When upstream suppliers drive traceability: A process study on blockchain adoption for sustainability

Abstract

Purpose

This paper aims to investigate the emergence of blockchain-enabled traceability in complex multi-tiered supply chains, focusing on the perspective of upstream suppliers. Blockchain technology receives attention for its potential to enable better traceability and thus sustainability risk management, yet there is limited empirical evidence on how actual implementation unfolds. We aim to understand how blockchain adoption unfolds in practice, particularly in critical mineral supply chains that are critical to the sustainability transition yet linked to severe environmental and human rights risks and to explore the role of traditionally non-focal firms in this process.

Design/methodology/approach

Adopting a process-based case study design, our research is grounded in data collected through participant observation (>12 months) within an upstream mining company, supplemented by interviews and document review. Our study employs the complex adaptive systems (CAS) lens and uses an abductive approach for data analysis.

Findings

In our case, blockchain-based traceability in the cobalt supply chain was co-constructed over time, fundamentally driven by a large upstream supplier but enabled through supply-chain-spanning collaboration with like-minded downstream actors and successive expansion into the opaque midstream, enabled through a stakeholder alliance forum and formalized in the blockchain. We find, however, that visibility, standards, trust and follow-up capacities need to exist in their own right, ideally prior to blockchain implementation.

Originality/value

Our paper provides empirical insights from an upstream (vs downstream) perspective and investigates blockchain’s implementation (vs potential) to complement and ground existing research. Further, we extend the CAS framework by emphasizing agency and visible horizon of traditionally non-focal firms.

Citation

Heldt, L. and Pikuleva, E. (2024), "When upstream suppliers drive traceability: A process study on blockchain adoption for sustainability", International Journal of Physical Distribution & Logistics Management, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/IJPDLM-01-2024-0022

1. Introduction

Blockchain technology has received attention for its potential to help improve environmental, human and labor rights issues in global supply chains (Marques et al., 2024; Saberi et al., 2019). Companies face pressure to gain visibility and control over sustainability issues far upstream, with increasing scrutiny from nongovernmental organizations but also investors, customers or regulators. New due diligence regulation is increasing this pressure, for instance, the European Union’s (EU) Corporate Sustainability Due Diligence Directive (European Commission, 2022), the EU Deforestation Regulation (European Commission, 2023a) that requires companies importing forest-risk commodities to provide traceability, geocoordinates and deforestation monitoring at farm-level but also the EU Battery Regulation (EUBR) (European Commission, 2023b) that requires digital product passports with full traceability data on critical minerals like cobalt, affecting business linked to electric vehicles and the green transition. With growing due diligence requirements, companies look toward technological solutions like blockchain.

While much has been said about blockchain’s potential contributions (and – to an extent – its risks) to sustainable supply chain management (SSCM) in general (Saberi et al., 2019), less is known about how its actual implementation unfolds (Sternberg et al., 2021). Similarly, pilot projects tend to be studied from the viewpoint of focal firms (Gligor et al., 2022; Hastig and Sodhi, 2020) and technology providers (Rogerson and Parry, 2020; Tuladhar et al., 2024), while perspectives of traditionally “non-focal” firms, such as midstream or upstream suppliers, essential to blockchain implementation, are underrepresented. This is striking considering that blockchain’s usefulness as an enabler of visibility (Rogerson and Parry, 2020) lies particularly in complex, dynamic and currently opaque supply chains where traceability remains a key challenge (Ahmed et al., 2022; Tuladhar et al., 2024). Likewise, these supply chain characteristics – complex, dynamic and opaque – mean that focal firms will struggle to single-handedly implement a new technology top-down without the involvement of multiple supplier tiers (Ahmed et al., 2022). In short, supply chains where blockchain could make the biggest difference appear to be supply chains where implementation could be most challenging (Marques et al., 2024). Whereas research on multi-tier SSCM is gaining attention (Mena et al., 2013; Tachizawa and Wong, 2014), studies that explicitly follow a multi-tier research design or highlight the perspective of lower-tier suppliers remain scarce (Gong et al., 2023; e.g. Jia et al., 2019), particularly concerning blockchain use. Likewise, scholars have called for diverse theoretical perspectives in SSCM that better capture the nuanced and multilevel nature of phenomena (Touboulic and Walker, 2015).

We therefore adopt complex adaptive systems (CAS) as our theoretical lens in order to elaborate the process and dynamics underlying the emergence of blockchain-enabled traceability in a complex critical mineral supply chain. The CAS lens specifically frames supply chains as resembling ecosystems, characterized by complex actor-constellations, self-organizing dynamics, external pressure and overall limited visibility and control by focal companies. Adopting this more complex, realistic view of today’s mineral supply chain, we explore how technology adoption and SSCM emerge when unilateral oversight is lacking. We aim to complement existing research on blockchain’s potential with a differentiated, empirics-based view of how implementation unfolds in practice and which pitfalls and dynamics emerge along the way. Simultaneously, the CAS lens has left upstream suppliers underresearched and we seek to elaborate theory in this direction. Our guiding research question is therefore as follows:

RQ.

How does a blockchain-based traceability system emerge in a complex supply chain?

To answer this, we adopt a process-based (Langley, 1999, 2007) case study design (Dubois and Gadde, 2002; Ketokivi and Choi, 2014) that is grounded in rich data collected through participant observation while embedded in an upstream company’s sustainability team and blockchain project, additional interviews and a document review.

Our contributions are threefold: First, by generating insights from the typically underresearched upstream supplier’s perspective, we contrast and complement existing multi-tier SSCM research that has predominantly focused on downstream firms’ perspectives. Second, by providing empirical evidence of blockchain implementation across an opaque multi-tiered supply chain, demonstrating how complexity and dynamics propel and/or challenge technology-enabled traceability, we complicate and ground prior blockchain research that has largely focused on expected opportunities and on pilot projects from the viewpoint of focal firms and technology providers. Third, by explicating the role of upstream suppliers through a set of propositions, we refine and elaborate prior research on supply chains as CAS which has implicitly overlooked the agency and influence of upstream actors, contributing to a more comprehensive, differentiated understanding.

The remainder is organized as follows: Section 2 reviews prior research on multi-tier SSCM and blockchain technology and introduces CAS to clarify our theoretical lens. Section 3 explains our research design and methods for data collection and analysis. Section 4 outlines our analysis and findings before we critically review and discuss these in Section 5. Section 6 provides conclusions, implications and future research.

2. Literature review and theoretical background

2.1 Multi-tier SSCM

SSCM research has long focused primarily on dyadic relations between focal firms and direct suppliers as the natural starting point and primary focus for practitioners in focal firms. In today’s globalized world, however, supply chains represent complex, dynamic (Beske, 2012) and often opaque networks with limited visibility (Sancha et al., 2019). Yet, it is becoming apparent that sustainability impacts – from modern slavery and other human rights violations (Marques et al., 2024) to carbon emissions and biodiversity loss tend to occur upstream in supply chains where they coincide with weaker institutional contexts and more vulnerable communities (Kshetri, 2021). While SSCM in focal firms, and research, has long focused on direct supplier relations, recent and upcoming regulatory changes are pushing companies’ legal accountability to include more distant upstream tiers. This reinforces earlier voices (Touboulic and Walker, 2015) highlighting the need for more nuanced, holistic studies of SSCM: “[t]he richness of the settings for empirical studies needs to be exploited” (Touboulic and Walker, 2015, p. 38) to generate theoretical insights. Research has shifted focus to triads of focal firm, for instance explaining first-tier suppliers’ twofold agency in cascading sustainability requirements (Wilhelm et al., 2016) and risks at lower-tier suppliers (Villena and Gioia, 2018). Yet as Dubois points out (2009), triads have rather straightforward links, so their characteristics resemble dyads because they lack the complex network and limited visibility of multi-tier supply chains and therefore cannot be viewed as an adequate proxy. Further, the high-sustainability-risk tiers typically lie beyond said visible horizon, requiring more comprehensive approaches (Sauer and Seuring, 2018) and specific targeting of midstream nexus suppliers like smelters (Sancha et al., 2019) or traders (Grabs et al., 2024; Grabs and Carodenuto, 2021). Consequently, multi-tier SSCM research is gaining more attention (Mena et al., 2013; Tachizawa and Wong, 2014), although studies that explicitly use a multi-tier research design or lower-tier suppliers’ perspective remain scarce, with notable exceptions on supply chain learning (Gong et al., 2018), supply chain leadership (Jia et al., 2019) and governance mechanisms to handle complexity (Gong et al., 2023). These studies build on the recognition that supply networks are complicated for unilateral management and explore mutual processes that actors use to cope, learn and collaborate for supply chain sustainability.

As upcoming regulations require companies to assume legal accountability beyond their visible horizon, companies are flocking toward technological solutions as (potential) means to efficiently collect and manage large data amounts required to conduct due diligence. Overall, this signifies a shift toward more efficient, (seemingly) objective and reliable ways of monitoring, verifying and managing sustainability issues than traditional approaches like onsite audits (Castka et al., 2020; Reid and Castka, 2023), e.g. through satellite monitoring of forest-risk commodity supply chains (Heldt and Beske-Janssen, 2023), or blockchain solutions for traceability under EUBR. The latter is covered in this paper and has received considerable attention in recent SSCM research.

2.2 Blockchain-based traceability

Blockchain emerged as a distributed ledger technology that provides a digital record of transactions and enables decentralized, unalterable and transparent information sharing within a network (Pournader et al., 2020). Blockchain has received much attention in supply chain management due to its potential for providing traceability in fragmented supply chains without the need for all parties to know – and trust – each other (Yavaprabhas et al., 2023) and supporting transparency toward stakeholders (Gligor et al., 2022). In theory, it facilitates secure information sharing, efficient processes and coordination, for instance, concerning food safety (Rogerson and Parry, 2020), anti-counterfeiting for pharmaceuticals (Babich and Hilary, 2020) or sustainability (Saberi et al., 2019). This is particularly interesting for SSCM: first because traceability per se is important for addressing sustainability deeper upstream (Kshetri, 2021) and second because blockchain can store additional information on sustainability practices or performance, e.g. certificates, audit protocols or chain-of-custody-related information (Yavaprabhas et al., 2023). With increasing due diligence requirements, companies look for ways to handle data efficiently and reliably.

With blockchain’s potential also come weaknesses: A central problem, termed the “garbage-in-garbage-out” problem, is the potential disconnect between digital blockchain and physical reality (Rogerson and Parry, 2020): As Babich and Hilary (2020, p. 4) put it: “There is little benefit in having an incorruptible ledger if the inputs are systematically corrupted,” either by intention or human error. Further challenges include lacking standardization and interoperability (Babich and Hilary, 2020) or scalability as blockchain was intended for “high-value transactions rather than high-volume transactions” (Pournader et al., 2020, p. 2073). Likewise, recognizing the resources, time and effort required by blockchain implementation, Ahmed and MacCarthy (2023) call for a more nuanced understanding of what traceability level and granularity is optimal for different supply chains rather than assuming a “more-equals-better” approach. Similarly, concerns around blockchain’s energy and carbon intensity (Babich and Hilary, 2020) suggest a need to assess where blockchain’s benefits warrant its costs. Lastly, Yavaprabhas et al.’s (2023) recent review demonstrates that scholars are divided on whether blockchain creates a trustless system, eradicating the need for interorganizational trust or whether it instead increases collaborative trust. Interestingly, many of these weaknesses correspond directly to blockchain’s anticipated strengths, suggesting that it eventually comes down to the nuances of implementation and supporting processes.

Blockchain has seen particular interest in supply chains of conflict minerals and other green transition minerals from similarly conflict-affected and high-risk areas, as Deberdt and Billon (2021) find reviewing (regulatory and voluntary) responsible sourcing initiatives. Early work by Hastig and Sodhi (2020) addresses blockchain-based traceability in cobalt mining specifically, synthesizing secondary data from academic and practitioner literature into business requirements and success factors. More recently, empirical qualitative studies are generating rich insights into blockchain’s practical working: Kshetri (2022) finds blockchain-based mineral supply networks demonstrate higher information flow and information authenticity and are better at including artisanal and small-scale mining (ASM) actors, thus empowering frequently marginalized groups; Ahmed and MacCarthy (2023) contrast blockchain in cobalt with other supply chains to identify adequate traceability levels; and Tuladhar et al. (2024) use information processing theory to explain how blockchain facilitates transparency and regulatory compliance with due diligence laws in conflict minerals sourcing by addressing information uncertainty and equivocality.

Overall, there is much research on blockchain’s potential and increasing empirical work on blockchain’s functionality but less is known on the nitty-gritty details of its implementation process. Extending prior work, we thus adopt a CAS perspective that helps explain how such innovations emerge in reality (Nair et al., 2016) and is gaining relevance in SSCM research.

2.3 Supply chains as complex adaptive systems

The CAS lens aims to explain supply chains “as living systems co-evolve with the rugged and dynamic environment in which they exist and identify patterns that arise in such an evolution” (Surana et al., 2005, p. 4236) and is deemed relevant for understanding complex supply chains (Carter et al., 2015; Choi et al., 2001; Surana et al., 2005). We adopt a CAS lens to capture the multi-agent, emergent and dynamic nature of the process and to explore upstream suppliers’ perspectives.

Choi et al.’s (2001) framework explains CAS through the following factors (see Figure 1): internal mechanisms (agents and schemata; self-organization and emergence; connectivity; and dimensionality), external environment (dynamism and rugged landscape) and their respective co-evolution (quasi-equilibrium; non-linear changes; and non-random future). Instead of viewing the supply chain as steered and controlled by a focal firm, the CAS lens emphasizes the agency of actors (beyond the focal firm) to influence the system and views the course of events as formed by the interaction of agents and dynamics. This agency is shaped by the schemata that actors adhere to, i.e. norms, rules and assumptions that shape actors’ cognition and behavior. Choi et al. (2001, p. 358) give the example of companies “work[ing] together through alliances based on shared norms and economic incentives.” Agents’ connectivity and dimensionality, i.e. autonomy/interdependence, refer to companies’ linkages across the chain and shape their agency. Self-organization and emergence refer to processes by which actions and responses of (relatively) autonomous actors shape the overall course of events with no single company deliberately steering it (Choi et al., 2001; Touboulic et al., 2018). Environmental factors include dynamism, i.e. temporal complexity, and a “rugged landscape,” i.e. a spatial complexity, leading to supply networks which are difficult to map (Touboulic et al., 2018). Carter et al. (2015) further introduce the idea of agents having visibility or rather a visible horizon, which limits the extent to which they can understand the system and its dynamics. This visible horizon is subjective to the focal firm or supplier (Carter et al., 2015). Together, these internal and environmental factors undergo a co-evolution shaping the CAS over time, which is characterized by intermittent quasi-equilibrium states, non-random futures and non-linear changes (Choi et al., 2001).

In concrete terms, focal firms attempting to manage sustainability issues beyond their visibility (and thus control) need to recognize that they operate in a self-organizing and evolving CAS where suppliers have agency and issues develop a life of their own (Touboulic et al., 2018). Establishing traceability through blockchain could extend each party’s visible horizon by connecting them systematically, thus creating the basis for managing sustainability issues along the chain. Taking a CAS lens suggests, however, that traditional control mechanisms may not work. The CAS perspective on SSCM has been applied to explain how environmental innovations emerge and diffuse in manufacturing supply chains (Nair et al., 2016) or how carbon reduction strategies are co-created in agri-food supply chains (Touboulic et al., 2018). We extend this research by placing emphasis on upstream suppliers’ perspective, exploring dynamics beyond downstream firms’ visible horizon.

3. Method

3.1 Research design

This study adopts a process-oriented (Langley, 1999, 2007) case study design (Dubois and Gadde, 2002; Ketokivi and Choi, 2014) to capture and analyze the emergence of blockchain-based traceability in critical minerals supply chains. Case studies are suitable for emerging, complex phenomena where prior research is scarce but opportunities to study it in the field exist (Verschuren, 2003). While blockchain-based traceability is no new topic, it remains a challenge to practitioners, and – while promising – we lack data regarding its practical implementation. Qualitative studies are considered important in the inherently complex SSCM field for generating differentiated understandings of underlying mechanisms and dynamics (Sauer and Seuring, 2019). In-depth single cases are viewed as instrumental to adequately capture network dynamics beyond triadic relations (Dubois, 2009) but also a necessary limitation given the labor-intensive nature of multi-tier research (Mena et al., 2013). Single case studies are not aimed at deriving universal rules based on statistical averages, but “at description and explanation of complex and entangled group attributes, patterns, structures or processes” (Verschuren, 2003, p. 137) and were used to study blockchain in coffee (Gligor et al., 2022) or mining (Tuladhar et al., 2024) supply chains. In this vein, case studies are considered effective for theory elaboration, i.e. using empirical data to challenge and further develop existing theory (Ketokivi and Choi, 2014), which we do for the “supply chains as CAS” lens by explicating the overlooked role of traditionally non-focal actors like upstream suppliers.

We further integrate process research methods since they are concerned with “how things evolve over time and why they evolve in this way” (Langley, 1999, p. 692). Process studies share key assumptions with our theoretical lens CAS and are therefore well-suited to capture emergence over time, multi-level interaction and nuanced complexity that we aim to explain. Process studies emerged as an alternative to variance-based approaches that understand phenomena in terms of variables and do not consider temporality or reduce it away for theoretical simplification (Tsoukas, 2017). Rather than measuring cause-and-effect between variables, processual approaches emphasize following events in real-time as they unfold and paying attention to dynamics underlying the perpetuation of seemingly “static” states, thereby lending themselves to studying how elements like traceability, sustainability or supply chain collaboration are created and maintained over time. Thus, process (Tsoukas, 2017) and CAS (Choi et al., 2001) scholars and advocates of theoretical pluralism (Cornelissen, 2017) argue that embracing – rather than avoiding – temporality, multi-level interaction and complexity provides avenues for relevant theoretical insights. Accordingly, it is unsurprising that seminal studies using CAS to study SSCM adopt processual approaches (Nair et al., 2016; Touboulic et al., 2018).

3.2 Research setting

We selected a critical minerals supply chain, specifically cobalt, as a case. Supply chains of critical minerals are notorious for being complex, dynamic and often opaque (Deberdt and Billon, 2021), flowing through numerous tiers (see Figure 2). Their extraction and trade are associated with serious environmental and human rights violations in sourcing regions such as the Democratic Republic of the Congo (DRC). Simultaneously, demand for these minerals is only growing – and pressure on supply chains increasing – due to their importance for the green transition. We managed to gain access to a large upstream mining company (UpstreamMiningCompany in the following) by embedding one author. Thus, this case was selected for its rather pointed situation from a sustainability perspective, combined with unique access upstream (Flyvbjerg, 2006).

In this article, we understand the internal mechanisms of CAS as those pertaining to the supply network surrounding cobalt sourcing for electric vehicle batteries, including both the physical and support supply chain (Carter et al., 2015) in the form of technology providers, certification schemes and similar actors. External mechanisms of CAS, in this article, relate to political, regulatory, non-governmental or civil-society activities.

3.3 Data collection

For data collection, we combined ethnographic fieldwork (participant observation) with expert interviews and secondary data from documents, industry webinars and events like conferences. Fieldwork “involves an intensive period of engagement (‘participant observation’) in a given [context]” (Atkinson, 2015, p. 12) and “actively participating in day-to-day affairs” (Emerson et al., 2011, p. 3). Participant observation provides thick data on empirical phenomena through in-depth participation in everyday work and meetings over an extended period. One researcher was embedded in UpstreamMiningCompany’s sustainability team, initially for an SSCM internship rather than research purposes, yet as intentions on a blockchain project emerged, we decided to use this unique opportunity to trace the project through to full launch. Over 12 months, the researcher participated in blockchain project meetings and everyday activities (see Table 1), taking field notes on observations and reflections concerning events, decisions, dynamics and challenges related to blockchain adoption. Aligned with our theory elaboration approach, observations were loosely guided by the CAS framework but open to emerging events. Whereas interviews are a well-established method in qualitative research and provide in-depth insights from practitioners not accessible through desktop research alone, participant observation is less common in SSCM research. However, the advantages and relative ease of conducting interviews have tipped into qualitative studies overly – or at least primarily – relying on interviews (Atkinson, 2015). Yet, we acknowledge that interviews provide a filtered view and are subject to limitations like hindsight bias, subjective views (which essentially is the point) and a certain blindness to underlying dynamics (Emerson et al., 2011). Thus, field notes and recurrent meetings with the non-embedded researcher were complemented with insights from additional semi-structured interviews, recorded and transcribed for analysis. For triangulation (Verschuren, 2003), we integrated data from external documents like annual reports and industry webinars to corroborate findings.

3.4 Data analysis

Data analysis is conducted in NVivo, following Mayring’s (2014) process (see Figure 3) that has been used in qualitative SSCM research (Beske et al., 2014). Following good practice in process methods (Langley, 1999), we wrote up a timeline of events, decisions and activities and then coded our data against the CAS framework as outlined above. In this process, we adopted an abductive approach, following Dubois and Gadde’s (2002, p. 558) ”tight and emerging framework” that lets researchers rely on the coding framework but remain open to divergences in data, thereby ensuring a clear focus while staying openminded. This is what Ketokivi and Choi (2014, p. 234) refer to as “the duality of being situationally grounded, but at the same time, seeking a sense of generality.”

During data collection and analysis, we worked individually and met at regular intervals to discuss observations and coding results, particularly aspects that showed discrepancies, complemented each other or were flagged as particularly interesting, using them as “key incidents” for further analysis into emerging themes. We, thus, iteratively shifted between the CAS framework as overarching codes and emerging subcodes from our data for different phases (see Table 2 for an excerpt, please see Online Supplementary Material A for full details). In line with adopting a CAS lens, our unit of analysis is multi-level: We use our case to analyze different agents at the firm-level (with particular focus on upstream) and their interplay with initiatives like the blockchain project or stakeholder alliance at network-level, against the backdrop of external dynamics on a systemic level.

4. Results

4.1 How does blockchain emerge?

Blockchain-based traceability is gaining momentum in supply chains. We take a CAS lens and analyze the emergence of a multi-tiered mineral supply chain over three phases (see Table 3 for an overview).

4.1.1 Phase 1: blockchain project preconditions and initiation

The supply network studied is highly complex with an opaque, loosely connected supply chain (see Figure 1). “We have this vision of having supply chains mapped, but the minute you have it mapped it has changed already. So how to manage that is a big question: the more dynamic, the less control” (I1, Supply chain sustainability manager, Downstream), i.e. dimensionality is high. Concerning sustainability schemata, the diversity of relatively independent agents creates diverging needs and interests, and while schemata around basic practices like joint reporting templates have co-evolved alongside conflict minerals laws, there is no overarching agreement or singular control. Blockchain is anticipated to bridge this control vacuum: “Traceability platforms for minerals [are] more of a supplement to existing standard schemes (…) there's more understanding that it's very difficult to collect and verify suppliers’ data without some kind of system or platform to keep track of it all. And because the legislation is getting stricter, like battery passport requirements” (I4, Sustainability manager, UpstreamMiningCompany). While control remains spread and decentralized with blockchain, it facilitates a more systematic, efficient process.

Due to the rugged landscape, downstream companies face substantial geographical, operational and cultural distance from upstream, but even midstream actors, limiting control and influence. In practice, much of downstream companies’ supply chain and sustainability issues lie beyond their visible horizon, leaving companies accountable and vulnerable to unexpectedly emerging issues. Further, sustainability issues are, reflecting a rugged landscape, often rooted in systemic challenges in sourcing regions, such as political instability, corruption and forced labor in DRC’s copper belt that in turn increase risks for child labor or environmental pollution.

Adding to the complexity, external dynamics affect supply chain actors differently, creating conflicting needs and divergent schemata on the best way forward. Downstream companies are exposed to external dynamics like stakeholder scrutiny from regulators, investors and NGOs. Similarly, since the upstream mining sector is quite concentrated, large mining companies face considerable scrutiny and, depending on headquarters locations, are subject to similar due diligence requirements as downstream firms. Midstream actors, by comparison, often go under the radar since even large companies’ names are unfamiliar to the public. This difference in public scrutiny and legal accountability complicate the process of building shared schemata.

External dynamics are pronounced in our case with sometimes conflicting effects: Critical minerals are essential to the green transition and demand for electric vehicles, batteries or photovoltaics puts pressure on critical mineral supply chains. “The problem and one of the biggest challenges at the moment is that geopolitics is playing against progress on business and human rights. With a race to access critical minerals related to the energy transition, security of supply is winning over human rights any day” (I6, CEO, Tech provider). Demand cannot be expected to decline soon and creates trade-offs between sustainability dimensions. Simultaneously, regulation is tightening across the globe, from earlier conflict minerals regulation to due diligence laws and EUBR.

The blockchain project’s (“BlockchainStartup in the following) start co-occurs with two external dynamics: first, the rise of blockchain technology for commercial applications, and second, the looming EUBR. “The legislation is getting stricter, like Battery Passport Requirements. (…) Mining companies see how difficult it is to collect data for clients and, ideally, they need to automate it” (I4, Sustainability manager, UpstreamMiningCompany). UpstreamMiningCompany senses this and decides to launch a pilot project for blockchain-based traceability to secure a future competitive advantage, providing traceable cobalt to downstream companies. Simultaneously, it serves to boost their sustainability reputation: “BlockchainStartup was started as a marketing initiative. (…) you can use this as your advantage in communications with clients and other stakeholders” (I4, Sustainability manager, UpstreamMiningCompany).

4.1.2 Phase 2: basic pilot trial – self-organization and momentum from upstream

Taking a pragmatic, self-starting approach, UpstreamMiningCompany and BlockchainStartup begin experimenting with practical implementation and understanding the needs for blockchain, to be able to hire the right company for the actual blockchain programming: Starting on a mining site, they test how it would work with scanning bags, collecting data, how material would flow hypothetically, how to link the digital and real-life part: “no one is stopping you from implementing traceability tech before you know [who is in your chain]. Start and just slowly, reeling it back in, you don't even have to do mapping, it's fine to just (…) say let's get some of these [RFID] tags or something and slowly work your way up” (I7, Founder, Tech provider, with background in upstream). Extending this, they identify a smelter headquartered in the EU, thus being under more pressure from external regulatory and stakeholder pressure. UpstreamMiningCompany here displays a sound understanding of the surrounding rugged landscape, diverging schemata and how external dynamics make others more/less inclined to join the pilot. Again, engaging downstream actors is prevented by the limitations of UpstreamMiningCompany’s own visible horizon. In this basic pilot, they are not tracking a specific batch of cobalt, but rather understanding what is needed further down the chain.

UpstreamMiningCompany, however, soon realizes they cannot shoulder blockchain and related expenses alone and need to engage others. “We initially partnered with [trader] that is already under the vertical chain …, but we realized that without other cobalt producers it would not make sense … you are mixing anyway, and it just doesn't make sense to track only your cobalt, like nobody at the top of the chain, nobody has the task to track everything … in the process of formulating hypotheses about what we're doing, it became clear that it didn't make a lot of sense” (I7, Founder, tech provider). UpstreamMiningCompany thus intensifies – rather than giving up or handing over – its agency: by identifying and specifically targeting competitors who share the schema of traceability as a competitive advantage, self-organization emerges between upstream actors.

Simultaneously, taking self-organization further, UpstreamMiningCompany were part of discussions to form a stakeholder alliance around battery supply chains as a joint forum on regulation and traceability topics, i.e. a platform for agents to meet and discuss schemata. “[StakeholderAlliance]allows companies to form a common language in terms of ESG targets and indicators, to form common standards (…) The fact that [StakeholderAlliance]is a multi-stakeholder organization makes it possible to integrate the vision of different parties” (I5, Program Manager, StakeholderAlliance). Having this joint forum facilitates recognition of weaknesses across actors and how to complement them selectively: “If it's a common goal, if it's directed in interest and plus you understand each other's limitations, then basically rowing in the same direction is not as difficult” (I7, Founder, Tech provider). While kept formally separate, one central idea with StakeholderAlliance was to develop schemata for sustainability metrics of practices and performance to track. “It was an opportunity to participate in the introduction of these new standards, talk to the whole supply chain, boost our power in supply chain” (I4, Sustainability manager, UpstreamMiningCompany). Again, agency emanates from upstream and showcases awareness of wider stakeholder dynamics and the need for collective action.

4.1.3 Phase 3: full pilot and launch – emergence through rugged landscape

Building on initial momentum, UpstreamMiningCompany and BlockchainStartup extend the pilot across the chain and integrate StakeholderAlliance’s sustainability standards, launching a complete blockchain-based battery passport in 2023.

The primary step for UpstreamMiningCompany is bringing further actors onboard. However, limited visibility and high dimensionality prove a challenge as midstream actors are reluctant: “it is challenging to get the full supply chain on board, especially midstream. To succeed on this market, companies need to have established networks and trust with supply chain members. You need to solve challenge on disclos[ing] one set of data but at the same time not disclosing the sensitive data. If you have something missed somewhere, that's a very big reputational risk” (I4, Sustainability manager, UpstreamMiningCompany). While blockchain, in theory, enables trustless operation, in practice it requires trust in setting up pilots and figuring out data sharing and interoperability. Essentially, it becomes “trustless” over time if companies involved put in the work, but hard to implement when there is no trust to begin with. “You have to have normal trust with them, and it's a very long process. So you can't just get together with three competitors and [expect to] have a f****** great time” (I7, Founder, Technology provider). Previously-built trust was essential in empowering UpstreamMiningCompany to onboard partners.

Self-organization thus remains volatile as companies cite fears of violating anti-trust rules against data sharing: “Anti-trust mechanisms are used as an opportunity to keep information out for fear that publishing excessive information about the supply chain would increase reputational risks” (I5 Program manager, StakeholderAlliance). Data integrity and unwanted public attention are one concern and larger refiners or smelters currently see mostly the additional work but limited incentive. “We need to be able to do it. So it needs to be a hands-on thing and it needs to be a balance between effort and benefits that is adequate. So if it does not create business value, we won't buy it” (W5, Director regulatory affairs, Midstream). Another concern, particularly for traders and intermediaries is the risk of appearing redundant and being cut from supply chains. One downstream manager is very direct about this: “Supply chains are not today set up for companies to start asking for visibility. You buy things on the market and you're not even asking where this is coming from, and there have been multiple aggregations. Almost every supply chain in the world has traders and the minute you hit a trader you have a problem: they bring little value to supply chains, they create blocks in information flow both ways. There's no efficiency, there's no logic of networks. It's just sub-optimized at all levels and I think the minute we are able to see full supply chains, we will start asking ourselves questions on how we can make it more efficient” (I1 Supply chain sustainability manager, Downstream). Whereas both upstream and downstream actors have higher interest in building traceability, it is hard to determine who is sourcing from whom, as visible horizons of each side are not reaching far enough into midstream to connect. As a workaround, UpstreamMiningCompany through StakeholderAlliance and BlockchainStartup engaged various downstream companies without either side knowing whether they share the same chain, “for example, AutomotiveCompanyA comes in and completely does not manage to track their whole chain. And even AutomotiveCompanyB [with decent visibility] actually said ‘we'll be interested to see, whether or not the cobalt actually ends up coming to us’” (I7, Founder, Tech provider). So here, UpstreamMiningCompany and downstream firms deliberately forged connectivity across the chain, hoping to meet in the middle. Working tier-by-tier, they gradually built traceability, creating a binary traceable/non-traceable system with a mass-balance approach at smelters and refiners: “in our system you can find out, what's coming to you from these three specific companies, and what's not coming to you from those three companies. So you have a binary system. Traceable, non-traceable. Here's X percentage in X volume coming in at X stage from a chain” (I7, Founder, tech provider). Our interviewee continues with an anecdote: “But it’s quite valid and useful information, especially if AutomotiveCompanyA thinks they have all their cobalt coming from MiningCompanyB, because two battery engineers say they buy it all from MiningCompanyB through TraderA, And then you realize MiningCompanyB doesn't have that kind of volume” (I7, Founder, tech provider). Midstream actors’ dimensionality allows them to act independently and exploit the chain’s opacity to use misleading or at least vague terms, whereas downstream firms’ restrictive visible horizon combined with wishful thinking/naivety allows them to believe in low-risk procurement – until traceability efforts uncover the actual situation.

The rugged landscape and dynamic environment create a scenario where an already complex challenge becomes more complicated the further you dive into it, with actors encountering issues they did not know they had – “as always the devil is in the details. And when you start doing something you discover all these nitty gritty things that you often more stumble with than the overall philosophy behind it” (W5, Director regulatory affairs, Midstream). This relates both to implementation practicalities and sustainability: One of blockchain’s strengths is that it can handle any data, from excel files to videos or certificates, creating flexibility. The basic trial revealed that placing QR codes on bags is unfeasible because printing and scanning involve manual labor, whereas, for radio-frequency identification (RFID) tags, scanners can be installed at site gates to scan bags automatically upon leaving or entering. In practice, however, this spawns hyper-local challenges like finding companies in rural DRC to produce metal frames for scanning RFID tags. It also showcases the heterogenous conditions found across the rugged landscape of mining supply chains. Consequently, the pilot focuses on large mining sites, whereas ASM is supposed to be involved only later due to the added complexity.

Similarly, concerning sustainability, getting full traceability means companies can no longer hide behind cushy opacity and instead may be directly tied to emerging issues. One interviewee summed up the industry’s underlying dilemma: “I think in many cases, the company is better off not knowing. It's the Don't-Tell principle. If we take away all the fluff and the pretty phrases with the big grandstanding, then systemically there's very little incentive for a company to know what's really going on in their chain” (I7, Founder, tech provider). Another adds that blockchain is only the first step and “[t]raceability is something that adds value to the system by tracing goods, but there is still a need to verify the actual working conditions and human rights impacts on the ground” (I6, CEO, Tech provider). In this case, blockchain stores information on both, suppliers’ sustainability practices (e.g. human rights risk assessment) and performance (e.g. carbon footprint), encompassing environmental and social sustainability aspects. However, this remains a static snapshot much like an audit, whereas new or undiscovered issues can emerge at any time. Interviewees indicated blockchain’s “garbage-in-garbage-out” problem of reliably linking digital and physical reality.

Overall, UpstreamMiningCompany and its partners now face a new level of blockchain challenges and costs: Besides the aforementioned practical challenges, developing the actual blockchain platform took considerable time. While the technology itself exists, it is not as “off-the-shelf” available (or affordable) as expected. “Because it's very expensive and requires a lot of resources, it seems to me that at least in the near term it's going to be a niche solution for some critical products, whether it's metal or non-metal. So it has to be something going on with a product, affect reputation or bring other risks, so that there would be such a powerful incentive” (I4, Sustainability manager, UpstreamMiningCompany). Similarly, interviewees are clear that blockchain is resource-intensive and has no umbrella solution: “Again, the question is why (…) Is it worth it? And especially now that everyone's saying you should, trace, I don't know, food. I'm kind of shocked because it's just some weird shit, you know, I mean, tracing a box of apples? A crate of apples? Like three crates of apples? What's the purpose of that?” (I7, Founder, Technology provider). This provides a sobering outlook but also hints at non-linear changes characteristic of CAS: whereas blockchain implementation currently is suited for high-value, high-sustainability-risk material, this scope could change if sustainability issues emerge elsewhere and through stakeholder pressure snowball so that costly blockchain solutions make sense.

This phase concluded with the launching of the first battery passport in Spring 2023, demonstrating a step toward EUBR compliance. Concerning CAS co-evolution, this launch marks the reaching and settling into a plateau, although not a full quasi-equilibrium state: With the pilot successfully launched and proof-of-concept delivered, the real work begins. Yet, involved companies got a reality-check, disillusioned about the challenges and costs ahead, giving a bump to previous momentum. “EU regulation is driving adoption [of product passports], right. But obviously, it's a blunt instrument. And the devil is in the detail, as we've kind of discovered on this pilot (…) we need to get that balance right between what the regulation is trying to achieve and what is practical because the regulations don't understand the realities of supply chains” (W5, Co-founder, Technology provider).

4.1.4 Phase 4: post-script developments

While the following developments occurred after the study concluded, they are essential enough to include them here. In Fall 2023, BlockchainStartup was placed on hold and some key staff left. The entire project had grown increasingly expensive, and whereas actors could afford extra projects on sustainability when metal prices were good, this changed when prices plummeted with a recent economic downturn and geopolitical instabilities weighing heavily on the industry. This climate makes it even harder for startups to ensure funding and to engage hesitant midstream companies or include ASM. At the time of writing, BlockchainStartup’s future is uncertain.

5. Discussion

Our study provides longitudinal insights into the co-evolution between agent-driven and environment-driven processes. We find evidence that upstream actors play an important and underexplored role in the emergence of blockchain-enabled traceability. Below we develop a conceptual framework (see Figure 4) and propositions (see also Online Supplementary Material B for overview table).

First, we demonstrate that upstream actors have agency and that exerting said agency is instrumental in the initiation (and possibly, discontinuation) of blockchain. In our case, UpstreamMiningCompany anticipated and even preempted downstream buyers’ actions by piloting a blockchain project and initiating a broader StakeholderAlliance in response to the EUBR’s traceability and battery passport requirements. Further, concerning motivations, UpstreamMiningCompany took action when expecting financial benefits from good marketing and more direct supply relations (and ceded action when cobalt prices and financial prospects dried up). Our empirical evidence thus challenges and extends Nair et al.’s (2016, p. 73) proposition that “[t]he emergence of environmental innovations in a supply network begins with a dominant buying firm sensing external environmental stimuli.” Concerning when and why innovations such as blockchain emerge, we therefore propose:

P1.

The emergence of blockchain-based traceability for sustainability in a supply chain begins with the sensing of external environmental stimuli by either a dominant buying firm or a traditionally “non-focal” firm such as an upstream supplier.

P2.

Firms that sense external stimuli but are not targeted directly by them (e.g. upstream suppliers) take action on blockchain adoption when they anticipate gains from demonstrating traceability and sustainability standards.

Similarly, upstream actors’ agency shapes the further emergence and self-organizing behind blockchain adoption by carefully reading supply chain dynamics in the rugged landscape and orchestrating who and how to approach to get the project rolling. However, limited visibility is a central hinder to multi-tier SSCM efforts (Sancha et al., 2019). Following from our prior assertion that non-focal firms have agency, we find that they are likewise restricted by their visible horizon. UpstreamMiningCompany initially intended to work more closely with downstream buyers but – unable to cut through the opaque midstream – struggled to identify which downstream firms were ultimately sourcing their cobalt. Eventually, they resorted to running the pilot with those within their visible horizon, i.e. a trader in their network and several competitors, for pragmatic reasons. Whereas the concept of “visible horizon” is well-established thanks to Carter et al.’s (2015, p. 93) seminal piece, we were surprised to find upon re-reading that their theorizing is – again – built around the downstream actor, positing that “[t]he supply chain is bounded by the visible horizon of the focal agent.” This is probably in line with and applicable for most SSCM research and we do not presume that Carter et al. (2015) meant to say that only focal agents have a visible horizon, yet they do (implicitly) say that only focal agents’ visible horizon matters. For the sake of encouraging future research into the matter, we see this as an opportunity to put the spotlight more explicitly on non-focal actors who have been underresearched and undertheorized as of now and propose the following extension:

P3a.

Focal firms have visible horizons, but so do traditionally non-focal firms such as upstream or midstream suppliers.

In our case, the visibility issue lies particularly with midstream opacity. We find evidence that it can be an effective workaround for upstream and downstream actors (regardless of whether they share a chain to begin with) to connect and work from their respective “visibility bubbles”: UpstreamMiningCompany’s workaround was to co-initiate StakeholderAlliance and leverage this cross-supply-chain forum to collaborate with downstream automotive manufacturers, irrespective of evidence that they shared a supply chain. This reinforces the significance of prior research highlighting nexus suppliers’ – like smelters – role for gaining upstream visibility (Sancha et al., 2019) and the need to extend current multi-tier SSCM approaches toward a cascaded perspective (Sauer and Seuring, 2018). Recent work in palm oil supply chains has similarly highlighted midstream actors like traders (Grabs et al., 2024; Grabs and Carodenuto, 2021). Whereas these studies have emphasized various midstream actors for their nexus position, our data demonstrates that upstream actors can assume a comparable function when their schemata align better than midstream actors’. This opens an alternative route for tackling traceability from both ends in cases of midstream opacity and inertia. Therefore, building on and refining this stream of research, we suggest the following proposition to guide further research into a more nuanced and differentiated understanding of non-focal actors’ roles:

P3b.

The relevance of companies in implementing traceability efforts is less determined by their position in the supply chain (e.g. nexus suppliers like smelters or mills) and more determined by (a) the complementarity of their visible horizon and (b) their commitment to sufficiently aligned sustainability/traceability schemata with their respective partner.

Further, we corroborate prior evidence that so-called boundary-spanning agents like industry consortia or stakeholder alliances are important platforms (Touboulic et al., 2018), particularly for “blockchain implementation [where] joining consortia is critical because it yields cost savings and accelerates learning among the stakeholders” (Yadlapalli et al., 2022, p. 298). Our data support Touboulic et al.’s (2018, p. 330) proposition that “[b]oundary-spanning agents, comprising internal and external agents in the network, can help overcome existing conflicts between the schemas of agents and facilitate the proliferation of environmental strategies in the network.” We find evidence that this holds true from an upstream perspective: Building shared schemata around traceability’s importance and sustainability standards was critical in onboarding midstream actors, yet this would have been hard without StakeholderAlliance as a discussion forum. Similarly, blockchain – while triggered by external dynamics like EUBR – emerged essentially as a platform to formalize these schemata on how to collaborate, define sustainability and trace material. Since these considerations transcend single supply chains, discussions need to be had at the industry level, involving third-party stakeholders and reflecting prior work on multi-tier governance mechanisms (Gong et al., 2023; Tachizawa and Wong, 2014). In our study, the higher-level StakeholderAlliance and more concrete blockchain application emerged as co-evolving, complementary arenas for formalizing and materializing shared schemata, therefore we suggest the following proposition on blockchain’s role in a CAS:

P4a.

Blockchain acts as a platform for formalizing and materializing shared schemata, thereby facilitating trust across tiers and the “onboarding” of new actors.

Blockchain’s trustless element is considered a central characteristic and some scholars view it as “likely to transform the current trust-based theories in supply chain management.” (Saberi et al., 2019, p. 11). In our study, the “trustless” component and operative trust in blockchain only become possible because processual trust is gradually built/extended among UpstreamMiningCompany and its partners during the implementation and standardization process by overcoming unexpected challenges and disagreements. Yet, our data demonstrate how trust is still essential for establishing blockchain in the first place and convincing companies to participate: UpstreamMiningCompany opted to partner with a trusted trader for the first trial, before building on the initial success by onboarding others. Besides, trust and visibility are important for managing suspected/detected sustainability issues, either for direct follow-up or effective cascading of requirements. Further, in treating blockchain as a proxy for trust, we see a risk that this could erode trustful relations as companies underestimate the need to build and invest resources in building such ties, although they remain critical for the management of detected issues. We agree with and provide empirical evidence for Yavaprabhas et al.’s (2023, p. 64) conclusion: “Ultimately, blockchain alone may not be capable of building an ideal trusted ecosystem of supply chain networks. In the long run, other instruments and mechanisms are required to be involved in the sustainable enhancement of trust in supply chain management”. We condense our contribution in the following proposition:

P4b.

Once established, blockchain can facilitate trust-building processes among involved actors and fast-track or replace trust-building in the “onboarding” of new actors, yet the adoption itself cannot occur without prior trust-building efforts among initial members.

While blockchain does have advantages over regular ledger systems (e.g. Excel), many of its heralded benefits need to be developed over time (e.g. its trustlessness, interoperability and efficiency) and/or independently of blockchain (e.g. visibility, standards and sustainability impact), plus its implementation may not be as straightforward and smooth as companies may expect/hope for. Whereas our findings support Rogerson and Parry’s (2020) point concerning the digital-physical nexus as a persistent weakness, Tuladhar et al. (2024) suggest an interesting mechanism, arguing that because actors are accountable and liable for information put on blockchain, the risk of fraud is reduced (i.e. data veracity improves, not via increased audits but via legal liability). We agree with Sternberg et al. (2021, p. 85) that “practitioners as well as researchers should have a critical attitude about the technology’s promised benefits.” However, if a supply network has sound SSCM practices in place, decent levels of trust and visibility and shared schemata, we consider blockchain a useful and effective tool.

Zooming out, blockchain adoption appears as another iteration of CAS’ co-evolution between industry and its external landscape, this time more technologically-driven: Minerals supply chains have shown non-random trajectories that tend to cycle through quasi-equilibrium, punctuated by external disruptions from regulatory or other stakeholder pressure like EUBR. Companies reinforce previous sustainability activities, invest in new ones and get together on industry-level to create compliance (in prior iterations, e.g. developing conflict minerals reporting templates) before efforts gradually taper off – due to rugged landscape and challenges with translating ambitions into on-the-ground change –, settling into a new quasi-equilibrium. Although EUBR is incisive, in CAS terms it represents a non-random event that was to be expected. While blockchain initially was the trigger/need for joint traceability schemata, it eventually turned into the mechanism by which such schemata are formalized and perpetuated.

6. Conclusion and implications

We set out to understand how blockchain-based traceability emerges in complex supply chains, such as those for critical minerals. We conducted a process-oriented case study, drawing on data from extended (>12 m) participant observation with one researcher embedded in UpstreamMiningCompany and participating in blockchain project meetings, plus triangulatory expert interviews and document reviews, analyzed abductively.

We find evidence that actors involved in blockchain implementation resemble a CAS and, therefore, do not follow classic downstream-controlled SSCM patterns. In our case, blockchain-based traceability in the cobalt supply chain was co-constructed over time, fundamentally driven by a large upstream supplier but enabled through supply-chain-spanning collaboration with like-minded downstream actors and successive expansion into the opaque midstream, enabled through the stakeholder alliance forum and formalized in the blockchain. With our study, we corroborate seminal work by scholars emphasizing the emergent nature of SSCM strategies and calling for a shift away from viewing SSCM strategies as “top-down and rational” (Touboulic et al., 2018, p. 30) by default. We develop propositions that extend CAS theory accordingly. In our case, upstream actors play a pivotal role in the emergence and expansion of blockchain-based traceability, by engaging midstream partners, building momentum upstream and initiating industry collaboration. The answer to our research question “How does a blockchain-based traceability system emerge in a complex supply chain?” is embedded in our propositions (P1-4b): We find that proactivity for establishing blockchain-based traceability need not originate downstream. Instead, traditionally non-focal firms such as upstream suppliers can take the lead (P1) when they sense favorable external dynamics such as emerging laws that render traceability a prerequisite (P2). However, they too are subject to a restricted visible horizon (P3a) and need to forge partnerships across the supply chain to implement blockchain-based traceability. We find that selecting appropriate partners is not solely determined by their respective position, e.g. gatekeeping intermediaries where the product gets mixed, but also by them holding aligned schemata and complementary visible horizons (P3b): if your midstream supplier holds key information on the upstream supply chain, this does not qualify them as an effective partner for implementing blockchain-based traceability unless they also share your interest in improving transparency and sustainability standards. Multi-stakeholder alliances play an important role in co-creating shared schemata, as prior literature suggests (Touboulic et al., 2018). Blockchain requires trust between the initial partners (P4b) but, once set up, acts as a platform that formalizes and propagates shared schemata, e.g. around data governance, supporting the onboarding of additional actors and extension of traceability across the system (P4a).

6.1 Research implications

Multi-tier SSCM research has mostly adopted a downstream-centric perspective, thus far, paying limited attention to upstream actors’ perspectives. Our study, however, demonstrates that upstream suppliers can play a crucial role in the initiation and implementation of blockchain-based traceability systems, particularly in complex and opaque supply chains, and that – in turn – the blockchain platform and wider stakeholder alliances help formalize and perpetuate sustainability and data governance standards across the supply chain. This highlights the untapped potential for more effectively rolling out traceability projects and achieving broader supply chain sustainability improvements. While requiring further validation (see Section 6.2), this suggests that multi-tier SSCM research in its current form may be leaving interesting insights on the table. Based on our contributions, we suggest three main implications for research from our study: First, our results imply that the current downstream-centric perspective in much multi-tier SSCM research could access new insights by explicitly taking the perspective of “traditionally non-focal” firms such as midstream or upstream actors. This implies that insights could be gained from building empirical insights into how upstream as well as midstream actors behave and interact with downstream companies in building traceability, supporting technology and sustainability standards. Even from a downstream perspective, it would be relevant to explore how multi-tier supplier relations unfold in practice in a case like ours. As downstream companies seek to build traceable and “clean” supply to comply with due diligence regulations – investing in technology, risk assessments and supplier development programs – incentives increase to capitalize on these efforts and, once that material is fully traceable, secure its long-term supply from upstream actors. In our case, the upstream supplier even stepped in and preempted downstream action. A stronger consideration of upstream actors and the possibility of longer-term supply relations in research and theorizing appears needed, potentially resembling decommoditization trends pointed out earlier in mining (Sauer and Seuring, 2018) and agrifood commodities (Heldt and Beske-Janssen, 2023). Second, and building on the first, the “supply chains as CAS” lens clearly lends itself to this “flipped” perspective on multi-tier SSCM from the upstream perspective by capturing multi-agent dynamics and complexities, as we demonstrate in our article. Yet, prior theoretical advancements have implicitly overlooked or underestimated the role of traditionally non-focal actors for instance concerning their visible horizon (Carter et al., 2015) or agency in initiating sustainable innovation (Nair et al., 2016). We thus suggest expanding the notions of the visible horizon and agency within CAS to recognize upstream suppliers’ potential and allow for more accurate capturing and study of empirical phenomena in multi-tier SSCM. This implies asking what insights emerge from reframing SSCM from something “done by” downstream companies to something emerging and shaped across the chain. Third, our work provides empirical nuance and grounding for prior blockchain research that focused more on the technologies’ potential and opportunities or pilot studies from technology providers and focal firms’ perspective. While retaining a cautiously optimistic outlook, our study points toward considerable challenges during the practical implementation of blockchain, particularly in the complex, opaque commodity supply chains where its potential usefulness could be largest, echoing recent work by Yavaprabhas et al. (2023). For research, this implies a need for more critical, empirical studies of blockchain implementation’s complexities and intricacies to produce a comprehensive, meaningful understanding of blockchain.

6.2 Limitations and future research

Reviewing limitations, the qualitative research design – while deliberately chosen for its strengths in disentangling nuanced phenomena – has its drawbacks. We investigate a single case and our strategic sampling is no basis for statistical generalization. We have limited data from the downstream perspective and participant observation data from being embedded downstream would have been interesting. Our case blockchain is a consortium blockchain (common for supply chains) rather than truly public (i.e. fully decentralized, permissionless, non-restrictive as for cryptocurrencies). Concerning future research, complementary qualitative studies that generate further evidence into transferability (e.g. to other commodity supply chains or to other technologies) are welcome, as is subsequent quantitative testing. The main avenue forward is further explaining the role of traditionally non-focal actors in implementing traceability, sustainability and supporting technologies in complex supply chains. Further, our research suggests that due diligence regulation affects downstream and upstream actors’ relations in complex commodity supply chains. Analyzing the role of traceability-oriented technologies in this appears interesting as they could either support focal firms’ efforts to secure “clean” supply from particular upstream suppliers, hinting toward decommodization tendencies, but may just as well create a “trustless” exchange of verified “clean” commodities from shifting origins.

6.3 Practical implications

Building traceability in multi-tiered supply chains is challenging to implement single-handedly for downstream firms. We derive several practical recommendations: Supply chain managers should recognize how important upstream actors can be for implementing traceability practices. Upstream (and/or midstream) actors have access to a distinct visible horizon, network and local knowledge and can help tackle opacity from both ends. Downstream managers should consider collaborating with upstream actors when implementing traceability projects when struggling with midstream visibility. As our case demonstrates, this can work even if it is unclear if partners even share a supply chain, and multi-stakeholder initiatives present useful platforms for identifying proactive partners and aligning schemata. Further, our study shows that to effectively manage expectations and implementation, managers need to recognize that blockchain is no panacea. In the end, blockchain is just a technology that relies on visibility, trust, sound standards and follow-up capacities to exist in the first place and in their own right. Managers should critically evaluate their visibility levels and SSCM’s maturity – or their earnestness in committing to building these – since traceability is likely to reveal sustainability issues they did not know they had.

Figures

Figure 1

Complex adaptive system dynamics

Figure 2

Simplified structure of battery mineral supply chain

Figure 3

Process of qualitative content analysis

Figure 4

Complex adaptive system dynamics during implementation of blockchain-enabled traceability in a multi-tier critical mineral supply chain

Table 1

Overview of data sources

Data	Details	Amount	Period
Participant observation
	Embedded time (upstream)	48 weeks	2022
	Internal weekly meetings (upstream)	30	2022
	Blockchain project meetings	30	2022
	StakeholderAlliance meetings	1	2022
	Industry conference, online	5 days	2022
	Industry webinars, online	5	2022–2023
	Informal conversations	>100	2022–2023
Additional interviews (cited as, e.g. I’x’, position, organization)
	Downstream - I1, Supply chain sust. manager - I2, Sustainability manager	2	2023
	Midstream - I3, Sustainability manager	1	2023
	Upstream - I4, Sustainability manager	1	2023
	StakeholderAlliance - I5, Program manager	1	2023
	Technology provider - I6, Founder - I7, Founder	2	2023
Webinars (cited as, e.g. W’x’, position, organization)
	Technology provider - W1, Founder	1	2021
	Midstream company - W2, Sustainability director	1	2021
	Technology provider - W3, Vice-president Marketing	1	2022
	StakeholderAlliance, tech provider, downstream - W4, Project lead, StakeholderAlliance - W4, Founder, tech. provider - W4, Digital solutions manager, tech. provider - W4, CSR manager, downstream	1	2023
	Midstream, upstream, tech. providers - W5, Director regulatory affairs, midstream - W5, Founder, tech. provider - W5, COO, tech. provider - W5, Director sustainability, upstream	1	2023
Documents
	Annual/sustainability reports, supporting documents	47	2020–2023

Source(s): Authors’ own creation

Table 2

Excerpt from coding structure

CAS codes, deduct		Emerging themes	Exemplary events and actions	Exemplary quotes
Internal mechanisms	Agents and schemata (Phase 1)	Diverging schemata across supply chain	Downstream actors face pressure from EU Battery Regulation (EUBR) but struggle with a lack of visibility and inertia from midstream “Hidden” midstream actors opposed to attention and reputational risk that traceability brings (…)	“You need very strong incentives to allow supply chain transparency. Unless there is a regulation, unless there is an alternative benefit for the suppliers, it's really hard to get that transparency. When we are asking our suppliers to get ready to start opening up their information sources, they refuse to do it until the moment when the law [EUBR] is signed”(I2, Sustainability manager, downstream)
	Agents and schemata (Phase 2)	Upstream supplier initiates blockchain pilot with schematically-aligned actors Difficult to onboard midstream due to diverging schemata	For participation in the blockchain pilot, UpstreamMiningCompany reaches out to actors along the supply chain, prioritizing actors that – for different reasons – have shared schemata: (…) UpstreamMiningCompany initiates StakeholderAlliance as forum for discussing schemata across chain	“If it's a common goal, if it's directed in interest and plus you understand each other's limitations, then basically rowing in the same direction is not as difficult”(I7, Founder, Tech provider) “You have to have a normal trust between them [partners in initial blockchain], and it's a very long process. So you can't just get together with three competitors and you're gonna have a f****** great time. You have to socialize together, you have to get to know each other, listen, argue, talk“(I7, Founder, tech provider)” “[StakeholderAlliance]allows companies to form a common language in terms of ESG targets and indicators, to form common standards (…)”(I5, Program Manager, StakeholderAlliance)
	Agents and schemata (Phase 3)	Shared traceability schemata formalized Shared sustainability schemata developed and formalized	During blockchain pilot, shared traceability schemata, e.g. on data sharing, were developed. Blockchain design formalizes these schemata Shared sustainability schemata developed between actors and formalized in StakeholderAlliance’s joint standards on sustainability metrics (…)	“this industry-wide body that provides clarity on the standards and approaches is a very good option. There are sticky issues around data governance that really need to be decided in a forum like [StakeholderAlliance]” (W5, Founder, Technology provider) “The main motivation [with StakeholderAlliance] was a unified goal of some sort, to realize that the whole industry has a common threat, that there is child labor, untraceable cobalt and so on. And having that common goal, it has helped a lot to actually unite the industry” (I7, Founder, tech provider)
Environment	Visible horizon (Phase 1)	Downstream faces visible horizon as barrier	Downstream actors face restrictive visible horizon Regulation (e.g. EUBR) extends downstream actors’ horizon of responsibility beyond horizon of visibility	“Visibility is very low in the supply chain (…) in most cases we don't know who the suppliers are beyond tier -one” (I2, Sustainability manager, downstream) “It's a fundamental question, how much you as a company can do, where is the boundary of your responsibility? And now the regulator, it starts to move you in the direction that the boundaries of your responsibility go much further than your perimeter of possession. Accordingly, your perimeter of impact becomes your responsibility” (I7, Founder, tech provider)
	Visible horizon (Phase 2)	Upstream actor’s visible horizon restricts pilot	UpstreamMiningCompany realizes that its visible horizon restricts range of pilot UpstreamMiningCompany initiates StakeholderAlliance to engage with downstream actors and extend visible horizon	“It was an opportunity to participate in the introduction of these new standards, talk to the whole supply chain, boost our power” (I4, Sustainability manager, UpstreamMiningCompany) “(…) nobody at the top of the chain, nobody has the task to track everything … in the process of formulating hypotheses about what we're doing, it became clear that it didn't make a lot of sense [to try without them]”(I7, Founder, tech provider)
	Visible horizon (Phase 3)	Upstream and downstream collaborate to extend visibility	Upstream and downstream companies join efforts for building visibility in midstream by working from their respective visible horizons	“for example, AutomotiveCompanyA comes in and completely does not manage to track their whole chain. And even AutomotiveCompanyB [with decent visibility] actually said ‘we'll be interested to see, whether or not [UpstreamMiningCompany’s] cobalt actually ends up coming to us’”(I7, Founder, Tech provider)

Source(s): Authors’ own creation; extended tables available in Online Supplementary Material

Table 3

Overview of CAS dynamics across phases

CAS codes		Key themes
		Phase I – Preconditions and initiation	Phase 2 – Basic pilot trial	Phase 3 – Full pilot and launch
Internal mechanisms	Agents and schemata	Diverging interests and goals (schemata) across downstream, midstream, upstream supply chain	Upstream supplier initiates blockchain pilot with schematically-aligned actors Difficult to onboard midstream due to diverging schemata	Shared traceability schemata formalized in blockchain Shared sustainability schemata developed and formalized in stakeholder alliance’s sustainability standards
	Self-organization and emergence	External pressure exerted on downstream sensed by upstream actor Upstream actor starts blockchain project to demonstrate traceability and sustainability	Blockchain pilot emerges between schematically-aligned actors Upstream actor exerts agency: includes competitors in pilot; starts stakeholder alliance to engage downstream	Full blockchain pilot, first battery passport implemented Distributed organization emerged yet remains volatile due to costs and practical challenges with blockchain Prior trust essential, but trust-building facilitated through blockchain pilot
	Connectivity	Complex, dynamic supply chain with low connectivity	–	Additional connectivity due to blockchain project and stakeholder alliance (“support supply chain”)
	Dimensionality	High dimensionality with autonomous actors and limited overarching control	–	–
External environment	Visible horizon	Downstream faces visible horizon as a barrier to traceability and legal compliance	Upstream actor’s visible horizon restricts pilot	Upstream and downstream collaborate from both ends of supply chain to extend their visible horizons
	Rugged landscape	Very rugged: low visibility and heterogenous conditions across landscape	Rugged landscape shapes which actors to involve in pilot Heterogenous on-the-ground conditions and infrastructure complicate pilot	Rugged landscape “hides” existing sustainability issues “Garbage-in-garbage-out” problem persists Midstream actors benefitting from lack of visibility
	Dynamism	Dynamic external environment (e.g. EUBR and HREDD laws, green transition and demand for critical minerals, tech progress with blockchain)	CAS-internally, blockchain more complicated to develop than expected, thus slower and more costly CAS-externally, geopolitical tension rising as Russia invades Ukraine, renewed urgency of green transition	Risk of uncovering emerging issues that actors did not know they had Blockchain’s immutability vs dynamic reality
Co-evolution	Quasi-equilibrium	Quasi-equilibrium shaken by EUBR	In flux	Still in flux, but settling into quasi-equilibrium as traceability and compliance appear feasible
	Non-linear changes	High hopes in blockchain technology as traceability/compliance panacea	In limbo: Uncertainty around blockchain’s economic impact (positive vs negative)	Blockchain as expensive and for high-risk contexts only; yet, high-risk status can change fast and non-linearly Volatile geopolitics persist
	Non-random future	EUBR incisive yet not unexpected	–	Regulatory compliance in sight, yet question of on-the-ground impact remains. New iteration likely Actors highlight need for systems thinking

Source(s): Authors’ own creation; extended tables available in Online Supplementary Material

Credit author contributions: LH: Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review and editing; EP: Conceptualization, Investigation (Data collection), Data curation, Formal analysis, Methodology.

Supplementary material

The supplementary material for this article can be found online.