among users (Table 2). This allowed us to identify key dynamic mechanisms and processes behind the multiplicity of institutional logics facing sharing platforms and uncover relevant events, actions and characteristics of stakeholders, including enabling and constraining contexts in relation to the green logic.
The research is based on (i) semi-structured interviews with key decision makers (founders or top executives) of sharing platforms and on (ii) an analysis of the online presence of the platforms. This form of interviews with key decision makers is also called 'key informant interviews', which is a well-established approach (Campbell 1955;Kumar et al. 1993;Squire et al. 2009) in qualitative and mixed-method research. In the context of our study, founders, chief operating officers and other top executives represent the most knowledgeable individuals in relation to prescriptive and demanding elements of institutional logics governing sharing economy platforms. The composition of interviewees by their role in ownership and management of respective platforms was as follows: 19 founders and chief executive officers (CEOs), 6 CEOs, 3 other senior executives (such as Chief Technology Officer, Chief Operations Officer) and 2 co-founders.
Before commencing the interview programme, the study obtained all necessary permissions from the local ethical review committee. The interviewees were guaranteed full anonymity and confidentiality of conversations, unless advised otherwise by the interviewees themselves (Creswell 2007). Most interviews were conducted face-to-face or via video link, with some interviews done via telephone. On average, the conversations lasted 45 min. Each interview was conducted by one of the two interviewers, who jointly designed the semistructured interview guidelines. The latter were centred on questions about key challenges experienced by platforms when it comes to successful growth (e.g. an increase in numbers of users and expansion in new locations). The questions sought the decision makers' perspectives on the factors behind the evolution of their business model, socio-technical requirements to operate two-sided platforms and the institutional demands and expectations of stakeholders (see Electronic supplementary material A.1). It was an explicit interview tactic to explore the perceived prerequisites for growth in terms of engaging with, and addressing the interests of, stakeholders and institutions. The interviewees had to be attentive to green impression management by asking critical follow-up questions to specify sustainability On-demand 13 6 Product service 8 4 Second-hand a 4 3 Business sharing a 5 3 a Second-hand economy and business sharing platforms tend to be hybrid, by integrating other types of sharing activities (such as pure sharing and/or on-demand, and/or product service) arguments put forward. Potential uncertainties regarding biased memorising or strategic impression management were addressed via credibility probes during the interviews and critical interpretation of the transcriptions (Patton 2002;Rubin and Rubin 2012).
To facilitate qualitative data analysis and cross-check of its validity and reliability, the interviews were fully transcribed verbatim (Miles et al. 2014). The transcripts were analysed in a multi-step iterative process, involving both interviewers and co-authors who were not engaged in conducting interviews. The coding process operated as an interplay between theoretical preconceptions influencing the analysis and inductive reasoning influencing conceptual development (Markusen 2003;Miles et al. 2014). To optimise validity via critical verification techniques (Morse et al. 2008), the coders crosschecked and enriched each other's interpretation of the data. In the first round (see Fig. 3), the interview data were coded according to the type of institutional logics (economic, social and green) as discussed in Sect. 2 and related constituency groups. When guided by these prespecified codes, the coders remained open to other analytic categories originating from the data which could indicate the presence of any other logics. The coders were satisfied that there were no indicators found which could fall outside the initial coding scheme based on the constructs of the economic, social and green logics.
This first-order coding was complemented by an examination of the online presence and design of the platforms studied to get further understanding of their value proposition and functioning. This supported the iterative process of analysing the interviews and enabled an understanding of logics that are utilised and communicated to potential stakeholders. We found that each of the platforms interviewed was addressing the demands of the economic and social logics. The reference to the green logic was observed in responses and web materials of 26 platforms. The remaining four platforms, which represented on-demand personal and professional services, did not explicitly communicate the green logic themselves but featured in media coverage related to 'sustainable' sharing economy generated by third parties.
The second round of coding involved understanding and interpreting the extent of the prevalence of the green logic and its relationships and interconnections with the economic and social logics. Driven by the data, this theoretical round of coding was also substantively informed by our conceptual discussion in Sect. 2 on multiplicity of institutional logics and organisational responses to it. We identified distinct configurations of multiple logics deployed by the platforms. These became valuable and nuanced theoretical constructs, especially in relation to developing our understanding of the way in which the sharing platforms can reconfigure their institutional logics to suit their constituent groups. This round of coding, however, was primarily concerned with interpreting existing (static) configurations of logics as they were shaped and deployed by platforms at the time of interviews.
Finally, in the third round of coding, we concentrated on understanding the interview data on dynamic shifts to and from the green logic across different logic configurations. The fluidity of the green logic and its relevance for platform growth emerged as one of the central aspects during the second round of data analysis but required a higher level of abstraction to conceptualise it. Hence, a separate coding round was needed. Similar to the process in the first round of coding, the higherlevel rounds of data coding were enriched and contextualised through analysis of web-presence of platforms, as well as data on platform age and web user statistics. 1

Findings and discussion
We now proceed to describe different patterns of interconnection between the green logic and other institutional logics as observed in our sample of sharing economy organisations, and how these logics are strategically orchestrated by their key decision makers such as founders and CEOs. Importantly, we are not examining the actual ecological performance of sharing platforms, but are focussing on how sharing platforms utilise green expectations and pressures from respective institutions (e.g. local authorities, societal norms, perceived beliefs of platform users, etc.), potentially in combination with alternative logics, in order to achieve growth of platform usership. 2

Logic multiplicity configurations
Keeping with our conceptual discussion in Sect. 2, we find that the degree to which the green logic influences strategic decision making and management of platforms can vary substantively. Our analysis of interview data, combined with web-materials and media coverage of the related platforms, reveals several distinct configurations of the relative importance of the green logic. The first configuration is represented by platforms where the green logic and its importance can only be implicitly traced in conversations with their founders and top executives (Table 3). This is indicated through their general recognition of the positive connotation of the notion of the sharing economy, and a 'very warm and friendly feeling' generated by it. As indicated in our early discussion, the sustainability aspect represents one important constituent of that feeling. In our sample, these platforms are representing on-demand professional and personal services only. Although it is the economic and social logics which are dominating this type of platforms, they appear to be willing to contribute to and benefit from the debate on sustainable sharing economy by participating in expert panels, round tables and featuring in the media coverage. The second configuration pattern of the logic multiplicity identified is represented by sharing platforms where the green logic is explicitly acknowledged by key decision makers, but allocated a secondary or complementary role. This is primarily because the sharing platforms in this category (which can represent pure sharing, on-demand and second-hand economy) do not normally consider themselves too dependent on sustainability-oriented groups of users. It is economic and social logics which are prioritised and nurtured by 1 For each platform web site, we used http://www.similarweb.com/ to track the average number of visitors per month from May to October 2015 to have an indirect and imperfect, but the best possible, empirical indicator of the success of platform growth. 2 An increase in the number of platform users may undermine the resource savings made through sharing economy model. As illustrated in Sect. 2.1, this ecological reality can diverge from the social reality. It is unclear how the actual ecological performance of the sharing economy with its multi-level uncertainties interacts with the social reality, and the management of platforms. This issue remains outside the scope of this paper, where the green logic is conceptualised as a social reality emerging through societal, norms, regulations, values and beliefs.  'So I think when it comes to sustainability I'm sure there are people who think that, but will they act that way? So there are some people...that will likely use a sustainable service versus a non-sustainable service...And out of those people that identify themselves as people that would choose a sustainable service over a nonsustainable service… how many people would actually use that factor maturely in their decision making process? So, we do it, but ultimately most people [platform users] are triggered by a valid proposition that is either convenience or economics' (Founder and CEO, on-demand personal/professional services, participant ID 26).
'...it's a two-sided marketplace...what people value most is that we connect them to others in the neighbourhood and they're able to do something to do good and feel good about themselves. I think for the borrower, it's the fact that they are able to use something for free, something that... they wouldn't want to keep in their house and maybe you know connect to a neighbour and save the environment at the same time. But those two are I think a bit more of a bonus for the borrower' (Founder and CEO, pure sharing services, participant ID 24).
17 platforms: pure sharing accommodation (6); transportation (3); personal/professional services (2); on-demand transportation \(1); personal/professional services (1); secondhand-based services (3); business sharing services (1) Critical prerequisite to deploy other logics • Legitimacy of platforms in the eyes of local and central government can only be obtained through 'ticking' the green logic box: ...municipalities or cities or… governments are looking for sustainable ways of improving their mobility or finding solutions for their mobility issues...We show cities that we are a good solution for part of their mobility problems...We also know that car sharing in general adds to the sustainable character of cities, in a way that the congestive traffic and pollution can be positively impacted by sharing cars' (CEO, participant ID 8).
'…operating a car sharing business of our sort in London requires very close relationships with the Boroughs, who control the parking. You have to take a fairly patient approach to that, you can't just turn the switch and you have parking available. And perhaps the most crucial thing was to bid aggressively, and win a tender put out by Borough Council' (Founder, participant ID 9).
• But, it is still other logics which need to be respected to increase the number of platform users: 'So it's not just ...the green aspect of our business; it's also our price point and it's also the convenience of our business, and where our vehicles are located. For example we're a 24/7 business so as opposed to the local car rental companies which close at 5pm...So there are several factors to our business, not just being a green company' (Founder and CEO, participant ID 13).
platforms within this configuration to achieve growth. This configuration is found to be most representative among the platforms interviewed. The third configuration pattern of logic complexity can be observed in sharing economy organisations where the green logic represents a critical prerequisite to let other valid logics be deployed. This is most typical for PSS companies in the transportation sector, which provide short-term car rentals. This type of organisations are normally unable to start providing their services in new locations without prior, often formal, agreement with local authorities and other infrastructure regulators, who need to be reassured in green credentials of incoming car-rental providers. Therefore, the green logic becomes central for this type of organisations to accommodate the demands of local gatekeepers controlling access to local markets and resources. This promoting of the green logic to regulators for gaining legitimacy is particularly important in the light of the ongoing regulatory backlash faced by sharing platforms across many cities, regions and countries (Economist 2014). The green logic appears to be less relevant within this configuration when communicating the message to another key constituency group such as individual customers, where the demands of the cost and convenience logics can be critical.
The fourth configuration pattern of logic multiplicity is associated with sharing platforms where the green logic is not only a prerequisite but also an integral part of a sharing business model. In this configuration, the economic value of the sharing platform is derived from its ability to fully exploit and embed the sustainability agenda. Among the examples provided by our sample of the interviewed platforms are those which are primarily aiming at business clients to help them achieve their own sustainability targets in relation to recycling and reducing carbon footprint (e.g. facilitating resale/re-use of redundant physical assets or providing car-pooling services to corporate staff). Interestingly, the green logic can be integrated into a business model to such an extent that it becomes effectively inseparable from the commercial logic.
In these cases, the green logic is usually combined with other logics, in particular with the logic of offering cost-effective or convenient service for users. Yet, within this configuration, there are a few cases where the green logic is prevailing, with other logics playing a secondary or subsidiary role. In these instances, it is the green logic which guides the start-up stage and may continue to exert major influence on all aspects of organisational functioning after that. In an extreme situation (which is also present in our sample), the green logic may even become a single logic behind the platform, entirely overshadowing other institutional demands. Importantly, whilst the dominance of the green logic over any other logic can be viable in the early stage of the start-up process, re-balancing and integration of other logics, in particular cost/convenience appears to be required for longer-term growth and success. Section 4.2 will elaborate further on these temporal shifts.
Overall, our results indicate a substantive degree of heterogeneity in relation to the way in which the green logic is manifested in sharing economy organisations, from being only implicit or of a secondary relevance to playing a critical and dominating role. Given that the sharing economy in its current digitally enabled form is a relatively new phenomenon, and includes a variety of sharing types across different sectors, it may not be too surprising that there is no common interpretation of the extent of the relevance of the green logic. The absence of shared interpretation of the green logic is further reinforced and reproduced by the platform-centric design of sharing organisations, which brings together multiple constituent groups (buyers, sellers, the platform founder team and external stakeholders) which may also disagree in their interpretation of the green and sustainability agenda in the sharing economy.

How strategies are chosen to utilise the green logic?
We find that the green and other relevant logics, such as those related to convenience and cost, and social experience, tend to be compatible with each other, with limited contradictory prescriptions. One explanation for that lies in the logic multiplicity being nearly exclusively demonstrated at the level of the means by which organisational goals are achieved. As indicated in the literature, at this level, the compatibility between multiple logics tends to be higher, compared to that at the level of the goals (Besharov and Smith 2014;Pache and Santos 2010).
For all the sharing platforms interviewed, one overarching goal is to increase the number of platform users. This is regardless of whether a given platform is primarily driven by commercial considerations or aims to make a positive social and/or environmental impact. Thus, the growing number of platform users becomes a homogeneous precondition for achieving success by any sharing platform. The presence of this single and underlying logic of platform scale and growth appears to distinguish platform-centric ventures from other multilogic-based ventures (e.g. social enterprises), where different logics present at the level of organisational goals may not have such a 'common denominator', creating hybrid organisational responses (Besharov and Smith 2014;Purdy and Gray 2009).
In the absence of the logics contestation at the level of organisational goals, the platforms are focussing on experimenting and finding the right strategy for accommodating different institutional logics as means to achieve their overarching goal of increasing the number of users. Again, irrespective of whether they are for-or not-for-profit, the platforms embrace institutional complexity to achieve their growth goals.
As discussed in Sect. 2, the institutional complexity facing a sharing platform can be attributed to its organisational and functional design, where the platform's entrepreneurial team has to deal with a variety of constituency groups critical for its growth. These are different groups of stakeholders such as platform users (sellers and buyers), local authorities, legislators and investors. When engaging with a sharing platform, these multiple key constituents may apply one particular logic (e.g. convenience/cost, social or green logic), or different combinations of them, which creates and reproduces multidimensional institutional prescriptions for the platform's entrepreneurial team. We find that the platforms' experimentation with different logics (as means) includes strategies which are aimed at both reducing and absorbing institutional complexity.

Reducing complexity
As indicated in the literature elsewhere (Dunn and Jones 2010;Simsek 2009;Tracey et al. 2011), one way of juggling different institutional logics is to match them with the relevant referent audiences, where and when appropriate. This, for instance, can be achieved via organisational compartmentalisation or structural differentiation. We find, however, very little evidence for this type of strategies, which may be due to the majority of sharing platforms being small and relatively simple in terms of organisational structure. In particular, we find no evidence for the green logic being implemented in a dedicated organisational subunit.
Instead, there is evidence for individual segmentation practices similar to those observed and theorised by Smets et al. (2015). In cases where the platform's decision-makers are confident that a particular key actor is interested in one particular logic, they select the respective cognitive frame of this logic and focus on this logic in their communication and engagement with this actor. For instance, deploying the green logic discourse (e.g. reducing the carbon footprint, bringing down congestion, improving air quality) becomes especially useful for engaging with public sector organisations that provide financial support or control access to public infrastructure. It is also relevant for potential business customers, charities and other organisations (e.g. big event organisers) which for different regulative and normative reasons are keen to demonstrate their positive impact on the environment.
The green discourse initiated by the sharing platforms often involves developing conceptual models or data analysis to estimate the environmental impact reductions generated by the sharing economy platforms. These models and calculations are usually selective representations and simplifications (e.g. number of owned cars which could be replaced by one shared car, implying 'definite' reductions in CO2 emissions), and normally not based on holistic analysis of complexity of multiple ecological dimensions.

Absorbing complexity
The ability to implement the segmentation strategy normally depends on a very good general and practical understanding of distinct and salient demands of key constituency groups (Schatzki 2006;Smets et al. 2015), which may not always be present in the case of the emerging sharing economy. Where there is a substantive degree of uncertainty about which institutional logic a particular key actor is following, the absorbing complexity strategy appears to be the most common choice for sharing platforms.
When a particular platform user visits the website or app, it may be difficult for the platform and its entrepreneurial team to determine whether the user is interested in the green logic, the cost/convenience logic or the social logic. It can be that certain users are attracted by the combination of logics, for instance by enjoying cost/ convenience benefits coupled with the fascination of the green benefits. In these situations, segmentation strategies do not work, and the platforms' decision makers have to embrace a different approach, resembling the bridging mechanism documented elsewhere in the literature (Purdy and Gray 2009;Smets et al. 2015) and discussed in Sect. 2.4. Essentially, what entrepreneurial and founder teams try to do (when facing uncertainty about the norms and values which guide their constituency groups in relation to the sharing economy) is to communicate the diversity of logics pursued by the platform. This is achieved, for instance, through presenting their business model as not only being green but also desirable by consumers who value convenience and low prices and attractive to individuals (potential providers/sellers) who want to make extra money. We found that platforms often use the construct of 'sharing' in order to bridge economic and green logics when articulating their value proposition and making an argument that sharing is the right way to both achieve economic efficiencies (e. g. saving time and money) and 'husband the world resources', with no need to manufacture new things. Hence, the construct of sharing (which can be, to a certain extent, interpreted as a proxy for the social logic) appears to be activating complementarities between all three logics present.
Bridging mechanisms (such as utilisation of all three logics or different combinations of them) tend to be shaped by judgement of entrepreneurial and founder teams about perceived relative salience of different logics for platform growth. It appears to be a rational approach of platforms to rely on multiple logics, either to deal with the uncertainty about the dominant logic followed by potential platform users or to address the interest of certain users in multiple logics. Whilst enabling and exploiting complementarities of these different logics appears to be critical for reinforcing growth of sharing platforms, we also find that the bridging processes, reconnecting segmented aspects (Smets et al. 2015) of the logics, may demonstrate slippage towards one relatively dominant logic, which is, in most cases, the economic logic of cost and convenience, with the green logic being at risk to be demoted and allocated a secondary role. This indicates temporal fluidity of the green logic, which represents one of our most intriguing findings as amplified in the ensuing section.

Temporal shifts to and from green
Given the uncertainty about the relative importance of the green logic as one of the means to achieve platform growth, the platforms appear to be under pressure to be flexible and dynamic in response to the fluctuating signals they may receive from their key constituency groups. We find that the ability to recalibrate logics configurations in response to different situations and at different stages of the platform lifecycle appears to be critical for the platforms at both sides of the logics configuration spectrum depicted in Table 3.
Our analysis indicates that for the platforms which are motivated and guided by the green logic in the first instance, the green agenda is very important for (i) motivating the entrepreneurial team, (ii) interacting with local authorities and (iii) attracting a core group of intrinsically green users at the very early stage of the platform development. These early adopters enable experiential learning processes as well as first income streams. In this case, however, the green angle as such can hardly ensure a long-term increase in platform users, pressurising green-motivated entrepreneurial teams to a certain extent concede the green logic at later stages of platform development. In their words, the reality becomes more about finding the factor which makes their business always work, and this is primarily about being able to compete on convenience and cost.
Making a transition from a model primarily dominated by the green logic towards that based on mixed logics inevitably involves trade-offs between the original vision for the company and its ongoing operational business needs. These trade-offs appear to be achieved relatively painless from the organisational point of view for all companies in our sample where the green logic was at the heart of their business model from the very beginning. Inability to initiate that transition, however, may be lethal. Indeed, there is one case in our sample of platforms, where the focus on the green logic only and the lack of integration of alternative logics appeared to be a key reason for the platform to run out of business during our study, with one more company with a similar approach seriously struggling to increase the number of users, despite being successful at the early start-up stage.
Related to the above is the question whether growth of the platform may be in conflict with the green logic, and whether the platforms should still accommodate it in order to achieve growth. The e v i d e n c e f r o m o u r s a m p l e c o n f i r m s t h a t sustainability-related regulations, norms and values constitute an increasingly important consideration when it comes to platform development and success. This is reflected in the business model evolution of the platforms which started from the premise of convenience and cost, and initially only marginally benefited from the green sharing economy discourse before they realised (with some variance across sectors and sharing types) they would need to do more at later stages of platform development to stress the green aspects of their offerings and business models. This realisation often comes through collecting their own evidence and data analytics on the factors important for maintaining the attractiveness of the platform among users, as well as interactions with regulators and other businesses. This is especially the case for all sharing types in the transportation sector, and to a lesser extent for accommodation sharing.
The green agenda appears to be least relevant for sharing models offering on-demand intangible assetbased services.
It is again worth stressing that in this context, the green logic and organisational responses to it are shaped by the social reality and pressures constructed by societal rules, norms and beliefs. This social reality can diverge from the actual ecological reality where an increase in platform users may overcome the ecological savings achieved through shared access to goods and services. 3 This actual or potential divergence does not appear to hurt the sharing platforms from the business point of view, given that it is the social reality which is providing guidance for platform management decisions and their success.
Overall, our findings point out to the temporal fluidity of the green logic, resulting in the following propositions. For deliberately green start-up platforms, the green logic tends to remain but diminish in its importance, giving the way to the cost/convenience logic over time. Conversely, for sharing start-ups that initially focused on the economic logic, integrating the green logic, at least to some degree, tends to become more important over time but may be influenced by the type of sharing and sector involved. In both cases, it is the construct of sharing (i.e. the social logic) which appears to be instrumental for bridging the green and economic logics.

Conclusions
Sharing economy research has primarily been driven by fragmented empirical pieces and media stories, with very little theorising to understand the nature of the phenomenon and complexities (Demailly and Novel 2014). This paper represents one of the first studies on the sharing economy taking advantage of robust theoretical foundations of the institutional logics approach. More generally, there is a substantive gap in knowledge on how local actors experience and manage complex terrain of multiple institutional logics and fields in business, as noted by many scholars (e.g. Greenwood et al. 2011;Kraatz and Block 2008;McPherson and Sauder 2013). Our study of how sharing economy venture founders, as green entrepreneurs, manage their enterprises at the intersection of multiple institutional logics contributes towards addressing this gap. We provide novel empirical insights into how complexity that arises from multiple logics is handled in the practice of running a sharing economy platform and how the coexistence of different logics is sustained over time by retaining the green logic on the radar. In particular, we illustrate how, regardless of the actual ecological performance of sharing platforms (which is outside the scope of this paper), sharing platforms consider and utilise the green logic in association with other institutional logics as a flexible social tool for engaging with key stakeholders. Our findings further corroborate the argument about multiple logic complementarity, and how the coexistence of multiple logics can be beneficial for organisations (see Mair and Hehenberger 2014;Mair et al. 2015). We find that the ability and willingness of entrepreneurial teams to draw on several institutional logics with a substantive degree of flexibility and agility contributes to the business sustainability and growth of sharing economy ventures.
Specifically, we make several distinct contributions to knowledge. Firstly, we extend the current literature on institutional logics by providing an empirical illustration for multiple logic configurations in the sharing economy field, going beyond such approaches that focus on dualistic nature of conflicting logics across different institutional contexts that prevail in the extant literature (e.g. Reay and Hinings 2009;Pache and Santos 2010;Ocasio and Radoynovska 2016). We contribute to the institutional scholarship longstanding interest in the interaction between agentic use of logics as tools and the consequences of these uses on strategy practice. In doing so, we enrich the relevant institutional logics literature by illustrating how different institutional logics can be employed over time and in a constructive manner. In contrast to much of the literature on institutional logics, where tensions and contradictions are emphasised, in the case of platform-centric business models of the sharing economy, it appears to be a key success factor for growth to combine different logics in a productive way. We have identified and presented four distinct configurations of engaging with the green logic such as 'implicit', 'explicit but secondary', 'critical prerequisite for deploying other logics' and 'core with other logics being combined or secondary'.
Secondly, our theoretical contribution extends to the domain of green entrepreneurship, by providing a foundation for a more extensive examination of the green logic engagement in entrepreneurship strategies. We have not found evidence for organisation-level complexity reducing strategies such as organisational compartmentalisation or structural differentiation (Boxenbaum and Jonsson 2008;Tushman and O'Reilly 1996;Perkmann and Schildt 2015). However, there is evidence for individual segmentation practices similar to those observed and theorised by Smets et al. (2015), whereby the cognitive framing of the green logic is communicated to specific stakeholders who are interested in the green credentials for normative reasons. Yet, as sharing platforms often appear to face situations of uncertainty about which institutional logic is of interest to their constituency groups, they frequently apply complexity absorbing strategies, explicitly communicating multiple logics in parallel. The vague term of sharing appears to facilitate this bridging mechanism (Purdy and Gray 2009;Smets et al. 2015), whereby stakeholders can 'pick and choose' logics that are of interest to them.
Thirdly, we enrich the existing literature by further bridging the institutional theory discourse and green entrepreneurship scholarship. In this context, one of our central findings is related to the temporal dynamics of the green institutional logic and its relevance for startup process of green ventures. Based on it, we have developed two propositions: (1) for deliberately green sharing economy start-ups, the green logic tends to become less important, with the economic logic increasing its importance over time to increase attractiveness for a broader range of users.
(2) For sharing economy start-ups initially driven by the economic logic, integrating the green logic may become more important over time to enhance legitimacy in the eyes of regulators, platform users and potential environmentally conscious organisations. As the green logic appears to be pertinent and constituted within the sharing economy-based entrepreneurship, the issues of temporal fluidity and complementarity of institutional logics warrant further attention for theorising in this domain.
Finally, we aim to reinforce efforts for meaningful scholarship with normative implications for decision makers and offer potential strategies, which can be deployed by entrepreneurial teams to navigate through institutional logics complexities. This type of institutional fields (such as sharing economy), which is characterised (i) by empirical uncertainty about which logic or combination of logics respective users are interested in or (ii) by a greater frequency of temporal changes in the needs of key constituency groups, requires agility in strategy formulation and emerging business model development. Temporal shifts to and from the green logic can be an integral part of business development. Careful examination of the situations, contexts and actions (of users and other stakeholders) and choosing an appropriate strategic mechanism of complexity reduction or absorption will drive platform growth while also maintaining green and social integrity of the venture.
In relation to the future research agenda, our findings indicate that in-depth examination of green aspects of venture development in the field of the sharing economy can substantively enrich investigations of the interaction between alternative logics, contexts and actors as currently pursued in the literature. Further research is needed to investigate the material ecological effects of various sharing models in a holistic fashion to provide more substance and critical reflections for the green discourse of the sharing economy. Additional insights into potential context-specificity can be brought by comparative studies of the use of logics in sharing economy ventures by entrepreneurial teams in different national, socioeconomic and sectoral settings.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.