Lessons on business model scalability for circular economy in the fashion retail value chain: Towards a conceptual model

Circular economy and especially circular business model (CBM), is currently being discussed as a way to enable the fashion industry’s transition to sustainable business models wherein pollution and resource waste may be reduced. However, one of the prime reasons for a slow transition is lack of scalability of CBMs operating in the fashion retail value chain. What is lacking in the current discourse is research that summarises and condenses the literature on strategies for how scalability can be attained and what that means in context to CBMs where not only economic values are in focus. Therefore, the main purpose of this paper is to explore the main strategic approaches to scale business models and how these can be applied to CBMs in the fashion retail value chain. To do this, a two-part method is adopted consisting of a systematic literature review of 57 business models and scalability papers followed by a review of activities reported by 76 fashion retail companies on how these have, or are planning to, increase the scale of their CBM initiative. Our suggested model provides a basic understanding of strategies for business model scalability seen from four different business model design perspectives. These are further contex-tualised for CBMs in the fashion value chain and lessons learned are generated in the form of four central propositions. The propositions account on how organisations can leverage resources from their existing conventional business model for eﬃcient scaling of their CBM initiative, how they can consider strategic partnerships to access complementary resources, while also embarking on adaptability by running business pilots either internally or by engaging in collaborative networks for industry-wide


Introduction
The fashion industry is not only one of the largest and most influential industries in the world, it is also one of the most polluting and resource-draining ( Ellen MacArthur Foundation, 2017 ;Pal and Gander, 2018 ).According to the European Environment Agency (2019) , the textile industry is the fourth largest user of raw material and water globally.As the circular economy has gained in interest, it has been identified as a way to mitigate the current environmental challenges by moving away from a take-make-dispose system.A circular economy, which is often referred to as "an industrial economy that is restorative or regenerative by intention and design" ( Ellen MacArthur Foundation, 2013 ) is of course an appealing concept for any industry since it implies that economic growth can be decoupled from environmental pollution.
From a business model perspective, this means finding new rationales for the creation and delivery of value, that build on cir-cular economy principles ( Lüdeke-Freund et al., 2019 ).Thus, by developing circular business models (CBMs), the fashion industry should be able to break the pattern of pollution and overproduction and move towards becoming environmentally sustainable ( Ellen MacArthur Foundation, 2017 ;European Environment Agency, 2019 ).With business models building on renting, reuse, or resale, resources stay at their highest value for a longer time.This will, in theory, result in less resources used and thus, less pollution, while still offering customers the fashion products they want ( Ellen MacArthur Foundation, 2017 ; The Business of Fashion 2019 ).There is no lack of examples of CBM initiatives and experimentations in the fashion retail value chain, as can be found documented in various industry reports (e.g. Ellen MacArthur Foundation, 2017 ;Global Fashion Agenda, 2019 ) and fashion brands and retailers sustainability reports.Nonetheless, most of these initiatives are niche activities or pilots.
However, for a systemic change towards a circular economy to take place CBMs not only need to be implemented they also need to have the ability to gain market shares from the currently dominant linear business models ( Bocken et al., 2018 ;Pal and Gan-der, 2018 ).In a report by the European Environment Agency (2019) , lack of scalability of CBMs was identified as one of the main reasons for the slow transition to a circular economy in the textile industry in Europe.Other studies have also found the lack of scalability of CBM to be a reason for the slow transition (e.g.Franco, 2017 ;Pal and Gander, 2018 ) even when demand for such circular models can be expected to be high ( Franco, 2017 ).Thus, it can be deduced that, even though pilots and start-up initiatives are important for the transition to a circular economy by contributing to innovation and new technology development ( Rok and Kulik, 2020 ), they also need the ability to scale if they are to have an impact.However, extant literature is currently scant when it comes to elaborating on what type of strategies can be useful for scaling CBMs.
The current notion of business model scalability is limited to the perspective of organisational growth or increased turnover for economic gains ( Jocevski et al., 2020 ;Nielsen and Lund, 2018 ;Stampfl et al., 2013 ), for example in terms of economy of scale to achieve economic gain in the creation, delivery and/or capture of value.This perspective fails to cover the triple bottom line perspective, i.e. environmental and social values beyond just the economic ones embedded in the CBM context.Some wider evidence on scaling can be found in social innovation literature in terms of impacting social values ( Bauwens et al., 2020b ;Bloom and Chatterji, 2009 ), this is however not full-fledged in CBM literature.Thus the notion of scalability of CBMs remains partially unattended due to the lack of more holistic and encompassing discussion on it beyond focusing only on the economic perspective.
In light of this, a fundamental argument made in this paper is that CBMs will remain as side/niche activities, decoupled from the main business model, as was also illustrated in Stål and Corvellec (2018) 's study of take-back schemes, unless the notion of scalability is deciphered both, concretely contextually and broadly in scope.There is therefore a need to investigate the underpinning logics and strategically how firms should approach scalability applied or adapted to CBMs in a fashion retail context.
Therefore, the main purpose of this paper is to explore the main strategic approaches to scale business models and how these can be applied to CBMs in the fashion retail value chain .To guide the study, two research questions are posed: What strategic approaches for scalability can be found in business model literature?In what way are these strategic approaches applicable for CBMs in the fashion retail value chain?
Given the dearth of literature specific to understanding and revealing CBM scaling strategies, we take a stance in reviewing general business model literature, and subsequently adopting and adapting this know-how specifically for fashion CBMs.This is done through analysis of "grey literature" in the form of industry reports, sustainability reports and press releases that describe activities used by fashion brands to increase the scale of their CBM initiatives.In doing so, the paper aims to contribute to the CBM literature by providing a conceptual model that summarises the strategic approaches for scalability and elaborates on the application to fashion retail value chains.This can also be useful for practitioners looking to understand potential paths to scale their CBM initiatives.Additionally, the paper contributes to a broadening of the scalability concept to cover aspects that support the decoupling of consumption from production which is embedded in CBMs.
The paper is structured as follows.First, the concept of CBM is introduced, both in general terms and in the context of the fashion value chain.This is followed by the method section, including data collection, descriptive statistics and thematic analysis.Subsequently, the development of the conceptual model is described in the results section, followed by a discussion and contextualisation of the conceptual model to CBMs in the fashion value chain.Fi-nally, the conclusion of the study is presented and limitations with the current study as well as future research areas discussed.

Literature review
This section provides a brief introduction to CBM, what it is and how it relates to scalability.The concept is first presented in general terms and then in the context of the fashion retail value chain.

Circular business models
CBM is a term used to describe business models that build on circular economy practices, highlighted as 3R: Reduce, Reuse and Recycle ( Lieder and Rashid, 2016 ).Even though there is no clear definition of what constitutes a CBM, the general notion is that it is a model for describing how an organisation creates, delivers and captures values, while keeping resources at their highest value and for as long as possible ( Frishammar and Parida, 2018 ;Nußholz, 2017 ).In a purely circular economy there is no waste, everything is looped back in different resource flows.In the influential framework by Bocken et al. (2016a) this is described in terms of closing, slowing or narrowing resource loops.Within these loops a large variety of CBMs emerge; while some cover the entire resource flow, others only focus on a specific activity and need to connect to other CBMs to create a complete loop.In a study by Lüdeke-Freund et al. (2019) , six different types of CBMs were identified that support resource flows (referred to as "major CBM patterns"); these are: Repair & maintenance, Reuse & redistribution, Refurbishment & remanufacturing, Recycling, Cascading & repurposing and Organic feedstock.
The discourse around circular economy and embedded CBMs often focus on environmental and economic benefits even though it is generally understood that CBMs should embrace a triple bottom line perspective ( Geissdoerfer et al., 2017 ;Kirchherr et al., 2017 ).Thus, CBMs are seen as a way to support sustainable development in terms of reduced carbon emissions and waste while simultaneously encouraging economic growth.However, the third dimension of sustainable development, the social dimension, has so far received less attention ( Kirchherr et al., 2017 ;Kristensen and Remmen, 2019 ).It is postulated that a circular economy will lead to more job opportunities ( Bauwens et al., 2020a ), but beyond that, the link to social values is less developed ( Geissdoerfer et al., 2017 ).
What is more commonly agreed is that CBMs play an important part in the transition from a linear to a circular economy.The body of literature covering the innovation and implementation of CBMs is growing ( Ferasso et al., 2020 ;Lopes de Sousa Jabbour et al., 2019 ).Lieder and Rashid (2016) discuss the implementation of circular economy principles with the emphasis on environment, resources and economic benefits.Bocken et al. (2018) draw from eight cases to show how experimentation can be an important key to innovation and initiate the transition to a sustainable business model.Scaling up the business model is in these studies seen as the last step of the implementation and innovation process, and even if it is an important step it is seen more as something that will inevitably happen if the implementation and groundwork are done properly and perhaps it is therefore less explored.

Circular business models in fashion retail value chain
In the fashion retail value chain, CBMs often operate within one or several of the following areas: take-back of used products, resale of used products (second-hand), renting, repair, redesign or use of recycled materials ( Pal and Gander, 2018 ).Except for the use of recycled material, these CBM areas work on retaining the value of the product as the key circular feature.Recycling, and to a lesser extent redesign, work further down in the waste hierarchy and instead retain the material value.Different design strategies are sometimes incorporated in the CBM to help facilitate recycling and/or extended product life ( Jia et al., 2020 ).This can be activities such as using more mono-material, designing for disassembly, designing for easier repair, optimising material usage or increasing the durability ( Global Fashion Agenda, 2018 ;Jia et al., 2020 ;Pal and Gander, 2018 ).
When it comes to the scalability of these business models both industry and policymakers have identified this to be a challenge.The European Environment Agency (2019) , as well as the Ellen MacArthur Foundation (2017) , recognise that many initiatives currently exist in the industry.Most established fashion brands and retailers frequently allude to circularity and CBMs in their sustainability reports and experiment with different circular activities like garment collection and resale schemes.However, in the report from the Ellen MacArthur Foundation they also write that "these effort s offer solutions and demonstrate promising progress in various areas, but are fragmented and often only effective at small scale" ( Ellen MacArthur Foundation, 2017 , p.26).A reason for this could be the decoupling between the general business of the company and the CBM activity that Stål and Corvellec found in their study of Swedish apparel firms (2018).CBM activities are to large extent handled as a separate and niche sustainability activity and are not allowed to influence the core (and mainstream) linear business model.Similarly, when studying companies developing cradle to cradle products Franco (2017) found that even though the general perception was that the demand for these products was on a rise, it was in reality still rather limited which affected the possibility to take advantage of economy of scale that comes from volume.This, in turn, affected further development and smaller firms from going mainstream with their circular collections.What makes CBM in fashion retail value chains even more complex is that customers do not only make up the demand side, they are also part of the supply side since they need to return their unwanted clothes for them to re-enter the resource loop ( Ki et al., 2020 ).

Methods
To explore the business model scalability a two-part approach was adopted.The first part of the study is a systematic literature review (SLR) where the focus lies on examining what has been written on the subject of business model scalability in peerreviewed journals.The literature is analysed both by descriptive statistics to get an overview and qualitatively to identify summarising themes for strategies suggested for business model scalability.
The second step aimed to contextualise the SLR findings to that of CBMs in the fashion retail value chain.This is done by building from CBM-activities and goals within the scope of the industry-wide initiative "2020 Circular Fashion System Commitment" ( Global Fashion Agenda, 2018, 2019 ).

Data collection method for the systematic literature review
The method of SLR was chosen since it represents a structured way of selecting and collecting academic literature within a specific field or scope ( Snyder, 2019 ;Wolfswinkel et al., 2017 ).The strength of the SLR lies in the rigours, transparency and replicability of the process that helps to limit bias, and reduce chance effects ( Reim et al., 2015 ;Tranfield et al., 2003 ).
As suggested by Snyder (2019) as well as Wolfswinkel et al. (2017) , designing a literature review is an iterative process.Different experimental search criteria and Boolean word strings were therefore constructed and evaluated before the final search criteria were chosen.For instance, "circular business model" was originally included in the search but it was revealed that very little has been published on scalability in relation to CBM.In particular, "scal * " in combination with "CBM" or "circular business model" gave only eight hits in Scopus and none of them were relevant to the purpose of this study.Thus, since the field of CBM scalability was proven to be rather limited, the broader scope of business model scalability was chosen for the SLR.This gives insights from a wider body of literature, covering related areas such as innovation, social entrepreneurship, strategy, and business model research.
The final design of the SLR, which is depicted in Fig. 1 , follows the steps outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement ( Page et al., 2021 ).The search string that was finally used in the database search was "business model" and "scal * " in the title, keywords or abstract.Other delimitations made in the database search were that only peer-reviewed articles and book chapters in English were included, and only those published in journals relevant to business research.The search was carried out using the databases Scopus and Web of Science due to their large range of peer-reviewed journals within business and management research scope.During the subsequent review process, which is depicted in the PRISMA flow diagram ( Fig. 1 ), 195 duplications were first removed followed by the screening of the remaining 900 articles.The screening was mainly performed by the first author but checked with the second author for validation whenever there was any ambiguity.During the first screening step, 751 articles that did not meet the criteria of discussing business model scalability were excluded.These articles were excluded mainly due to two reasons, first, the articles included the word scale only in terms of something used to measure (e.g. the noun "a scale").Secondly, the articles discussed the scaling of technology without a business model perspective.This first screening through title and abstract reading yielded 149 articles which were all read in full to further scrutinize their relevance according to the purpose of detailing business model scalability rather than referring to it in a "passing" discussion, thus removing 92 more articles.For example, Agrawal and Tapaswi (2017) mention how cloud computing offers scalable business models but focus on the threat of distributed denial-of-service attacks (DDoS attacks) on those business models, not the business models themselves; Ingemarsdotter et al. (2019) discuss the use of internet of things (IoT) for circular strategies, with the focus on the implementation of such strategies rather than CBM scalability.These type of articles were subsequently removed to create a final list of 57 articles for analysis.

Descriptive statistics of selected papers for the systematic literature review
To get a representative overview of the collected literature, each paper was classified according to three predetermined descriptive elements: (i) the distribution over, (ii) journals of publication, and (iii) primary research method.
As is illustrated by Fig. 2 , the first relevant article on the subject of business model scalability found in the SLR is from 2001.This fits in well with that it was during this time that the concept of business models started to emerge more frequently in academic research ( Zott et al., 2011 ).However, what is noticeable is that there is an increase of papers from 2018 and onwards which indicates a growing interest in the notion of business model scalability.This seems to be driven by the fields of social ventures and information and communications technology, to which approximately 40% of the papers are dedicated.
The paper corpus represented in the SLR cover 39 different journals and one book chapter.Of these, the Journal of Cleaner  Production has the highest number of represented papers (8).Additionally, six other journals have more than one relevant publication and 33 journals featured only a single paper relevant to the subject of the SLR (see Table 1 ).Qualitative methods, particularly case studies, are by far the most commonly used methods in the included papers.As can be seen in Table 1 , 39 of the 57 papers use case studies as their main research method.The use of multiple cases is somewhat more common than single cases.

Thematic analysis of the systematic literature review
The analysis started from the literature included in the SLR, but also took inspiration from existing business model literature and relevant theoretical underpinnings, in what can be described as a qualitative and abductive approach ( Alvesson and Sköldberg, 2008 ).This was done in what Kovács et al. (2005) describe as creative and iterative learning loops which helped to develop and fine-tune emerging themes.The first step of the analysis was to re-read the included papers and synthesise the content ( Flick, 2009 ) with respect to strategies for business model scalability.This was done by the first author.Initial themes were then developed from the synthesised document in discussion with the second author and a third researcher.
In the second step, the initial themes were further developed by going back and forth between coded excerpts and the data.Different techniques were used for coding categories reorganising, re- examining, merging or separating them.The sorting and organising of coded excerpts were primarily done in tables using excel sheets, but more hands-on techniques such as "cutting and sorting" ( Ryan and Bernard, 2003 ) was also used to get closer to the material.The first author took the lead in this while the second author kept a distance and instead had the role of asking critical questions and push for clarification and elaboration to ensure relevance and avoid bias.
In the third step, which in reality, following an abductive process, happened simultaneously as step two, theoretical insights were drawn from other relevant literature to find patterns and fine-tune the overarching themes.The final themes, mainly draw theoretical insights from two prominent thematic orientations captured within the business model design/innovation literature.The first thematic orientation address the boundary of the organisation and the aspect of "closedness" vs. "openness" for managing business models ( Osterwalder and Pigneur, 2010 ).That is, whether or not strategies rely on boundary-spanning resources and capabilities, or on resources and capabilities within, or that can be absorbed into, the organisation's control ( Chesbrough, 2006 ).The second thematic orientation draws on business model design from the point of view of efficiency vs. novelty ( Miller, 1996 ;Zott and Amit, 2007 ), as these reflect fundamental alternatives to create value.This thematic orientation thus deals with efficiency-centred strategies which have the aim to minimise costs and increase efficiency ( Williamson, 1983 ), and adaptability/novelty-centred strategies that take advantage of changing conditions and environments and use that to improve business model scalability.
During the process of alternating between step two and three, it was noticed that strategies for improving scalability could be assigned to both of the thematic orientations simultaneously.This lead to the conclusion that organisational strategy for scaling can be viewed along two dimensions, in other words, along a two-bytwo matrix where the relationship between the different themes are more profound (see Fig. 5 ).This setup, with four strategic approaches to business model scalability, was therefore used for final coding and analysis and constitute the foundation of the concep-tual model.Fig. 3 show the final data structure that resulted from the analysis.
Inter-coder reliability was ensured through the process of discussing and jointly reflecting on the different interpretations.Seuring et al. (2012) call this approach "discursive alignment of interpretation" and it is a suitable approach when the themes are not clearly distinct, as is the case here.Through assessing and resolving individual differences in judgement, the coding was adjusted to achieve consensus amongst the researchers.

Data collection and analysis of grey literature
To contextualize as well as complement the findings in the SLR of peer-reviewed papers on business model scalability, the study also draws on literature from practitioners and industry, what is generally known as "grey literature" ( Adams et al., 2017 ;Schöpfel, 2010 ).This approach is commonly used in management and organisational studies for adding contextual information missing when reviewing peer-reviewed papers.Thus, it allows for emerging conclusions from one literature stream to inform the other ( Adams et al., 2017 ).
For this study, the grey literature corpus consists of (i) two status reports from the multi-stakeholder fashion industry initiative "2020 Circular Fashion System Commitment" which summarise the activities of numerous leading fashion retail brand's attempts to scale their CBM initiatives, and (ii) sustainability reports and press releases of the participating brands and collaboration partners.The "2020 Circular Fashion System Commitment" initiative was chosen primarily for two reasons.Firstly, the aim fits well with the scope of this study, which is to explore how organisations can scale up their CBMs, thus aiding the industry's transition to a circular economy, and secondly due to the wide coverage of the initiative.Goals and activities from 76 leading fashion brands worldwide on how they have scaled their CBM initiatives, or planned to, were reported in the latest status report ( Global Fashion Agenda, 2019 ).This gives a good representation of the type of activities organisations in the fashion retail value chain are currently working with and what type of strategic approaches they adopt.The data collection was done systematically by using the industry initiative "2020 Circular Fashion System Commitment" as a starting point for the data gathering.Complementary information was collected from publicly available sources, such as company homepages and sustainability reports.This was done purposefully using google search as well as company homepages' search functions, and for two reasons.First, if the actions and goals reported in the status report did not reveal sufficient elaborations on how a company planned to fulfil their commitment.Secondly, to acquire additional information on the representative actions and companies that were selected as illustrative cases.
All data was saved in an excel sheet and sorted company-wise.An initial mapping was then carried out by the first author.This roughly followed the four CBM activity areas that the "2020 Circular Fashion System Commitment" is shaped around, that is: increasing the volume of used apparel collected, increasing the volume of used apparel resold, increasing the share of apparel made from recycled post-consumer textile fibres, and implementing design strategies for cyclability.This was followed by analysing the goals and activities within each of the four areas to identify and code the types of strategies the companies planned to use to fulfil their commitment.The coded strategies were then plotted against the conceptual model developed from the SLR of peer-reviewed papers; this was done jointly and in discussion between the two researchers.An example of the coding is shown in Fig. 4 .Lastly, representative actions and companies were selected based on how well they illustrated the approach as well as the availability of complementing information.

Results
This section presents the conceptual model for business model scalability analysis derived from the SLR.
As is described in the method section, the analysis revealed four strategic approaches to business model scalability along two dimensions in a two-by-two matrix (see Fig. 5 ).In this suggested conceptual model, closed business model strategies verge towards being either efficiency-centred or adaptability-centred.The closed and efficiency-centred strategies, represents a "do it yourself" approach, by not only relying on the resources within the boundaries of own organisation but also by organising these to obtain efficiency through standardisation, automation and/or centralisation.Whereas the closed and adaptability-centred strategies, which represents the fundamentals of an "absorb external ideas and opportunities" approach, also rely on the resources within the boundaries of the own organisation but listen, experiment or in other ways absorb ideas and opportunities that can be used for scaling up.In the same way, the open business model strategies could be either efficiency-centred or adaptability-centred in nature.Open and efficiency-centred strategies, which represents more of a "divide the labour" approach, look outside the boundaries of own organisation but still strive for efficiency in those collaborative strategies.Whereas the open and adaptability-centred strategies, resonating the fundamentals of a "create together" approach, also look outside the boundaries of own organisation but use these collaborations to experiment and create new connections and opportunities.
Below the four strategic approaches of the conceptual model are elaborated on and the rationale behind them described.

Do it yourself
This approach focuses on how organisations can utilize their internal resources and capabilities to improve the scalability and the process of scaling.Here the focus is on how the own organisation can be organised for improving scalability, by executing strategies such as standardisation of products and process routines ( Gao and Liu, 2020 ;Kohler, 2018 ;Mathur et al., 2020 ;von Krogh and Cusumano, 2001 ), centralisation and automated processes ( Bohnet-Joschko et al., 2019 ;Niemczyk et al., 2019 ;Stampfl et al., 2013 ;Yoho et al., 2018 ).This makes it possible for the organisation to exploit economy of scale and keep fixed cost low when scaling up.Standardisation is also a prerequisite for replication strategies since replication require that the business model and its operations can be described and transfer to new locations and understood by new people ( Winter and Szulanski, 2001 ).Even though some local adaptions might be necessary, the key lies in the ability to transfer vital elements of the business to the new location in a standardised way ( Dobson et al., 2018 ;Gebauer et al., 2017 ;von Krogh and Cusumano, 2001 ).Winter and Szulanski (2001) refer to this as the "arrow core" and discuss how the dynamic capabilities of the focal company are important to deepen the understanding of these key elements with every replication.Replication can also take the form of franchising where the franchiser remains in control of the concept and centralised administrative areas ( Gillis and Castrogiovanni, 2012 ).
Connected to standardisation the literature gives examples of centralising core activities and automation.von Krogh and Cusumano (2001) suggest that centralising core activities can give better conditions for handling increased transactions when the organisation scale, as well as reducing administrative costs.Using technology for automation is another way to reduce administrative costs and facilitate scale.However, even though technology has great potential for supporting scalability ( Kohler, 2018 ;Mohan and Potnis, 2010 ) by supporting more users and transactions, Stampfl et al. (2013) caution that relying too heavily on technology can have a negative effect on the quality of customer service and mean that there needs to be a balance between human interaction and reliance on technology.
An organisation can also find opportunity to scale by building on its existing resources and competencies, thus using them more efficiently.Using existing distribution channels to reach new customers ( von Krogh and Cusumano, 2001 ), develop complementing products or distribution channels ( Bigdeli et al., 2016 ;Blundel and Lyon, 2015 ;Mancha et al., 2019 ;von Krogh and Cusumano, 2001 ) and increasing the perceived value of the offer by marketing activities ( Bailetti et al., 2020 ;Biloshapka and Osiyevskyy, 2018 ;Walske and Tyson, 2015 ) are some examples of this.

Absorb external ideas and opportunities
This approach is similar to the "do it yourself" approach when it comes to building upon own resources and competencies.However, in contrast to focusing on efficiency to achieve scale, this approach builds on being alert to changes and adapting to them.By being alert and absorbing ideas and opportunities into the organisation, new learnings leading to new opportunities for scale can be created.Blundel and Lyon (2015) for example, show how a social enterprise working with childcare in London was able to scale its business by adapting to new legal and social conditions.Where childcare previously had been offered in the form of a social charity, changes in the legislation made it publicly funded which presented an opportunity for scaling the operations.This is not easily planned for, instead, it is an approach that builds on being prepared and having the capabilities to absorb ideas and opportunities into the organisation as they emerge.Another example often suggested is to take advantage of emerging technology and by adapting to the opportunities that may bring, reach new markets in novel and more cost-efficient ways ( Biloshapka and Osiyevskyy, 2018 ;Bocken et al., 2016b ;Yoho et al., 2018 ).
Listening to, and learning from customers and other stakeholders are also highlighted as a strategy to find new opportunities to scale ( Bailetti et al., 2020 ;Biloshapka and Osiyevskyy, 2018 ;Gebauer et al., 2017 ;Giudici et al., 2020 ).This however needs to be combined with experience to know what to make of that information.One difficulty in scaling geographically is for example that a new market may have local legislation or customs that clash with the current business setup.Hiring local managers and giving them the flexibility to adapt to local conditions, still keeping the core of the business model intact, can be a way to get around this ( Mohan and Potnis, 2010 ;Stampfl et al., 2013 ).Replicating the business model in a new location can in this case also serves as an opportunity to experiment, which can help to develop the business model further for increased scalability ( Gebauer et al., 2017 ).Overall, running pilots and experiments is suggested as an important part of the strategy to scale, not only as a first step towards something bigger but as a tool for learning and exploring opportunity as part of the scaling process ( Cavallo et al., 2019 ;Esposito et al., 2012 ;Gebauer et al., 2017 ;Goyal et al., 2017 ) In this approach, mergers and acquisitions can also be found since this is a way of absorbing competencies and resources that is not initially within the bounds of the organisation.This can be a way of adapting to changes and leapfrog competition to scale up quickly, especially if high investments are needed to reach a new market ( Acquier et al., 2019 ;Holm and Günzel-Jensen, 2017 ;Niemczyk et al., 2019 ).

Divide the labour
Following this approach, organisations find ways to increase their business model scalability in collaboration with others with the purpose of increasing efficiency.With strategic partners, organ-isations can access resources and experiences that the organisation do not have, or that would be too costly to develop and maintain ( Acquier et al., 2019 ;Ćwiklicki, 2019 ;Nielsen and Lund, 2018 ;Shrimali and Sen, 2020 ).It can also be a way to access local knowledge or gain legitimacy ( Bocken et al., 2016b ;Walske and Tyson, 2015 ) by partnerships instead of hiring own staff.Interestingly, in a study of how social enterprises scale, Walske and Tyson (2015) found that securing partnerships often had the additional benefit of making it easier to secure other partnerships.The first partner gave legitimacy to the organisation which made it easier for others to join.This could be important if the business model requires external distributors or investors to scale.
In addition to giving access to resources, capabilities and experiences outside the boundaries of the own organisation, the SLR revealed another type of strategic partnership that allows the business model to scale: the "free-franchising" model.When the aim of scaling up is to increase the impact of the business model rather than the revenues, some organisations (often social enterprises) find that an efficient strategy is to encourage partners to replicate the whole business model, free of charge ( Blundel and Lyon, 2015 ;Goyal et al., 2017 ;Minoja and Romano, 2021 ).If the main goal is to impact as many people as possible, encouraging and teaching others to replicate a successful business model actually make sense.However, for this to be successful the business model first needs to be fully understood and standardised so that it can be communicated to entrepreneurs outside the organisation.Goyal et al. (2017) discuss a case where an organisation working to bring energy solutions to the rural areas of India build incubation centres where they teach local entrepreneurs how to replicate the concept.In this way, the organisation could scale up the impact of their business model without being directly involved.The key was to be able to transfer the knowledge to these local entrepreneurs so that they understood how to successfully build their own local social enterprise, and this is only possible by standardising the information.Another example can be found in Acquier et al. (2019) study of sharing economy where resource-hungry initiatives, such as MIT Fab labs (also known as makerspaces) aim for their network to reach a global scale by presenting a standardised concept for engaged people in other parts of the world to replicate.In this way, they replicate their business model and reach a global scale by relying on local labour and investments.
Collaborative networks or platforms can in some cases also work as a way to better take advantage of economy of scale by pooling resources ( Esposito et al., 2012 ).It can also be a way to distribute cost by letting others work for you.Nielsen and Lund (2018) and Zhang et al. (2015) highlight that engaging and leveraging the work of customers can help with scalability, especially as a cost-effective way of marketing.This relates to the freemium models often used in digital business models where non-paying users (e.g.free users) are used to take advantage of informational increasing returns through viral adoption and network effects ( Holm and Günzel-Jensen, 2017 ;Täusc her and Abdelkafi, 2018 ;Zhang et al., 2015 ).However, to be able to handle a large number of platform interactions and be able to scale up to even more, standardisation and automation is key ( Holm and Günzel-Jensen, 2017 ).

Create together
Creating together requires engagement and openness.Not only do these strategies build on collaborating with organisations and customers outside the boundaries of the own organisation, they also require letting go of some control and allowing partners access to resources or information (at the same time getting access to other resources or information).Thus, even though they might touch on the "divide the labour" strategies, they differ in the closeness of the engagement and the search for new opportunities and connections.The "free franchising" model of Fab Labs discussed as a "divide the labour" approach earlier, also touch on "create together" strategies.Where the "free franchising" model use standardisation to be able to scale, it also has an element of a community ( Acquier et al., 2019 ) and new opportunities could be created if the local replications and the learnings generated are used to adapt the main business model for scale.
Collaborative networks and platforms were also previously mentioned as a way to "divide labour" and take advantage of economy of scale.However, it can also be an adaptability-centred strategy where customers or even competitors are invited to participate in the development process.This can for example be about encouraging suppliers to develop complementing products as Zhang et al. (2015) suggest, and in the process also scale up the reach of the original organisations business model through increased informational returns.
The "create together" approach can also be found in the social venture literature, but perhaps in a slightly different way.When studying the scaling of sustainable business models at the base of the pyramid, Palomares-Aguirre et al. ( 2018) find that engaging the community in projects can be a key to scaling, not only as a way to expand the project by word-of-mouth but also to create a lasting systemic change.Thus, by engaging with the local community adaptions are made based on local customs and habits, which not only facilitates adoption but scalability of the business model.

Discussion
In this section, the conceptual model for scalability developed through the SLR and its four strategic approaches are discussed and contextualised to circular business models in fashion retail value chains.Well-represented cases listed in the initiative "2020 Circular Fashion System Commitment" have been analysed and is here used to illustrate how the four strategic approaches for business model scalability relate to the context of CBM in the fashion retail value chain.Table 2 provides an overview of the mapping of the CBM scaling goals and activities pursued by the majority of the companies listed in the initiative, in terms of how these are aligned to, and conforms, the conceptual model.
Starting with garment collection schemes, evidence can be found of activities that can be classified under all the strategic approaches suggested in the conceptual model.The efficiency-centred approaches, i.e. "do it yourself" and "divide the labour" may be more prominent, but there are a few companies that allude to more experimental strategies where they adapt to, and absorb, new opportunities both within the boundaries of the organisation and in collaboration with other stakeholders in the industry.For the area relating to resale, again the two efficiency-centred approaches are the more prominent.However, in the area relating to recycling, evidence is instead more focused on adaptability-centred approaches.Both in the form of "creating together" in different projects aiming for industry-wide solutions, and by absorbing and adapting internal routines to circular economy practices.This can also be seen in the design process where most refers to strategies relating to how the companies can absorb the concept of circular economy into their current setup by educating design teams as well as customers on how to apply these principles.
Sections 5.1-5.4further elaborates on these different strategic approaches together with some illustrative examples of activities from representative companies in the "2020 Circular Fashion System Commitment".

Table 2
Strategies derived from the goals reported to the "2020 Circular Fashion System Commitment".

Do it yourself
> Make take-back available to more customers by more take-back points or easier access (e.g.replication and use of existing resources and capabilities) > Make customers aware of the collection scheme by marketing and information campaigns (e.g.use of existing resources and capabilities) H&M Group, Inditex, Lindex, Gina Tricot Divide the labour > Collaborate with logistics firms to make take-back available to more customers, the organisation is only the facilitator (e.g.strategic partnerships) Esprit, PVH Corp.

Implementing design strategies for cyclability
Absorb external ideas and opportunities > Make customer care information and repair kits more available to prolong garment life (e.g.listen and adapt) > Training for designers and other internal teams about circular design principles (e.g.listen and adapt) > Implementing design principles/guidelines such as design for reparability and recyclability (e.g.listen and adapt) Nike, Filippa K, KappAhl, Kering, Marks & Spencer, Eileen Fisher

Do it yourself -how companies leverage their existing business to scale circular business model activities
Common for the "do it yourself" strategies found in the "2020 Circular Fashion System Commitment" are that companies often piggyback on their linear business models.Most of the fashion retailers and brands in the initiative already have established linear business models, and therefore have access to resources such as sales channels and logistics centres.These channels and resources are taken advantage of in order to scale the CBM activities, without reaching outside the boundaries of the organisation for support.Companies like the H&M group, Inditex, Lindex and many more, use their existing network of stores to implement and subsequently scale up their garment collection schemes using strategies such as fine-tuning and simplifying the routines followed by replication of the take-back concepts to more stores.The H&M group, for example, started their garment collection scheme in 2012 after first running a pilot in Switzerland ( HandM Group, 2013 ).The following year 3047 tonnes were collected in H&M stores around the world ( HandM Group, 2014 ).The volumes have since increased as the concept was scaled up further by replicating to more stores as well as to the different brands within the H&M group.This was combined with supporting marketing activities.In 2019 they reached 29,005 tonnes, which is more than their goal of 25,0 0 0 tonnes yearly by 2020 ( HandM Group, 2020a ).
The same logic is applied for the resale area as well.Companies that use strategies related to a "do it yourself" approach take advantage of the existing resources and build on those to scale the CBM activity.Nudie jeans for example started to offer resale of their jeans in their own stores in 2013 ( Nudie Jeans Co., 2020 ).In addition to scaling the offer to include all stores (i.e.replication), they later added the offer to their own online store as well (i.e.build on existing resources).An interesting point in relation to closed strategies for scaling is that Nudie Jeans sell their new products wholesale as well as through their own retail channels, even so, the resale offer has so far only been scaled up using their own channels.
From the SLR we found that a key to efficient scaling is to keep fixed costs low.Keeping the process simple and piggybacking on the existing infrastructure seems to have that effect.In other words, some of the fashion retailers/brands are building on their core business and areas of experience to gain efficiency in the process of scaling up.A similar logic was found in the SLR where von Krogh and Cusumano (2001) wrote that "to grow by scaling, a company expands product development around core technologies and offerings, expands product lines and increases the intensity of marketing by using existing distribution channels to reach new customer groups with related needs."(p.54) Based on the discussion above we propose: Proposition 1: Organisations that already have access to resources and capabilities within their existing business model can leverage these efficiently, through centralisation, replication, standardisation and automation, for scaling of CBM activities.

Absorb ideas and opportunity -how companies scale up their circular business model activities by adapting their internal routines to follow circular economy principles
The participation in the "2020 Circular Fashion System Commitment" can in itself be seen as a strategy within the "absorb ideas and opportunities" approach.It is a platform for coming together, learning from each other and sharing information to change the direction of the industry ( Global Fashion Agenda, 2018 ).The purpose of this collaboration is to get new ideas and learnings that can be absorbed into the own organisation and/or facilitate change at an industry level, not to co-create or be involved in each other's or-ganisations.Identifying the need for a transition to a circular economy is the first step, but to make the transition and scale up, the companies need to understand what that means to their organisations and adapt accordingly.
Changing the design process by implementing strategies for cyclability is stressed by the "2020 Circular Fashion System Commitment" status report ( Global Fashion Agenda, 2019 ) as an area where companies need to learn and adapt their current way of working to be able to make a lasting change towards a circular economy.There is no uniform solution, taking part in information and training is the first step, each organisation then needs to translate that into best practices for their organisations.For Nike this meant implementing scoring tools in the design process so that the teams can better analyse the footprint of each product connected to design choices and make more informed decisions ( Nike Inc, 2019 ).Others have a more straightforward approach, like Filippa K that, amongst other things, aim to make 25% of their collection in mono-material for easier recycling at end-of-life ( Filippa, 2020 ).
Ideas and opportunities to scale up CBM activities can also come from experiments.In the SLR we found that pilots and experiments could be a tool for learning and exploring opportunity as part of the scaling process ( Dobson et al., 2018 ).amongst the goals for scaling up garment collection we find that Nike indicates using this approach for scaling up their collection of post-consumer footwear "…building upon Nike's nearly 30-year history collecting post-consumer footwear, Nike will launch new pilots to test methods of collection and gather insights on consumer engagement" ( Global Fashion Agenda, 2019 ).They are however not alone in this and there are several examples of pilots in the commitment's status reports (2018,2019).However, there is a difference between a pilot that is launched with the intention of being the first step before scaling up an initiative, and pilots running with the intention to generate learnings and testing ideas.Here it is the second type that we refer to.These type of experimental pilots can generate learnings that can be transferred to the CBM to increase the potential to scale.Just as the Nike pilots above that are used as an approach to test methods and gather insights.
Thus, when using adaptability-centred strategies for scaling up, the process is not continuous but episodic and interrupted by new learnings and opportunities.These can come from industry initiatives (such as the "2020 Circular Fashion System Commitment"), own experimentations and pilots or changing status-quo conditions.By exploring those learnings and adapting to them, opportunity to scale is created and exploited.Blundel and Lyon (2015) also noticed this phenomenon in the study of scaling and replicating social ventures.In the study, they write that organisational growth "…may not be occurring in a steady period of change, but rather occurs following critical incidents such as the winning of a new contract or the development of a new market" (p.91).
Bas ed on the discussion above we make the following proposition Proposition 2: Organisations can create or discover opportunities for scale by actively searching for information from industry networks and customers, or by running own pilots and experiments.The key is to act on these opportunities by absorbing them into the organisation.

Divide the labour -how companies use strategic partners to help scale their circular business model activities
Instead of relying on their own resources, some companies choose to form strategic partnerships where they can divide the labour for scaling the CBM activity.This can be seen in two of the areas from the "2020 Circular Fashion System Commitment"; garment collection and resale.Esprit, for example, collaborates with Packmee to scale up the volume of garments collected from their customers ( Global Fashion Agenda, 2019 ).Packmee is an organisation that offers a service to customers that makes it possible for them to donate their unwanted garments directly to a charity by packing them in a box and printing a free shipping label ( Esprit, 2020 ).This not only helps Esprit's customers with a convenient standardised collection system but also adds to Esprit's goal to increase the amount of garments collected from their customers, without the need for Esprit to make additional investments or reconfigure their business model.
However, this approach is possibly more widely used for resale.Even though many of the larger retail companies have implemented garment collection schemes in their stores, they turn to partners to help scale up the resale of those garments.Inditex for example has the goal "by 2020, we will have established partnerships with 40 local non-profit organisations and recyclers for the re-sale of used garments" and the H&M Group write that "by 2020, 40-60% of all garments collected in H&M brand's stores will be resold and re-used through our partnership with I:CO" ( Global Fashion Agenda, 2019 ).
A reason for this is most likely the resource-heavy process of sorting which requires both expertise and special equipment ( Sandberg et al., 2018 ).It is, therefore, more efficient to establish partnerships with professional sorting companies or charities and donate or sell the collected garments to them instead of taking care of them in-house.Previous studies have also found that these partnerships lead to more cost-efficient processes and economies of scale since they give access to complementary resources and capabilities ( Pal, 2017 ), which lead to better scalability.
It should be added that it is not only labour that will be divided using this approach, the created values will also be distributed between the partners.If we take the example of I:CO and H&M.Even though there is an economic exchange involved which create a revenue stream for the H&M Group, the larger part of the economic value of the collected garments goes to I:CO after their sorting.On the other hand, H&M can still claim that they, through this partnership, scale up the reuse of garments and thus help minimise their environmental impact.The same goes for the partnership between Esprit and Packmee.Even though Esprit never really touches the products, they can claim that they help to facilitate garment collection and by donating to a charity they also help scale up social value.This suggests that it is important to consider what value across the triple bottom line is important to scale up for the own organisation.Is it the revenue or the environmental impact that the organisation want to scale?In relation to this it is also interesting to note that, just as other research on circular economy has concluded before ( Kirchherr et al., 2017 ;Kristensen and Remmen, 2019 ), there is very little mention in concern to the social dimension of the triple bottom line.
This discussion is summarised by the following propositions: Proposition 3a: Dividing the labour and resources between different organisations can be a way for organisations to take advantage of complementary resources and capabilities for a more costefficient process when scaling up CBM activities outside their areas of expertise.Thus, finding the right partnership can be a source of business model scalability.
Proposition 3b: Organisations that use a "divide the labour" approach for scaling their CBM activity need to consider what value, across the triple bottom line, they aim to scale.

Create together -how companies come together to create new, industry-wide, opportunity to scale up circular business model activities
Some of the companies in the "2020 Circular Fashion System Commitment" initiative also collaborate on a more industrial level to develop solutions that can facilitate scale, especially in the area of recycling.For instance, both Gap Inc. and the H&M Group run research projects with the Hong Kong Research Institute of Textiles and Apparel (HKRITA) to find scalable techniques for recycling ( Global Fashion Agenda, 2019 ).One recently launched project stemming from H&M's collaboration with HKRITA is "Looop", an in-shop garment-to-garment recycling system currently offered to customers in Sweden ( HandM Group, 2020b ).Customers can bring their old garments to a store in Stockholm where the machine shreds them, spin yarn and knits new garments.Even though H&M is the first to offer this technology, HKRITA will licence the technology to make it available to the whole industry ( Hennes and Mauritz, 2020 ).Whether or not this technology will help to scale up recycling or not remains to be seen, and perhaps that is not the intention.This type of technology invites customers to take part in the recycling process which could influence attitudes to recycling and create a better understanding of the value of old garments ( Hennes and Mauritz, 2020 ), thus scaling up the environmental impact by making circular consumption more acceptable and mainstream.
Another industry collaboration is with Renewcell, which is a company that has collaborated with several retail companies in their effort to develop their technology for transforming textile waste to new fibres, at scale.The H&M Group and KappAhl have both invested in Renewcell ( Kejonen, 2019 ) but other retailers such as Levi's ( Renewcell, 2020a), Gina Tricot (Gina Tricot, 2020 ) and Beyond Retro ( Renewcell, 2020b ) has also been involved and contributed in various ways.
In these projects the different organisations are pooling resources, ideas and knowledge to create solutions that will help to overcome bottlenecks for scale for the whole industry.However, by being at the forefront of these collaborative projects, the companies also create a better position for themselves to be the first to take advantage of these opportunities to scale up.This is both in terms of being amongst the first to get access to the technology and in terms of media coverage.It might also be easier for these companies to integrate it into their organisations since they are part of the process from the start and therefore have early access to knowledge and information.
Thus, we can propose: Proposition 4: Creating new ideas and technologies together in collaborative networks and communities is a source of learning, as well as a way to change the industry and thus also creating better conditions for scaling own CBM activities.However, to capitalise on this created opportunity it needs to be integrated in the permanent business model.

Conclusion
This paper is a contribution to the research field of CBM and business model scalability in the context of fashion retail value chains.The suggested conceptual model provides an opportunity to analyse the strategies for scaling up a business model that draws from the notion of "closed" and "open" business models as well as efficiency-centred and adaptability-centred strategies.In doing so, four strategic approaches were identified that adds to the understanding of strategies for business model scalability.Additionally, contextualisation of these four approaches to that of CBMs in the fashion retail context, by analysing strategies currently used in the fashion retail value chain, further results in generating a set of propositions.These build on a larger notion of scalability than that of economy of scale and production volumes.It takes into account other conditions for circular business opportunity (e.g.changing conditions, collaborations) that support the decoupling of consumption from production.As such it broadens the perspective of what business model scalability means in a CBM context.
In terms of practical implications, the conceptual model should be seen as a first step towards a more comprehensive tool for analysing CBM scalability.As such it could help managers and business model developers analyse appropriate actions for their CBM initiatives.Additionally, through the contextualization lessons learned were generated in the form of propositions that can work as guiding principles for industry practitioners aiming for scaling CBMs.These principles can provide important insights for what to consider when choosing different approaches to scale, such as how pilots and experiments can be a tool for generating learnings and increase the CBMs potential to scale through adaption, or how organisations can leverage resources from their existing conventional business model for efficient scaling of their CBM initiative.The discussion also shows the importance of considering what the organisation aim to scale with a specific approach.Different values are for example not always equally distributed in collaborations and one organisation may be able to scale their CBM initiative along both economic and environmental dimensions while other only capture values along the environmental dimension.
In this study, the suggested conceptual model is both contextualised and elaborated on for industrial application.However, more research would be needed to further explore the specific resources and capabilities that dictate what would be the most beneficial match between the scalability approach and CBM activity areas.Moreover, secondary data on the activities for CBM activities could be seen as somewhat shallow.For a fuller, more in-depth understanding of the complexity of CBM scalability case studies could be a better source of data for further studies.Additionally, a dynamic element was noticed where transitions occur between different approaches in the conceptual model.The adaptabilityfocused approaches for example help to feed new ideas and opportunity to the organisation where they can be fine-tuned using more efficiency-centred approaches.Why and when these transitions occur is not evident from this study but it is likely that it is connected to the maturity of the industry and the organisation's dynamic capabilities.In emerging industries, which the circular economy is for the fashion retail value chain, an organisation likely need to explore many avenues for scalability before they can hone in on any particular strategy.An interesting future area to explore is thus how business model scalability relates to a CBM's dynamic capabilities.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.PRISMA flow diagram of the study selection process.

Fig. 3 .
Fig. 3. Final data structure of the analysis of selected papers.

Fig. 4 .
Fig. 4. Example of coding of strategies from the activity area "increasing the volume of used apparel resold".

Fig. 5 .
Fig. 5. Conceptual model representing two thematic orientations leading to four strategic approaches.

Table 1
Journals of publication and primary research methods.
Journal Journal of Cleaner Production Journal of Business Research Journal of Business Strategy Sustainability Technology Innovation Management Review Corporate Governance (Bingley) MIT Sloan Management Review Others (one article each) Method Case study, multiple cases Case study, single case Interview/focus group Literature review Survey Modelling and simulation Historical narrative Other (viewpoint papers, unstructured reviews etc.)