Exploring how retail and logistics service provider managers make sense of sustainability in last mile delivery

2.1 Sustainability in last mile delivery

In recent years, last mile research has become more focused on sustainability, as complexity has increased and global challenges concerning climate and social issues have become more topical. In the last mile scene, this has a direct implication for managers as company decision-makers. Table 1 illustrates the context by outlining examples of the multitude of sustainability challenges, including emissions, pollution, urban biodiversity, noise and related safety and health issues for workers, customers and city inhabitants alike. It is in this complex and often ambiguous environment, where issues such as payment models, collaboration concerns and upholding social values in new delivery schemes can cause conflicts between sustainability and business outcomes (Moncef and Monnet Dupuy, 2021), that managers need to make sense of sustainability. Sustainability challenges are made more complex by the transformation of LMDs and the many heterogeneous stakeholders involved (Mangano and Zenezini, 2019). The transformation of the last mile landscape refers to the introduction of innovations, such as drones and delivery robots (Boysen et al., 2021), crowdsourcing and dynamic pricing (Mangiaracina et al., 2019). Ambiguities are created when, for example, the use of delivery robots is both cost-efficient and green but creates social sustainability challenges involving safety and equity (Garus et al., 2022). There are many different types of LMDs, such as home delivery, pick-up points and parcel lockers (Buldeo Rai et al., 2019), so there are many sustainability considerations for managers. For example, the sustainability of different delivery methods is difficult to evaluate, as the environmental sustainability of LMDs depends partly on the degree to which the deliveries can substitute for customers’ personal car travel and by which transport means the deliveries are organized (Bjørgen et al., 2021). In addition, consumer preferences are changing (Villa and Monzón, 2021), underlining the importance of delivery speed (Bjørgen et al., 2021). At the same time, consumers’ possibilities to make green delivery decisions are impacted by the communication between LSPs and retailers (Sallnäs and Björklund, 2023).

Both retailers and LSPs play important roles in making LMD more sustainable (Bask et al., 2016). However, retail and LSP interests might not converge when LSPs focus on cost reduction rather than environmental issues and when retailers are occupied with competing to satisfy customer needs (Kiba-Janiak et al., 2021). Retailers have mostly been passive toward sustainable LMDs (Kudla and Klaas-Wissing, 2012); their interest in environmental issues seems to decline when criteria such as price, reliability and load factor take precedence, leaving sustainability aspects dependent on the LSPs (Huge-Brodin et al., 2020). Retailers in end-consumer-oriented industries are more likely to be interested in and bring pressure to the sustainable activities of LSPs (Kudla and Klaas-Wissing, 2012), but many issues remain, as there is a lack of trust and knowledge sharing, which blocks the adoption of sustainable practices in the retailer–LSP relationship (Huge-Brodin et al., 2020). With the several barriers to adopting sustainable practices, including organizational, financial, retailer–LSP market, retailer–consumer market, governmental and technological (Sallnäs and Björklund, 2023), there is a need to understand the subjective interpretations retailers and LSPs have (Reinecke et al., 2023). As managers are primary stakeholders in the cascading effects impacting sustainability (Prataviera et al., 2023), this study explores how managers make sense of sustainability in LMDs by using CFs as an analytical lens.

2.2 Managerial cognitive frames

There is a growing need to examine cognitive and behavioral aspects of supply chain management (Fahimnia et al., 2019). CFs are part of sensemaking, i.e. the process of constructing meaning (Weick, 1995) and play crucial roles in managing sustainability tensions in organizations (Carmine and de Marchi, 2022). Depending on their CFs, individuals interpret information differently (Weick et al., 2005). Managers use CFs to make sense of various complex and ambiguous signals (Hodgkinson and Healey, 2008; Walsh, 1995). Cognitive elements and their links make up a CF (Hayes-Roth, 1977). The complexity of a CF is impacted by its structure and content, as well as the interaction between these two (Preuss and Fearne, 2022). The content of a CF relates to the domain (Hahn et al., 2014), such as sustainability, and the elements of the frame relate to how a person distinguishes or groups the attributes within this domain (Scott et al., 1979). When looking at CFs, one should consider structure in the sense of content and logic (Hahn et al., 2014). CFs are not static but change over time and are likely to become more complex as a person gains experience (Rousseau, 2001). CFs can therefore be both extended and transformed, as well as diffused within organizations, to become collective action frames (Benford and Snow, 2000). While CFs have been discussed as crucial for understanding the conditions under which companies can address sustainability challenges in their supply chains (Corsini et al., 2022; Hahn et al., 2014; Preuss and Fearne, 2022; Xiao et al., 2019), little is known about the CFs that retail and LSP managers hold in the context of LMDs or about their operational manifestations.

Decision-making is often viewed as a rational process, and the sensemaking part is neglected (Ericson, 2010), even if the process is entwined with sensemaking as an underlying dynamic (Boland, 2008; Ericson, 2010). Sensemaking provides the retrospective reflexive context and foundation for decision-making, and decision-making gives the direction and purpose necessary for navigating and making choices (Boland, 2008). When performing tasks, CFs provide managers with a structure for their assumptions and boundaries for what to include in their sensemaking (Smith and Tushman, 2005). Once CFs have been used to interpret an ambiguous environment, managers are likely to respond on that basis, leading to different sustainability decision-making processes (Hahn et al., 2014). Therefore, the CF perspective constitutes an important viewpoint for exploring sustainability in LMDs. Building on previous sustainability-related CF conceptualizations by Hahn et al. (2014) and the work on integrative complexity by Suedfeld and Tetlock (1977), among others, Preuss and Fearne (2022) conceptualized three different CFs held by supply chain managers: unidimensional, hierarchical and paradoxical. In their conceptualization, adapted to a framework for this study (Figure 1), they illustrate how the CFs that managers hold make a difference in how they address sustainability challenges in the supply chain.

The framework presented in Figure 1 depicts the ideal conceptualizations used in this study as an analytical lens. A unidimensional frame is indicated by a narrow view, and it focuses on one core element, such as cost or economic gain. A manager with this frame will most likely not consider sustainability (Preuss and Fearne, 2022), and this “business-case” frame is deeply rooted among managers (Menon, 2022) and embedded in organizations (Cornelissen and Werner, 2014). A hierarchical CF is indicated by a supply chain manager who focuses on alignment and economic issues but is willing to consider additional dimensions of sustainability if these align with other interests (Preuss and Fearne, 2022). Paradoxical sensemaking is beneficial for learning, but convenience might prevent managers from adopting paradoxical sensemaking (Pinnington and Meehan, 2023). In a paradoxical frame, the structure of the frame includes many connections between the frame elements so that flexible information processing and openness to new information are possible (Hahn et al., 2014; Suedfeld and Tetlock, 1977). In the case of a supply chain manager, a paradoxical frame allows multiple elements beyond cost or economic gains, making the manager aware of and likely to consider sustainability aspects regardless of their economic relevance (Preuss and Fearne, 2022). A paradoxical CF indicates that the manager can accept tensions between conflicting elements and can observe a wider variety of sustainability aspects; however, both time and other resources are needed for the manager to adopt a paradoxical CF (Hahn et al., 2014). Effective replacement of a unidimensional CF necessitates sensebreaking to pave the way for adopting a new frame (Menon, 2022). To engage with the paradox, a manager needs to allow malleable categorial boundaries and appreciate differences, as well as act to solve tensions collaboratively (Sharma and Bansal, 2017). Sustainability is driven by company policy and strategy, but individual managers are responsible for the implementation process (Preuss and Fearne, 2022), so their CFs manifest in operations (Hahn et al., 2014).

3.1 Embedded multiple case study

To explore how managers make sense of sustainable LMDs and what operational manifestations follow, this study adopted an embedded multiple case study approach to provide rich, nuanced data that can enhance the theoretical understanding of CFs in the last mile context. Case studies enable in-depth analyses and rich insights (Yin, 2018), and multiple case studies allow for the contrasting of results (Ellram, 1996). The focal point of the present study is the individual manager within an organization, but as Rousseau (2001) and Weick (1995) pointed out, there is a socialization aspect to cognitive processes. Therefore, managers can be seen as embedded cases within their respective organizations. The unit of analysis is the manifested sensemaking that can indicate the CF of a manager and thus explains how the manager makes sense of sustainable LMDs. The analytical lens is the framework in Figure 1, which is juxtaposed with the secondary data representing the organization’s view on sustainability and the view presented in the literature.

3.2 Selection of cases and data collection

The selection criteria in this study followed a purposeful criterion logic to focus on information-rich and high-quality cases (Patton, 2002). The case companies were retailers and LSPs, as they jointly executed LMDs. Retailers in consumer-oriented businesses were chosen because they were more likely to be interested in the sustainable activities of LSPs (Kudla and Klaas-Wissing, 2012). Companies of different sizes with different product segments representing a variety of retail environments and with a substantial presence in e-commerce were selected. The interviews were conducted during autumn 2021 and spring 2022, and retailers were interviewed first. The LSP selection criteria were that they worked with the retailers interviewed and had substantial LMD experience. The case company was first selected, followed by the managers to be interviewed. The manager selection criteria were that they were operationally involved with LMDs and had significant work experience in e-commerce. In addition, the managers needed to have positions of such managerial relevance that their opinions had strategic value within their organizations. These criteria were chosen because they implied that managerial CFs could be expected to have an impact on operations. All managers included in the study also shared the socio-political background and having working experience in European and Nordic organizations, so they could be assumed to share basic societal values. The managers were chosen to obtain rich information, and the selection of the appropriate respondents could be considered more important than their number (Saunders and Townsend, 2016). In this study, there were 13 cases: 8 retailers and 5 LSPs. An overview of the cases can be found in Table 2.

Semi-structured interviews were conducted to obtain new and previously unidentified insights (Patton, 2002). An interview guide was developed based on the last mile and sustainability literature, and feedback from other researchers was sought to ensure quality. The guide contained background questions to understand the professional background of the manager and the company, followed by questions about what the manager thinks about sustainability in LMDs, delivery strategies, general strategies and sustainability, as well as ad-hoc questions related to the respondents’ answers. Each interview was recorded and transcribed. As the use of secondary data can strengthen theory building by triangulation of evidence (Eisenhardt, 1989) and enhance the validity of a study (Yin, 2018), secondary data were collected to obtain a thorough understanding of each case company (Eisenhardt, 1989). Special attention was given to capturing data sources that allowed for investigating a broader range of attitudinal and behavioral issues (Yin, 2018). For the retailers, observations that were captured in screenshots of what delivery options they offered at checkout were collected.

3.3 Coding and analysis

Cognitive processes and structures are not obvious to outsiders or even the persons themselves, but they can be identified from both reports and behaviors (Scott et al., 1979). A simultaneous coding process, outlined in Figure 2, was utilized to explore the data. CFs are made up of two things: the context-specific content that can be described as attributes (Hahn et al., 2014; Menon, 2022; Scott et al., 1979) and the relationship between these attributes (Menon, 2022) that describe the CF structure (Hayes-Roth, 1977; Preuss and Fearne, 2022). Thus, the analysis process focused on simultaneously discerning between the CF context-specific content of the frames, here named content domain and the CF structure. The theoretical framework in Figure 1 provides three CF structures: unidimensional, hierarchical and paradoxical (Preuss and Fearne, 2022).

The context-specific content domain consists of elements that could be exemplified by specific operational activities to make LMDs sustainable, such as purchasing electric vehicles with the aim of emission reduction. Structural coding (Saldaña, 2009) was used in the first cycle to discern what content elements emerged from the data. As seen in Figure 2, these elements were grouped into themes utilizing pattern coding and further grouped into the CF content domain (Hahn et al., 2014). The outcome of this process is visible in the results section in Figure 3. The content domain can be described as the operations that managers associate with sustainable LMDs. This content was compared to the areas of LMD sustainability found in the literature (Table 1) and to the secondary data to understand what content and topics were not mentioned.

As the content domain is intertwined with how a manager might make sense of it and associate the different content elements, a simultaneous process of understanding what CF structures could be indicated in the data took place. During the first cycle of coding, in vivo coding of quotes was used to capture the managers’ languages, worldviews and perspectives (Saldaña, 2009). An abductive theory matching approach was then adopted (Dubois and Gadde, 2002) in coding cycle two to elaboratively match the in vivo quotes with the frame indicators in the theoretical framework (Figure 1). This analysis focused on determining whether similar frame indicator logic could be found in the empirical data. It is important to note that the in vivo quotes functioned as a chain, in which logic, structure and associations emerged throughout the interview, not in a single quote. The connections between elements and themes that emerged in the content domain coding supported this process. The resulting empirical frame indicators, presented in Table 4, describe how the frame indicators in Figure 1 appear when managers make sense of sustainable LMDs.

3.4 Quality of research

The quality of the study was ensured through various measures reported in Table 3. Multiple sources of evidence were utilized and a chain of evidence was established through the theoretically derived interview guide and the structured coding process, ensuring construct validity (Ellram, 1996; Yin, 2018). Quotes were used to support the chain of evidence and to strengthen the most important findings (Stuart et al., 2002). Objective data collection was feasible because of access to original data sources (Voss et al., 2002). Preliminary results were presented to respondents and retail and LSP managers at a webinar to validate the findings.

4.1 Operational manifestations of cognitive frames in sustainable LMDs

In this study, CFs are utilized as an analytical lens when describing the operative activities that managers associate with sustainable LMDs. This content seems to make up the content domains of a CF and can, in a last mile context, be categorized into three general themes: CO₂ and emission reduction, market impact and stakeholder collaboration (Figure 2). Analysis was also made about the content not described by the managers. It is apparent that environmental challenges are well acknowledged on a general level; however, the discussion lacks details and nuances that include, for example, biodiversity, and there is no discussion of air pollution or noise impact. Nevertheless, some positive impacts, such as LMDs being enablers of the circular economy, are highlighted. The managers did not include social sustainability aspects, such as accessibility for customers or the effects that LMDs can have on society at large.

4.2 Empirical frame indicators in sustainable LMDs

This section is a theoretically grounded description of the empirically derived frame indicators that describe how managers make sense of sustainability in LMDs. This section presents the CFs that the data indicate managers hold when making sense of sustainable LMDs. Empirical frame indicators were developed based on the theoretical framework (Figure 1) to discern the content and structure that could indicate a CF in a last mile context. These indicators, manifested in the sustainability actions and observed in the data, are presented in Table 4, together with a grouping of managers in accordance with how they seem to frame sustainable LMDs. For both retailers and LSPs, the main CF content considers environmental sustainability, paired with aligned topics that support the balance between business interests and emission reduction goals. The CF content domains (Figure 2) of all managers remained similar, but the relative importance of activities, the way content was discussed, and the importance of sustainability activities varied in accordance with either a more unidimensional or hierarchical structure.

Out of 12 managers, 8 indicated predominantly hierarchical CFs, in which sustainability is an important topic but is subordinate to economic interests. Four managers indicated thinking that aligned with unidimensional CFs. Although some managers acknowledged the complexities, none of the respondents could be said to indicate paradoxical CFs regarding sustainable LMDs. Common for the managers indicating unidimensional CFs was focusing on LMD optimization for profit reasons rather than environmental considerations. They all shared instrumental and profit-driven views of LMDs and had little interest in sustainability. They also tended to view other actors and stakeholders more as competitors than as partners. Common among managers indicating hierarchical CFs was that their companies had sustainability as one of their core values, but the operational execution showed that economic interest was prioritized, even if the managers themselves acknowledged the importance of environmental sustainability. These managers had more positive views of stakeholders as sustainability collaborators. There was no notable variation in content or structure that would differentiate between LSPs and retail managers.

6.1 Limitations and future research

Current research on paradoxes widely shares the assumption that paradoxical frames or mindsets are of benefit to organizations. While this study adheres to this assumption, it can be noted that there is research indicating otherwise. The geographical context of this study, which is primarily Nordic, accounts for some study limitations and might influence the content domains of managerial CFs. The Nordic welfare context might impact managers’ views of social sustainability. This study is also limited to individual managerial frames and provides a situational snapshot. The dynamic process of CFs that are likely to change over time is a limitation of this study. Further research could consider the evolving dynamics of CFs and sensemaking over time and under varying conditions through conducting longitudinal studies with contextual data and alternate data collection techniques. As CFs are malleable and can blend or transition (Cornelissen and Werner, 2014) or can function as mediators for organizational or social change (Benford and Snow, 2000), the socialization and organizational perspectives of CFs within LMDs and their impacts on sustainability provide ample opportunities for further research.