The
 UN
 Sustainable Development Goals as a North Star: How an intermediary network makes, takes, and retrofits the meaning of the Sustainable Development Goals

In this paper, we investigate how a network of informal intermediaries – including international organizations, consultancies, business alliances, and standard setters – has contributed to the persistence of the universalistic meaning of the UN Sustainable Development Goal (SDGs). Based on our analysis of 26 interviews and 121 online resources produced by the 22 most prominent intermediaries, we find that SDG diffusion is distinct from linear depictions, such as the regulator-intermediarytarget model. This is because the intermediary network acts via three dynamic mechanisms that lend to an inclusive meaning of the goals; the core intermediaries lead efforts to make the perspective one that can accommodate a range of different audiences and activities, then intermediaries who subsequently join the network accept that broad perspective. Concomitant to their making or taking of the perspective, each intermediary individually works to retrofit the SDGs onto their unique tools and activities and to create their spot within the network. The combination of perspective making and taking, and retrofitting, propels the persistence of the SDGs as a “North Star” rather than a more specific blueprint for companies.


Introduction
The United Nations' (UN) Sustainable Development Goals (SDGs) are a set of 17 objectives for governments, businesses, and society-at-large to achieve by 2030. In this paper, we argue that the nature of the intermediary network involved in the diffusion of the SDGs into the business context has contributed to the persistence of its broad meaning, rather than developing more specified indicators. We use Howells' (2006, p. 720) conceptualization of an intermediary as any organization or body that acts as an agent or broker in any aspect of a diffusion process between two or more parties. Through our empirical analysis of the mechanisms underpinning the interactions of the SDGs' intermediary network, we strive to advance existing analytical tools for examining diffusion via intermediary networks and, relatedly, the global diffusion of objects. To do so, we combine and extend stateof-the-art literature on diffusion (Ansari et al. 2010;Klingler-Vidra & Schleifer 2014;Kaplan & Kinderman 2020), intermediaries (Scott 2003;Abbott et al. 2017;Brès et al. 2019), and social networks (Finnemore & Sikkink 1998;Seabrooke & Tsingou 2014).
By bringing network dynamics into the study of intermediaries (Abbott et al. 2017), we align our analytical outlook with the "relational turn" in social science (Emirbayer 1997). We first reveal how intermediaries taking up the responsibility for SDG diffusionsuch as Global Reporting Initiative (GRI) who have become a de facto standard-setter for sustainability reportingare unofficial and unformalized (Brès et al. 2019, p. 135); put differently, they give themselves authority to diffuse the SDGs and define their individual roles through processes of role appropriation (see Kourula et al. 2019). By pairing this noncontracted characteristic of informal intermediaries with a more relational understanding, we then show how the dynamics of the intermediary network helps to transmit the perspective of the SDGs as being universally applicable for business strategy, reporting, and evaluation.
Building upon insights garnered from 26 interviews and an analysis of 121 online resources by the 22 key intermediaries, we argue that intermediary networks undergo processes of perspective making and taking among one another, while individually retrofitting the prevalent perspective. By creating a broad, shared understanding of the diffusion object, the network of intermediaries helps its members to navigate their contrasting needs for collaboration and differentiation while building the necessary critical mass in uptake. As a result, the meaning of the SDGs is understood as a guiding principle for work on sustainability; or, as several of our interviewees put it, the SDGs serve as a guiding "North Star," rather than a specific blueprint. As a North Star, the SDGs enable a range of intermediaries to participate in the diffusion process, which confers broad consensus-building, while also allowing individual intermediaries to fit the SDGs with their preexisting organizational identities and practices.
By answering the research question: "how do dynamics within intermediary networks shape the meaning of the objects that they help diffuse, particularly the SDGs?," our article makes three primary theoretical contributions. First, we bring novel insights into the functioning of informal and unofficial intermediaries in the global landscape. These types of intermediaries are understudied, even though their theoretical and empirical significance can be ascertained by looking at their increasing numbers, influence, and transnational reach (see Brès et al. 2019). Our analysis helps to bring the conceptualization and study of informal intermediaries into mainstream literature on diffusion and regulatory intermediaries.
Second, we develop analytical tools for exploring the role of intermediary networks in diffusion processes. While notable studies engage with the role of consultants in such contexts (particularly Seabrooke & Sending 2019), there is a need to pay further attention to the ways in which intermediary's individual practices interact with network dynamics, and how the need for individual distinction interacts with, and is at odds with, the network's desire for shared consensus. By conceptualizing causal mechanisms for how intermediary networks interact, we both help to advance future studies of intermediary networks and to further mainstream relational approaches in diffusion literature.
Our third contribution is advancing the study of the SDGs in diffusion research. In doing so, we offer novel insights into the SDGs as a diffusion object, particularly how they operate as an overarching set of goals. By bringing the SDGs into diffusion literature, we contribute to scholarly tools for studying these types of global norms and objects (Sommerer & Tallberg 2019). In addition, we strive to advance analytical approaches for research on the empirical area of corporate social responsibility (CSR) and the SDGs (van Zanten & van Tulder 2018; van den Broek 2020).
The article is structured as follows. In the next section, we develop an analytical approach for conceptualizing the role of intermediary networks. The following section then discusses our empirical case (SDGs) and details our methodological approach. In section 4, we draw on our primary interviews and analysis of intermediaries' online resources to show how intermediaries gave and transmitted meaning to the SDGs through perspective making and taking, while individually retrofitting their existing practices, and in so doing, carving out their distinct functional roles within the network. The final section discusses the importance of relational approaches and identifies avenues for further research.

Theoretical framework: Diffusion, intermediaries, and network
Our analytical approach extends political economy literature on the role of intermediaries in international diffusion processes. Diffusion is widely understood as the spreading of an objectbe it policies, ideas, norms, or practicesfrom one actor to another; this oftentimes implies from one territory, typically a state, to another (Shipan & Volden 2008), although it has also been studied as diffusion among international organizations (Sommerer & Tallberg 2019). In diffusion processes, intermediaries act as crucial mediators and collaborators by providing advice and expertise through various mediating activities (Brès et al. 2019, p. 129). The role of intermediaries in the diffusion process has been well-theorized in the context of regulation. Most notably, the work of Abbott et al. (2017) on the regulator-intermediary-target (RIT) framework draws on the model of orchestrationwith a regulator enlisting the assistance of an intermediary (Abbott et al. 2015). In their RIT approach, Abbott et al. (2017) explain that regulatory intermediaries operationalize rules by facilitating the flow of information from (global) regulators to (local) implementations and vice versa.
This RIT framework has been particularly prevalent in research on the diffusion of CSR, which led to the 2019 special issue in Regulation & Governance on the formal and informal roles of regulatory intermediaries in transnational multistakeholder regulation. This special issue predominantly echoes the implicit presumption of the RIT-model that diffusion happens on an uncoordinated basis, with two important exceptions. First, Fransen and LeBaron (2019) note that consultancies and NGOs collaborate in shaping governance agendas; however, they do not specify how or to what extent these interactions are able to set the agenda. Making the case for an approach that takes intermediary interaction into account, Kourula et al. (2019, pp. 151-153) argue that intermediaries cannot be understood in isolation as they operate, and interact, within a dynamic complex environment in which intermediaries are constantly influencing the roles that other intermediaries perform.
We pick up on this by focusing on the "I" in the RIT model; we focus on diffusion processes in which informal intermediaries interact in a network, rather than operate in silos, or through formal roles. Acknowledging the dynamic and complex nature of diffusion implies stepping away from linear-depicted processes in which the regulator's object is transmitted via individual intermediaries to a target. We conceptualize and empirically test how a network of intermediaries, who each adapt the object for their own purpose, work in concert with one another. In doing so, we extend analytical tools for exploring ways in which intermediaries come together to give meaning to diffusion objects, beforeand whileobjects are transmitted to end users. We refer to this as network diffusion.
Our approach helps to extend existing research on the mechanisms of consensus-building in professional networks (Finnemore & Sikkink 1998;Seabrooke & Tsingou 2014) and epistemic communities (Haas 1992). Research on professional networks examines how these networks transmit and assign value to objects. Networks are succinctly conceptualized by Hafner-Burton et al. (2009, p. 560) as "sets of relations that form structures, which in turn may constrain and enable agents." In transnational environments, research has revealed how networks can foster knowledge flows that "create far more complex and decentralized, two-ways networks of exchange" (Welch & Hao 2013, p. 234). Intermediaries co-construct networks, in which they collectively foster multipartner exchanges that codify knowledge in the form of consensus building.
The significance of intermediary networks in the diffusion process derives from their translation of the objects they diffuse. Individual intermediaries are not neutral transmitters; they imbue meaning on to the diffusion objects. Scott (2003, p. 884) asserts that intermediaries take objects that "are abstracted, named, codified, and converted into models." In this process of theorization, or abstraction to model, intermediaries necessarily insert their own understanding and valuation of the object that is being diffused. Rather than being passive diffusers of objects, intermediaries necessarily choose which aspects of the object they focus on, which values they instill and so forth, effectively assigning distinct meaning to the diffused object (Scott 2003;Ansari et al. 2010).
An insightful distinction can be made with respect to the specificity of the object and how varying degrees of specificity affect the extent to which adaptation occurs in the diffusion process. Weyland (2007) distinguishes between principles and models with the former being "general and vague on details," whereas the latter are "concrete, specific blueprints." In the diffusion process, a higher degree of variation can be expected when principles diffuse because they offer larger room for interpretation (Weyland 2007, p. 18). Others use the language of "boundary objects" to refer to diffusion objects that are "both plastic enough to adapt to local needs and constraints of the several parties employing them, yet robust enough to maintain a common identity across sites" (Star & Griesemer 1989, p. 393).
Elsewhere in the diffusion literature, scholars have demonstrated that intermediaries do not simply adopt "dominant discourses but filter them through their own cultural lenses so as to produce something new and hybridized" (Hobson & Seabrooke 2007, p. 17). As each intermediary purposefully adapts a model for local use, a process called transmutation (Yeo & Painter 2011) or constitutive localization (Acharya 2010) occurs, in which objects are reformulated in a way that fitsand benefitsthe local context. In adapting objects for the local context, Klingler-Vidra (2018) emphasizes contextual rationality, in which normative environments dictate the form and extent to which objects are adapted for use. These localization processes do not only occur geographically but at an organizational level as well. Diffusion, in this context, is shaped by organizational fit that depends on the "degree to which the characteristics of a practice are consistent with the (perceived) needs, objectives, and structure of an adopting organization" (Ansari et al. 2010, p. 68). Intermediaries are able to create, present, and sustain an organizational identity that is congruent with the diffusion object (Brown 2015). Rhetorically, intermediaries will start by reinterpreting their past and present practices through this new lens, which is referred to as "narrative fidelity" (van den Broek 2020). Thus, it is established in the literature that each intermediary in the diffusion process is expected to rationalize some degree of fit through their own normative or cultural lens.
Each intermediary brings their own identity, which individually impacts the meaning of the diffusion object. Consequently, when intermediaries interact in a network, there is pressure toward vague understandings, in order to accommodate these multiple, and distinct, identities. In this way, intermediary networks are not necessarily battlegrounds where individual intermediaries fight to assert their own preferred meaning on the network. Instead, led by core intermediaries who prioritize consensus, or act as "perspective makers" (Boland & Tenkasi 1995), intermediary networks act to establish an inclusive agreement on the meaning of what they diffuse. The ecology of the network, thus, gives meaning to the diffusion object (Lainer-Vos 2013). Core intermediaries take a broad perspective that allows room for individual intermediaries to articulate the consensus for their local use. This space for localization allows intermediaries to maintain legitimacy as experts by offering their own strategy formula, reporting guidelines, and other services. Haack et al. (2012, p. 821) speak in this context of consensual narratives, understanding a consensus as the stabilization, or institutionalization, of a dynamically developed set of narrative elements.
This tendency is mirrored within professional networks in which actors exhibit group think due to "a strong incentive to maintain their position in the network by excluding others who do not agree with their understanding of issues" (Seabrooke & Henriksen 2017, p. 10). As a result, intermediaries work to link themselves with groups who share a similar institutional agenda, and then accept or "take," this established perspective (Boland & Tenkasi 1995). By forming and exploiting these ecologies of shared interests, they collaborate on the agendas of others, which creates mutual dependency (Faulconbridge & Muzio 2017, p. 230). In this process, "leader" and "follower" roles are established and maintained (Brown 2015, p. 24). This research, then, suggests that intermediary networks will seek consensus around an accommodative meaning of the object they diffuse.
Returning to Weyland's (2007) lexicon on the specificity of diffusion objects, network pressures toward an inclusive consensus informs our expectation that the SDGs will diffuse as a (loosely defined) principle, and not a (highly specified) blueprint, offering room for how they are put into practice. In this sense, diffusing the SDGs as a principle serves as strategic ambiguity, which can foster the co-existence of pluralistic viewpoints (Eisenberg 1984;Star & Griesemer 1989) and, thus, accommodate the involvement of multiple intermediaries in the network. We expect that the network dynamics engender the establishment and persistence of a broad understanding of the goals, so that each intermediary has room to adapt the goals to suit their individual circumstances.
We draw on these theoretical tools to help account for the ways in which intermediaries engage in the network vis-à-vis the diffusion object. Synthesizing these insights for the context of network diffusion, we contend that there are three mechanisms at work. First, there is the "perspective making" mechanism that fosters the development of a wide consensus to achieve fit for all members. Second, a "perspective taking" mechanism is instigated, in which new intermediaries, accept rather than contest the established consensus. Finally, while taking on the perspective, the "retrofitting" mechanism is present, as each network intermediary translates the consensus to fit with their own unique contribution and, based on this, carve out their distinct functional roles within the network.

Empirical setting: SDG diffusion
Our empirical focus is on the SDGs, which were officially adopted in December 2015 by the UN member states after an extensive multistakeholder consultation process (Fukuda-Parr 2016). Through this multistakeholder approach, the SDGs were the first development agenda that included business from the start, as is illustrated by Unilever's active role in the making of the SDG agenda (van Zanten & van Tulder 2018, p. 225). The SDGs have diffused across multiple channels; in the corporate context, this process is led by the UN's arm responsible for mobilizing companies to advance broader UN principles and norms, the Global Compact (hereafter "UNGC"). Then, informally, intermediaries, states, corporations, consultancies, alliances, and NGOs work as intermediaries to further distil the SDGs for corporate use. The diffusion of the SDGs provides an opportunity to examine the interplay between global agenda setting and a network of unformalized intermediaries. Furthermore, as Weyland (2007) explains, because of their lack of specificity, principles are expected to be more prone to different interpretations as they are being diffused (see Star & Griesemer 1989 on "boundary objects"). We unpack how the network of intermediaries' dynamics contributes to the maintenance the objects' broad meaning, rather than turning to a more precise understanding.

Data-collection and analysis
Step 1: Establishing SDG intermediaries and collecting resources. Our unit of analysis are the intermediaries who help to diffuse an international object (SDGs) into digestible meaning to inform business and investment activities around the world. The intermediaries include international organizations, the Big Four accounting firms, boutique sustainability consultancies, standard setters, and reporting platforms. 1 Logically, the first step of this study was to identify who the relevant intermediaries are that contribute to the translation of the SDGs into the business context.
To determine the population of key intermediaries, a natural starting point is UNGC. Specifically, we began by hand-collecting all UNGC's online resources on the SDGs since their launch in December 2015 through to January 2020. Through this compilation, we identified all intermediaries who at any point collaborated with UNGC. We then gathered all available online resources for the SDG activities of these intermediaries, including their SDG guidance documents, platforms, reports, and standards. We included only those that were explicitly SDG themed, meaning that the SDGs were specified in the document title, abstract or introduction.
To ensure we had captured the intermediaries that are embedded in the network, as opposed to those who were simply involved in one-off collaborations, we imposed a threshold of three entry points. This filter led to us cutting five organizations from the study. 2 However, two of these intermediaries (IMP and B Lab) were ultimately included in the analysis on account of multiple mentions of them as key collaborators during our interviews. They only had two and one co-authored online resources, respectively. In total, we identified 22 intermediaries.
To analyze the interaction of these intermediaries, we coded each online resource with the following: the official title, type, year of publication, names of the main author(s), collaborating organization(s), and technical partner(s). In total, we coded 121 online resources. Table 1 presents a summary of the types of online resources, giving descriptions and examples.
We verified the comprehensiveness of this intermediary list by asking each interviewee (see step 2), starting with UNGC, to identify key, standing collaborations with other intermediaries. Consequently, through the online resources, we could ascertain the form and extent of collaborations between intermediaries in terms of their shared participation in an Action Platform, or through their co-authoring of Data and Rankings, Guidance, or Reports on the SDGs.
Step 2 -Interviewing professionals within the intermediaries. As the population of 22 intermediaries is relatively small, we set out to interview SDG specialists within all intermediaries. Access to such specialists is notoriously difficult, and amplified by challenges of geographic locations, time, and gatekeepers, such as secretaries and personal assistants (Harrington 2017, pp. 43-44). We started by using our own professional networks and, in addition, submerged ourselves in the intermediaries' networks by attending 12 key events and meetings, including the SDG local goals event in London in June 2019. 3 In addition to sourcing interviewees, the events and meetings also provided useful content; as we listened to keynote speeches and panels on the SDGs, we were able to observe what the various intermediaries said and how they interacted with one another. Hence, we used participatory observations to inform our studies and further triangulate the data. In total, we were able to conduct 26 interviews with SDG specialists from 20 intermediaries, meaning we interviewed at least one representative from nearly every intermediary. Unfortunately, we were unable to interview two of the intermediaries (UN Environment Program and DNV GL).
Step 3 -Data analysis. We have two sets of data that we analyze: (i) online resources and (ii) interviews and event observations. The main function of the online resources is to provide a description of the network, first by identifying the intermediary members, and then giving insights into the ways the intermediaries interact with one another. The main function of the data gathered through interviews and event observations, on the other hand, is to make inferences around how the intermediaries act both individually and collectively, and through which mechanisms those network activities shape the meaning of the SDGs. In particular, interviews and event observations helped to explore the ways and extent to which the specificity of the meaning of the SDGs is affected. More specifically, participant-observations allowed us to observe intermediaries, and in particular their interactions, within their own environment, whereas interviews allowed us to better understand their subjective meaningmaking.
Furthermore, to discern patterns and uncover interactions within the intermediary network, we employ basic social network analysis, which allows us to assess the degree and closeness centrality between intermediaries (Chalmers & Young 2020, p. 67). In particular, we use the software Gephi to visualize how intermediaries collaborate on the co-production of online resources. The Gephi software codes the degree in terms of the volume of online resources produced by each intermediary through the size of the node and through color visualizes the closeness centrality across intermediaries based upon the volume of collaborations (red indicates working with many collaborators, while blue means few). It draws particular attention to the active, core positionsindicated by their large size and red colorof the early, core members -UNGC and GRIwho remained a driving force behind the broad meaning of the SDGs throughout our study. The intermediaries who collaborate less often, who we described as "perspective takers," are represented again by their color (blue is least close) and small size. Deloitte, for instance, only collaborated with one other intermediary, and so is bright blue, whereas UNGC is bright red as it has the largest number of collaborators. The Gephi in Figure 1, below, illustrates the both degree and closeness of each of the intermediarieswho are indicated as a nodein the network.
Thus, Figure 1 underscores the central position of UNGC, along with GRI and PRI, in leading the network. UNGC's and GRI's degree and closeness centrality with other intermediariesindicated by the large size and red color of their nodeshas been important to the persistence of the broad meaning of the SDGs; they have been core in making the North Star perspective.

Empirics
Our analysis of 121 online resources (Action platforms, Consultancy, Data and Rankings, Guidance, and Reports) helped to construct a picture of the intermediaries involved in the diffusion of the SDGs into the business context. Combined with our specialist interviews, we do not only know what the intermediaries produced, but also how and why they collaborated.
Based upon our analysis, we conceive of three SDG diffusion phases thus far: 2015-2016 is characterized by retrofitting and establishing the North Star consensus; 2017 marked the advance of collaborations and the ascent of the North Star; 2018 to early-2020 is typified by deepening collaborations and ensuring the breadth of the SDGs. Figure 2 illustrates the three phases, overlaid on the trajectory of the intermediary network, in terms of number of activitiesi.e., action platform, guidance, reports, consultancy and rankingsand members over the five-year period.
As illustrated in Figure 2, the phases follow the S-shaped diffusion curve in terms of the trend shape and proportions of the timing of adopters (see Rogers 1962); Phase I includes innovators and early adopters (40 percent of intermediaries), Phase II the early and late majority (46 percent of intermediaries) and Phase III the laggards  (14 percent of intermediaries). Table 2 summarizes the consecutive phases by explaining the network configuration as well as how the three mechanismsperspective making, perspective taking and retrofittingfeature in each of the three phases.

Phase I (2015-2016): Retrofitting and the making of the North Star perspective
In 2015, there were few intermediaries ("innovators and early adopters"), which were all involved in providing input into the SDG framework through the multistakeholder advisory committee. The first attempt by intermediaries to give meaning to the SDGs in the business context was the "SDG Compass," as UNGC, GRI and WBCSD ("World Business Council for Sustainable Development ") shared a sense of urgency in the need to come together to publish business guidance when the SDGs launched. 4 One interviewee explained that "it [was] critical to be involved [with the SDG Compass] as the 2030 agenda is a consensus on the key issues we need to work on." 5 This collaboration marked the start of the network of informal SDG intermediaries. The first three intermediaries had overlapping corporate membership, which propelled some alignment of their interests. However, from the outset they aimed to maintain their individual contributions, as an interviewee explained that: We had a partnership as we are the three biggest, most prominent NGOs on business sustainability. Everyone came with their own expertise, perspectives and skill set; UNGC with their UN connection, the WBCSD with its members and firms and GRI with their reporting. 6 Given the three intermediaries' desire for distinct roles, they needed to work to align the "SDG Compass" to the principles and standards of the different organizations. It had to be broad enough to accommodate the materiality rationale underlying GRI standards as well as the 10 principles that underline UNGC's work, while keeping the WBCSD member-firms satisfied.
This desirability of making a broad meaning was justified by the common understanding that the SDGs do not encompass new issues on sustainable development, they instead provide a new framework which builds on

Mechanisms at work
Perspective makinga core group of intermediaries comes to an agreement about the meaning of the SDGs in the business context.
Perspective takingby joining the network, the majority of intermediaries adopts the meaning of the SDGs as has been stipulated by the early adopters.
Perspective takingthe latest intermediaries to join the network adopt both the meaning of the SDGs as well as reiterate the established discourse (of the North Star). Retrofittingeach intermediary finds fit with their organizational identity; they, in turn, steer the meaning of the SDGs to be compatible with their identity and establish their legitimacy to co-make the emerging perspective.
Retrofittingnew intermediaries joining the network seek coherence between the established meaning of the SDGs and their organizational identity; and, based on this, articulate their functionally distinctive roles within the network.
Retrofittingsimilar to Phase II, and also, as most functional roles have been fulfilled, the articulation of the functional roles become more specific. the same "business case." 7 Additionally, the "SDG Compass" had to keep in line with earlier efforts in order not to disregard past and present work. As a result, the aim was to make the SDGs widely accessible for companies and prevent the requisite reporting from becoming too specific. Keeping the "SDGs accessible, easy and generic, served the broader goal of laying the foundational work of all efforts to translate the SDGs to the business context thereafter." 8 The broad perspective underpinning the core intermediaries' understanding of the SDGs in 2015, however, was not the only option available. Different, and more specific, perspectives "crossed the table, such as circularity and Key Performance Indicators, but they were found to be too radical, specific, and conflictual." 9 Efforts for advancing the SDGs as specific actions or expectations, as put forth by WBCSD, were rebuked. Peter Bakker, the CEO of WBCSD, expressed his disappointment that the "SDG Compass" was not transformative enough. In stark contrast, an interviewee at UNGC commented positively on the broad meaning relayed at the outset of the SDG Compass, The guide [SDG Compass] translates easily. It is easy to use and is applicable to every company. It doesn't provide detail; it just nudges companies in the right direction. 10 Two of the Big Four consultancies -KPMG and PwCentered the intermediary network in 2015 and performed a critical role in the early phase of shaping the open meaning of the SDGs. Both strived to make the SDGs relevant for all companies and, hence, their early inclusion in the network contributed to making the SDGs widely applicable. PwC undertook a survey in order to pinpoint what business felt were the important SDG and CSR issues in the coming years. This helped the consultancy to carve out their role in the intermediary network, as a specialist in formulating the SDG business case: We contributed a guide to make the business case, that translate the government SDG agenda to business, as a lot is expected from them [business]. 11 In the context of perspective making, there was a push for more specificity from the other Big Four intermediary, KPMG. They were advocating for hands-on relevance to translate the SDGs for a wide range of businesses by producing industry-specific guidance. To this end, KPMG reached out to UNGC directly and proposed to contribute by developing industry matrices that translated the SDGs into practical corporate action. They saw their opportunity in making the SDGs more practical while still relevant to all businesses. 12 However, the momentum was in favor of the broader perspective, with intermediaries like KPMG already retrofitting the broad SDGs into reporting matrices in their own practice. 13 Further fuel for the broad meaning came in 2016, as new intermediaries increased the range of business actors for whom the SDGs were deemed relevant. As PRI ("Principles for Responsible Investment") entered the network, for instance, investors were brought into the frame. While each new intermediary (like PRI) brought their own focus, they helped make the broad perspective. For instance, the PRI promptly repeated the chorus of the general aims in co-producing a guidance titled "Transforming Our World Through Investment" with UNGC and "What Do the UN Sustainable Development Goals Mean for Investors?" through collaborations with UNGC, UNEP, and PwC.
While bringing investors into the SDG agenda, PRI worked to translate the broad meaning into their own investment context. PRI's SDG Investment Case, for example, makes the explicit link between the PRI requirements and the SDGs, saying that their principles for responsible investment strive to "better align investors with broader objectives of society," and that the SDGs act as welcome guidance as to what their "broader objectives of society" are (p. 11). The SDGs, in this way, were retrofitted to substantiate what PRI had already been doing, by making the "objectives of society" a more tangible construct. PRI, as others would do after them, further propelled the broad consensus and worked to retrofit the SDGs to their particular context.
The UNDP ("United Nations Development Program") also retrofitted the SDGs onto their work and articulated their distinctive functional role in the network, focusing on impact measurement and the context of developing countries. They initially did this by collaborating with GRI to produce guidance on how businesses can measure their impact and accelerate the SDGs through sustainability reporting. The guidance strived to start "mapping existing principle sets" and "built on what already exists," underlying the perspective that the SDGs did not present new issues, but rather an opportunity "to speak the same language." 14 UNDP's unique value proposition focused on its geographic focus, explaining that it is "difficult to allocate capital to emerging markets and very little rigor or transparency [exists] around reporting for SDGs." 15 Thus, UNDP's involvement in the intermediary network began with collaboration aimed at perspective making, and retrofitting the SDGs onto their impact measurement and reporting practices in emerging economies.
Around this time, the language that the SDGs provide a "North Star" entered the intermediaries' discourse. It was apparent for intermediaries that had already joined the network, as well as for intermediaries that engaged with the SDGs individually and were, thus, still placed outside of the network. For example, boutique consultancy SustainAbility, which started engaging with the SDGs by doing individual consultancy work, elucidated how, from their perspective, the SDGs made sense for businesses. As an interviewee asserted, SustainAbility accepted the established SDG perspective, and linked them with what they were already doing.
The SDGs are the North Star within our field. They were introduced with a big splash and were the topic of every conference and conversation. But in the end, they are really nothing new, they are just issues articulated in a new framework that makes sense to both business and society. 16 This is mirrored by EY, who in 2016 picked the SDGs up as part of their mega trend scan, though remaining outside the network until 2017.
The SDGs give you a great North Star. They are so comprehensive about what a better world looks like, that companies can say where they want to focus. It gives a sense of targets and ways to measure it. Then we help the company articulate its unique capabilities. 17 In sum, during this first phase, a small number of intermediaries initially shaped the perspective for what the SDGs would mean for business in a way that is inclusive, all-compassing and strives to accommodate all the key players.
As will become evident in the years that follow, this phase has been formative for establishing the core network, and relatedly, the making of the SDG perspective. These first intermediaries would continue to drive the SDG agenda, with intermediaries new to collaborating with network members coming to "take" or accept this broad perspective. Thus, the first phase is underscored as a crucial period in which an emerging consensus that the SDGs provide a broad frame for activities was cultivated. Intermediaries, even in this first stage of perspective making, also each worked individually to retrofit the SDGs to their previous sustainability efforts. This both guided the perspective making and justified their core position within the network. The perspective of the SDGs as a broad North Star allowed intermediaries to accommodate it to fit with a wide variety of applications, including investors in emerging economies. Table 3 shows how the mechanisms of retrofitting and perspective making manifested for every intermediary entering the network in Phase I.

Phase II (2017): Perspective taking, retrofitting and the ascent of the North Star
The second phase saw the further increase in intermediaries ("early and late majority" adopters) as well as the uptake of SDG action in general, which is evident from the increase in the number of online resources published during this year. As demonstrated in Figure 1, above, the volume of SDG resources more than doubled from 2016 to 2017, as more intermediaries began providing guidance documents, offering data and rankings, and coauthoring reports.
In our interviews, numerous intermediaries spoke of the central role of UNGC as convener and agenda-setter, and ultimately, driver of the intermediaries' pursuit of the broad consensus. They "really put a lot of resources in place to engage firms with the SDGs" 18 and, as a result, created an infrastructure for broad, inclusive SDG action. Speaking of the central role of UNGC in 2017, an interviewee at GlobeScan explained that, "UNGC is heavily involved in monitoring and implementing the SDGs." 19 By framing the SDGs as central to sustainability, they ensured their role as a "central authority figure." 20 The other core intermediary, GRI, was also cited as being instrumental in supporting these efforts; as a reporting platform, they have been extremely successful in centering the agenda around generic corporate reporting. PwC, for example, explained that "UNGC is central to making business better" and, as a result, they are leading the agenda, along with: GRI…[with] the most adopted set of standards. There's lots of guidance coming out on the goals, but it felt like those two organizations particularly were kind of the leading organizations. 21 The SDGs offer an opportunity to "build capacity and make firms use the GRI standards"; as such, their aim is to "promote to GRI standards" and work to "align the SDGs" with these standards. 22 The GRI, therefore, had a keen interest in keeping the SDGs broad and avoiding the development of new and more specific reporting tools.
Volans, a boutique consultancy, tried to challenge the broad perspective, in a way similar to what WBCSD had done earlier on. An interviewee at Volans explained that "we can only achieve the SDGs if we see them as a breakthrough mechanism, not just as incremental steps, or as we say, change as usual." 23 Their founder, and creator of the triple-bottom line, John Elkinton, gave a critical speech on the SDGs at an UNGC event. 24 Afterwards, the CEO of UNGC reached out and said "I think your challenge is very relevant and we want to work together." 25 UNGC subsequently hired Volans, with technical support from PA consultancy, to produce the guidance document "The Breakthrough Pitch." The difference between Volans' pressure for greater specificity and WBCSD's 2016 push is that this was initially framed as the only way forward, challenging the broader meaning of the SDGs. UNGC, as the de facto convener, worked to co-opt Volans into the network, by co-authoring a thought leadership piece in line with the broad agenda. In doing so, the North Star consensus was restored as Volans soon began to echo the network's language: SDGs are just framing for CSR and allow firms to somewhat align with the global agenda. The fundamentals of the sustainability agenda are still the same as in the 87 Brundtland report. The SDGs are not a new concept but just a new framework on the same agenda. 26 Both the IMP ("Impact Management Project") and the WBA ("World Benchmarking Alliance") were launched in 2017, and quickly after their formation, both joined the network. WBA began by developing benchmarks to assess companies' engagements with the SDGs and aimed to have these picked up by other civil society, governments, and investors. Even though they were established after the SDGs, they felt inclined to develop a new framework that the SDGs map onto. They "reshuffled the SDGs into seven systems lenses," which they based on "other frameworks, research, and consultations" by "consciously looking at what is out there." 27 A central part of their strategy was to foster collaborations, which includes work with intermediaries such as UNGC and GRI, as there: was an awareness that we can't achieve it on our own. The Alliance is our soundboard and shows whether what WBA does is relevant for them. 28 IMP similarly set out to align with other intermediaries on a common language for impact measurement. Though IMP had initially been working to standardize the measurement and reporting of impact investing without specific reference to the SDGs, they quickly acted to engage with the consensus-building work taking place in the context of the SDGs. Their dedication to this perspective came from their contention that "there is not a single truth or silver bullet. The standard is a social construct, so buy-in is everything." 29 Thus, the importance of collaboration and a shared perspective, and particularly the SDGs as a North Star, becomes even more explicit with these intermediaries joining the network.
The SDGs have become a globally agreed kind of North Star for at least the next decade for us to work toward. People will take their own way of doing, their own data, their own framework, and then bring that match to the SDGs language. 30 In 2017, UNGC, GRI, PRI, PwC, and SustainAbility further collaborated in the action platform, "Reporting on the SDGs." The first guidance published as a result of the action platform was the "Business Reporting on the SDGs: An Analysis of the Goals and Targets." This initiative aims to provide more concrete guidance on how business can make sense of the goals, 31 while remaining "simple" and "accessible" to all, 32 which illustrates the continuation of the adherence to the broad meaning of the SDGs in order to bring together, and keep relevance for, this range of intermediaries.
Interestingly, the WBCSD, who had pushed for specificity from the start, left this collaboration. Although the collaborators going their own way is claimed to have "just grew that way," 33 one is left to wonder whether the challenges brought up by their CEO earlier, about their desire for the SDGs to have more specific application, played a role, as is substantiated by the comment of their SDG lead that "the SDGs are so broad… We decided to really go for industries. There is a lot of communication on progress but not a lot of action and change." 34 This is to say that not all intermediaries continued with accepting the perspective that the SDGs should be understood as a broad, North Star; those who disagreed with this perspective, and instead favored a more specific perspective on the SDGs (WBCSD in particular), retreated from activity in the network. As such, those who remained were only those who continued to accept the broad meaning of the SDGs.
In sum, the second phase (2017) had seen an extensive growth in the number of intermediaries collaborating on the SDGs. Collaborations helped keep the broad consensus on what the SDGs mean for business, ensuring that every intermediary had a role to perform, and that their range of remits could fit under the SDG umbrella. Table 4 shows how the mechanisms of perspective taking and retrofitting play out for every intermediary entering the network in Phase II.
4.3. Phase III: (2018-January 2020): Consolidating the North Star consensus, deepening collaborations, and claiming distinctive functional roles The third phase further institutionalized the broad meaning of the SDGs as the North Star, which allowed the intermediaries to carve out their unique roles, both within and outside the network, and deepen collaborations between them. The year 2018 saw the focus on two broadly applicable business issues: reporting and finance. In this phase, only a few intermediaries are still joining the network ("laggards"). The core members, especially those active from 2015, continued to be central in shaping the perspective. The collaboration between UNGC, GRI, and PRI deepened as the action platform on "Reporting on the SDGs" continued and produced a new practical guide on "Integrating the Sustainable Development Goals into Corporate Reporting." This guidance was "meant to integrate the SDGs in existing reporting practices" 35 and, in GRI's case, to enable retrofitting to the GRI framework. 36 The newly established action platform "Financial Innovation for the SDGs" by UNGC, PRI, and UNEP further propagated the broad meaning of the SDGs across the finance-focused intermediaries. This inclusive perspective allowed intermediaries to advocate for mainstreaming SDG investments. Again, rather than developing a specific blueprint, the goal of the finance-focused platform is to "establish guiding principles and best practices for SDG-aligned corporate finance and investment opportunities." 37 Structured interactions, led by IMP and UNDP, further aimed at consensus. In 2018, the IMP established the "Structured Network," in which regular communication channels were established among the senior leadership of each intermediary as well as monthly calls held for those at the technical and implementation levels. The Structured Network strives to align frameworks and bring people together 38 ; as an interviewee asserted, it deliberately only includes "generalists" and is designed to facilitate "coming to consensus." 39 The UNDP, for its part, created the "SDG Impact" team in an effort to institutionalize their work in translating the SDGs for impact investments in emerging markets. Collaboration with other intermediaries has been central to their efforts. UNDP's SDG Impact team signed a memorandum of understanding with the IMP, which stipulated that both teams are embedded in the structured network. The SDG Impact team's impact specialist remarked that they strove to have buy-in from that group [IMP Structure Network] first. 40 Thus, collaboration around the broad consensus has been central to the activities of the SDG impact team.
UNGC continued to play a central role in convening the network activities that propelled the broad meaning of the SDGs. More precisely, they set up additional action platforms that each address specific goals and issues, such as "Decent Work in Global Supply Chain," "Health is Everyone's Business," "Pathways to Low-Carbon and Resilient Development," "Sustainable Ocean Business," and "Water Security through Stewardship." They explained that collaborating on specific issues is "critical to achieving the SDGs." 41 While the goal-focus does provide specificity, it is under the guise of broad applicability. For example, in the action platform on supply chains, the most widely reported challenge is "achieving transparency" and companies are advised that "technology enables transparency and oversight" and "collaboration is essential" as well as "building the business case." 42 Similar rhetoric is used in online resources coming from the other action platforms. Thus, the narrative underlines the aim that the SDGs must remain broad so that everyone has a seat at the table.
The year 2019 marks the further consolidation of the North Star consensus. Specific projects and guidance documents are repeated on a yearly basis, and the same set of intermediaries initiated more and different types of collaborations aimed at the SDGs' broad applicability. For example, Accenture and UNGC began to co-produce their CEO study centered around the SDGs on a yearly basis. UNGC and GRI are still at the core; they build upon their previous efforts to keep the SDGs broadly applicable by emphasizing that each intermediary has a role to play in the "decade of action." To illustrate, their joint "Reporting on the SDG" action platform continues to be operative, although it now aims to "build capacity among pioneers, showcase how pioneering firms reported and disseminate further these insights to drive action." 43 The late entrance of B Lab provides an interesting example of how latecomers must accept the broad perspective, and also, negotiate their spot within the network by specifying their own value-add. Rapid international growth of the use of the SDGs led B Lab to join the network by collaborating with UNGC.
We [B Lab and UNGC] have been looking for a collaboration opportunity and share the market space. 44 In 2019, other key intermediaries, including UNDP and GRI, mirrored a strategy of "getting everyone on board," in a way similar to that taken by IMP and WBA. The mechanism of retrofitting was imperative to B Lab's maintaining of its certification expertise within the context of the SDGs. An interviewee explained that most of B Lab's work predates the SDGs and, as a result, they had to retrofit the goals to their tools: The SDGs arose in this space, and we needed to think about how they are related to our approach. Our impact assessment tool was not created for the SDGs, it pre-dates the SDGs. We needed to think about "retro-fitting," to see how our approach can be mapped onto the SDGs and identify where there are differences. 45 Their retrofitting is similar to how intermediaries who joined before them wanted to both be part of the SDGs, and concomitantly, not lose their unique value proposition. In explaining their understanding of this tension, an interviewee at UNGC explained that they knew that "no one wants to give up their identity, so everybody [intermediaries] finds their niche SDG specialty and looks at where they can establish authority." 46 Echoing the desire to join the consensus and still be unique, our B Lab interviewee asserted that the SDGs are an essential North Star, and that B Lab brings value by offering "an education management platform" and, as such, are "different than GRI as that is just reporting." 47 In sum, during the third phase (2018-January 2020) we observed the persistence and consolidation of the North Star perspective along with individual intermediary's retrofitting, as well as efforts to formalize interaction across the intermediary network, notably in the context of the IMP-led structured network. As such, the SDGs were now firmly entrenched as a shared aim that stimulates collaboration. The early, core members of the networkparticularly UNGC and GRIcontinued to make the SDG agenda broadly inclusive. In 2019, the need for a widely-held consensus became further embedded in the language of all intermediaries and this mantra was Established the SDG Action Manager, in collaboration with UNGC adopted by newcomers. Table 5 shows how the mechanisms of perspective taking and retrofitting play out for the intermediaries entering in the third phase. In sum, over the course of the five-year period, intermediaries joined and took to embracing the need for the SDGs to be understood as broad business goals, and simultaneously, worked to retrofit the SDGs as a North Star on their existing practices and offerings. In the few instances where intermediaries were critical, or pushed for greater specificity in translating the goals into specific reporting practices, these intermediaries either left the network activities (WBCSD) or were co-opted to be on board with the broad meaning (Volans).

Discussion and conclusion
Since its launch in 2015, the intermediaries have purposefully madeand maintainedthe perspective of the SDGs as a North Star. The reference to the North Star serves as a powerful metaphor; it indicates a fixed reference point around which all others rotate. The use of such a metaphor "implies a way of thinking and a way of seeing" (Morgan 1986, p. 12). In this case, associating the SDGs with the North Star endows the diffusion object with a broad and inclusive meaning rather than stipulating specific routes of operationalization. Metaphors assert similarities and overlapping between two objects (the SDGs and the North Star) and divert attention away from the dissimilarities (Oswick et al. 2002). By studying the GRI's early emphasize on the similarity between sustainability reporting and financial reporting, Etzion and Ferraro (2010), for example, showed how analogies and metaphors "map a novel institution to the natural order of things" (p. 1094). This study highlights the importance of such metaphors in shaping the diffusion object, and hence, delimiting diffusion pathways.
We found three primary mechanisms underpinning this North Star phenomenon. First, there is a commitment to framing the SDGs as broadly applicable by the core intermediaries in the network. The mechanism of perspective making shows how intermediaries collectively imbue meaning on the object they diffuse, building on earlier work by Scott (2003) and Ansari et al. (2010). From the start in 2015, and then throughout, through numerous collaborations the core intermediaries reassert the need for an accommodating perspective. The SDGs, as a diffusion object, successful remained a "boundary object" or "principle" which allowed for productive interpretive flexibility and collaboration between intermediaries (see Star & Griesemer 1989;Weyland 2007;Lainer-Vos 2013). In particular, UNGC, from the outset, took on the task of ensuring universality for the SDGs. Hence, one of the main contributions of this study is the revealing of how diffusion objects are neither static nor objective as their meaning is collectively constructed.
At the helm of intermediary network efforts, the core intermediaries have worked to ensure that each intermediary that joins the network needs to implicitlyand sometimes, explicitlytake the established perspective, which propels the wide social acceptance of the SDGs as a North Star. Intermediaries join the network through collaborations as co-authors of reports or co-conveners of action platforms, often with UNGC and GRI. Through these interactions, the chorus builds that the SDGs serve as an essential North Star and should not be understood in too specific of terms. This finding reinforces earlier work by Kourula et al. (2019, pp. 151-153) who show how intermediaries are constantly influencing the roles that other intermediaries perform. As such, this study reveals the leading role of early adopters in both setting the conditions for the diffusion object and also shaping how later adopters engage.
Joining the intermediary network and taking the North Star perspective is acceptable to intermediaries because it affords them the latitude to carve out their own space. A shared understanding that requires specific action would conflict with the practices of some members of the group; so, either those intermediaries would leave the networkas WBCSD didor would be co-opted into pursuing the universalistic perspectiveas happened with Volansshowing the incentive to exclude those who do not agree with this shared perspective (see Seabrooke & Henriksen 2017). The generalist and universalistic perspective of the SDGs provides flexibility for all intermediaries to stipulate their own SDG actions underneath the broader SDG umbrella (see work on "recursivity," Broome & Seabrooke 2020).
Thus, the SDGs as a North Star also means that each intermediary is not pitted against one another; they avoid competition by having the space to stake out their own, distinct contribution. As we saw with B Lab most recently, they established that their practice is one of "education" rather than "reporting," helping them to differentiate themselves from GRI. GRI, for their part, was also committed to the SDGs as a North Star from the start, rather than the SDGs comprising precise reporting; this helps ensure that their (GRI) reporting framework could continue to fit within SDG reporting. This study provides new insights into the challenges underpinning the constructing of diffusion objects as generalist and universalistic in order to accommodate a wide variety of actors and, ultimately, create a critical mass.
The third mechanism is that of retrofitting, in which each intermediary translates the SDGs to their preexisting practices. Retrofitting is possible given the breadth of the consensus, which means that localized translations do not challenge the consensus. Each intermediary works to retrofit the object to their raison d'être and, as a result, are able to create, present and sustain their organizational identity (see Brown 2015). They go to great lengths to prove that their practices are either consistent with the SDGs, or were designed explicitly for the pursuit of the SDGs. To illustrate, several intermediaries stressed that the SDGs contain longstanding issues and merely provide a new framework for viewing these issues. The logic of this mechanism is consistent with the process of achieving organizational fit (Ansari et al. 2010), or what is referred to as localization (Acharya 2010), transmutation (Yeo & Painter 2011), and adaptation (Klingler-Vidra 2018) in other scholarship.
By shedding new light on these retrofitting practices, this study speaks to the questions of how actors translate diffusion objects into their local context. Different from this existing literature, though, the retrofitting of the SDGs as a diffusion object takes place at the level of practices, such as reporting tools, rather than at the level of institutional or policy isomorphism. Retrofitting is also distinct from localization as the intermediary's existing framework is not subjected to substantive change. It is, instead, simply given new labels or frames, that fit the meaning of the diffusion object. As such, retrofitting has a clear temporal dimension; it aims to create a sense of continuity and stability between past, present, and future practices (referred to as "narrative fidelity" in van den Broek 2020), as was illustrated by the need of the "SDG Compass" to keep in line with earlier efforts of GRI, UNGC and WSBCD.
Collectively, the combination of perspective making and taking, and retrofitting mechanisms has ensured that, since their creation in 2015, the operationalization of the SDGs allows room for everyone. This process has allowed SDG intermediaries to proclaim their authority and define their distinctive functional roles, which sheds further light on the functioning of the unofficial and unformalized intermediaries (see Brès et al. 2019;Kourula et al. 2019). The initial interest in the SDGs' breadthasserted by member states and members of the multistakeholder consultations, including UNGC and GRIhas maintained by the extensive collaboration on action platforms, reports, guidance, and rankings by a growing number of intermediaries. The perspective underpinning each of these activities is that of universality, which allows for retrofitting by individual intermediaries. This study, thus, offers new insights into dynamic, networked diffusion processes, and challenges more linear understandings which overlook collaboration between intermediaries.
The SDGs as a North Star can accommodate industry-specific guidance, country-focused approaches, and also, intermediaries offering different strategy and reporting advice. As such, each intermediary can define their role (referred to as "role appropriation" in Kourula et al. 2019) based on their unique experience, expertise, and stakeholders. The conjunction of perspective making and taking, and retrofitting is in everyone's interest; the broad meaning of the SDGs does not challenge any existing practices nor does it demand fundamental change at the intermediary level. Consequently, the leniency of the diffusion object (see Fransen & Conzelmann 2015) allowed it to attract a critical mass of intermediaries during the first five years of the SDGs. Our finding advances the scholarly understanding of framework updating. Rather than being pitted against each other and discarding previous work, this article provides insight into how this North Star approach promotes cooperation and the prolongation of other frameworks.
In sum, while our empirical case is the SDGs, the finding that network diffusion is an enabler of broad diffusion objects is of wider relevance to the diffusion literature and substantiates calls for a relational turn in social science (Emirbayer 1997;Kourula et al. 2019). For such global goals, diffusion is less linear than is depicted in the RIT model (Abbott et al. 2017); multiple intermediaries may be active and, as a result, more attention needs to be given the relationship between them within diffusion studies. Additionally, we show that the study of intermediaries is not limited to formalized and official intermediaries that are contracted to implement regulation (see Brès et al. 2019, p. 135). On the contrary, diffusion scholarship can build on our insight that intermediaries may come in various forms and may diffuse a variety of objects. Furthermore, this study not only corroborates but also shows how intermediaries are not neutral transmitters (see Scott 2003), but express agency in the diffusion process by shaping the specificity of the diffusion object. In short, in a world where diffusion is "messy" in that it is happens dynamically across multiple levels and regions, we need further analytical tools for researching how numerous intermediaries interact, and what the nature of their interactions means to the diffusion object.

Acknowledgments
This research benefitted from support from a King's Together Seed Fund, from King's College London, which supported the organization of, and attendance at, SDG-themed events throughout 2019, including the Political Economy of Social Impact Measurement workshop in July 2019. Feedback provided at the "Microfoundations of CSR Paper Development Workshop" held at Cass Business School on the 27 September 2019, on an early draft, proved instructive in shaping the paper's analytical framing. The authors would like to thank the significant number of intermediary contacts who shared their time and insights, and invited us to industry events in order to complete this research. Finally, our thanks to Adam Chalmers for his constructive suggestions on a previous version.

Data availability statement
Research data are not shared.
Endnotes 1 See our Appendix S1 for a complete list of intermediaries. 2 The intermediaries cut from the analysis on account of them not having three or more online resources published are: ARM, Enel, ILO, UNICEF, and Vivid Economics.