From purchase to pantry – exploring archetypes and strategies in the context of e-grocery fulfilment

ABSTRACT

Fulfilment is a key driver for the digitalisation of the grocery sector and thus a major source of value creation. Although value creation is contingent on a multiplicity of contextual factors, extant research has mainly explored fulfilment in the generic realms of e-commerce and does not provide empirically grounded insights into the collective interplay of fulfilment attributes. Because e-grocery features unique exigencies and trade-offs that require non-generic theorisation, a functional and cross-domain perspective is required to comprehend the role of fulfilment in creating value in this sector. Following a multi-method approach, this paper combines a literature review with an empirical investigation of 111 e-grocers to produce three main outcomes. First, a conceptually-grounded and empirically-refined taxonomy to capture important decision attributes of e-grocery fulfilment. Second, six archetypes to indicate common combinations of these attributes. Third, a contingency model to disclose value relationships between attributes and outcome performance. Our results contribute to the body of knowledge on e-grocery value creation as well as the explanatory understanding of operational configurations and digital transformation conditions in this industry.

KEYWORDS:

• Digital transformation
• contingency theory
• e-grocery
• archetypes

1. Introduction

The ubiquitousness of digitisation requires the retail sector to adapt its services, business model(s) and infrastructure (Trenz et al., 2020). Advancements in information technology (IT) have allowed retailers to integrate stationary infrastructure with online-channels and invent novel, pure-digital business models (Gu & Tayi, 2017). As a consequence, the online-channel has become indispensable for the retail landscape. Driven by changes in customer expectations and recent events such as the COVID-19 pandemic, e-grocery is a prospering retail opportunity space that yields economic and societal benefits (Milioti et al., 2021). With an increase of 1.2 billion USD from 08/2019 to 7.1 billion USD in 03/2021 (Statista, 2021) and a projected compound annual growth rate of 9.31% until 2030 (Verified Market Research, 2021), the e-grocery market is attractive for both, established brick-and-mortar retailers and digital players. Implementing novel business models or transforming existing e-grocery business models, however, is a complex task that is contingent on multiple factors. E-grocers need to reflect on specific challenges, including delivery constraints (e.g., preservation of fresh produce, Hays et al., 2005), diverse delivery schemes, limited shelf lives of perishables, and low margins (Enders & Jelassi, 2009). Accordingly, they face a heterogeneous set of options concerning order fulfilment in particular – a fact that is even accentuated in the digital age (Mithas et al., 2016; von Viebahn et al., 2020).

Information systems (IS) research has suggested that fulfilment attributes such as lead times can have a remarkable effect on the performance of online operations (Qureshi et al., 2009; Saeed et al., 2005). Yet, developing e-grocery fulfilment (EGF) strategies – the combination of attributes – is impeded through a series of context-dependent trade-offs. For example, the minimisation of inventory costs might lead to long lead times and vice versa, whereas short lead times require large quantities of stock. The impact of flawed decisions can be illustrated by various real-life examples, such as the Silicon Valley-based e-grocer Webvan. Webvan built fulfilment centres with advanced robot technologies to deliver groceries within 30 minutes of ordering. After being valued at 1.2 billion USD, Webvan went bankrupt within a few months. Among other reasons for their bankruptcy, they failed to balance efficiency and value creation. For instance, they built fulfilment infrastructure from scratch instead of using infrastructure through cooperation and they sold to the mass market instead of targeting upmarket consumers who would be more profitable in the face of low profit margins (Cleophas & Ehmke, 2014).

Because digital transformation can mitigate these trade-offs, choices made to achieve fulfilment efficiency need to be aligned with technological choices (Davis-Sramek et al., 2020; Milioti et al., 2021; Sandberg et al., 2014). Therefore, research stresses the importance of integrating digital transformation with order fulfilment (Lim et al., 2018). Given that “the grocery retail sector is generally known to pioneer the adoption of innovative technologies to enhance efficiency” (Kurnia et al., 2015, 1906), previous studies have contributed to our understanding of the use of digital technology for order fulfilment. Among others, this includes self-service technologies to increase shopping convenience (Saldanha et al., 2021) and Blockchain to foster process efficiency (Yacoub & Castillo, 2022). However, although several studies have attempted to conceptualise individual fulfilment attributes or supporting technologies (Appendix A), our review of the extant literature on value creation in e-grocery (Table 1) indicates that research in this area is still fragmented.

Table 1. Contextual factors for value creation in e-commerce fulfilment.

Category		Example references
Context: E-commerce (e.g., fashion)
Influencing factors	Automation, assortment size, buyer uncertainty, channel, company objectives, consumer habits, digital capabilities, digital opportunities, logistics capabilities, logistics information systems, order volumes, product characteristics, routing, supply chain networks, warehouse configuration	Chircu and Mahajan (2006), Cotteleer and Bendoly (2006), Davis-Sramek et al. (2020), Y. He et al. (2019), Kembro and Norrman (2019), Kembro and Norrman (2021), J. Luo et al. (2016), X. Luo et al. (2021), Pavlou and Fygenson (2006), Pavlou et al. (2007), Polites et al. (2018), Saldanha et al. (2021), S. Scott and Orlikowski (2022), Zhang et al. (2021)
Value outcome: Consumer response	Behavioural (re-purchase) intention, channel adoption, commitment, loyalty, satisfaction, trust, willingness to pay	Deng et al. (2010), Hulland et al. (2007), Lee et al. (2003), Milioti et al. (2021), Otim and Grover (2006), Qureshi et al. (2009), Shankar et al. (2002), Tan et al. (2016), Trenz et al. (2020), Wollenburg, Holzapfel, et al. (2018)
Value outcome: Organisational impact	Costs, managerial actions, revenue	Bressolles and Lang (2019), Fang et al. (2021), Hulland et al. (2007), Jeyaraj et al. (2009), Li et al. (2019), X. Luo et al. (2020), Saeed et al. (2005)
Context: E-grocery
Influencing factors	Communication technology, competition, logistics capabilities, order volumes	Hardgrave et al. (2008), Müller-Lankenau et al. (2005), Tang et al. (2004), Yacoub and Castillo (2022), Wollenburg, Hübner, et al. (2018)
Value outcome: Consumer response	Behavioural intention, loyalty	Alagarsamy et al. (2021), Boyer and Hult (2006)
Value outcome: Organisational impact	Costs, service quality, product quality, revenue	Boyer and Hult (2006), Enders and Jelassi (2009), Kurnia et al. (2015)

Following calls to investigate the big picture (Hübner et al., 2016) and given that operational trade-offs confine value creation (Barua et al., 2004), there is a demand for a comprehensive understanding of EGF. Moreover, and even more importantly, in light of the immense number of EGF choices that can potentially be effective, it remains unclear as to why and how different combinations create value. IS scholars agree that value creation is contingent on a multiplicity of attributes and their interplay (Rai et al., 2012; Trenz et al., 2020). Yet, in the case of EGF, there is no holistic understanding of the effectiveness of EGF attributes and their interdependencies for value creation. This can be illustrated by looking at recent findings from research and practice, which have provided rather contradictory insights on individual attributes. For example, whereas Anshu et al. (2022) and Qureshi et al. (2009) highlighted delivery experience as a major driver of purchase intentions, Otim and Grover (2006) did not uncover any effects of delivery arrangements on (re-)purchase intentions. Similarly, the use of automated fulfilment centres and short delivery lead times did not work out for the aforementioned e-grocer Webvan, whereas it is highly profitable for Amazon and Tesco (Cleophas & Ehmke, 2014). In response to the ambiguous knowledge base, there is an imminent need to conceptualise the combination of EGF attributes as an influencing factor for value creation in e-grocery.

Due to the plurality of feasible decisions as well as the highly dynamic environment (Van de Ven et al., 2013), this paper builds upon a contingency view (Luthans & Stewart, 1977). This view considers the individual business contexts in order to identify a combination of attributes capable of achieving an intended outcome. Researchers thereby develop functional relationships between the environment and the performance to cope with different situations (Miller, 1981). Following this idea, it is not about finding the one optimal combination of attributes, but rather a combination that reflects the current (external) setting and its goals (Turedi & Zhu, 2019). Consequently, we ask: What are the attributes of e-grocery fulfilment and how does their combination affect the value creation of an e-grocer?

Drawing upon interdisciplinary literature related to e-commerce, this paper opts to categorise EGF attributes and their combinations. By linking value creation to EGF attributes under consideration of fulfilment combinations, competitive focus, and organisational assets, we uncover distinct value relationships and provide a broad foundation for theorisation. For this purpose, we develop a contingency model for order fulfilment that relies on a review of 80 scientific literature sources and an empirical investigation of 111 e-grocers. The contribution of our paper is threefold. First, we organise EGF attributes through a taxonomy. Second, six EGF archetypes are proposed that represent the main contingency factors. Moreover, these archetypes advance the understanding of trade-offs and pave the way for designing comprehensive solutions for the digital transformation of the grocery sector. Third, we present a contingency model to disclose value relationships between EGF attributes and performance outcomes. This model elucidates the moderating role of EGF archetypes as well as archetype-related competitive focus and organisational assets as contingency factors for value creation in e-grocery.

In the ensuing sections, we first outline a synopsis of the related work and theoretical background to introduce the main constructs relevant to our contingency model. Following our research approach, we present the findings of our study. Finally, we discuss the paper’s implications and outline its limitations as well as future research directions.

2. Research background

2.1. Fulfilment-based value creation and influencing factors

The prosperity of e-commerce has fuelled scholarly interest in exploring online-channel value creation (Qureshi et al., 2009; Tan et al., 2016; Trenz et al., 2020). From our review of the extant literature on value relationships in e-commerce, we have found that fulfilment is essential for value creation (Bressolles & Lang, 2019; Davis-Sramek et al., 2020) and a prerequisite to successful transactions of any kind (Otim & Grover, 2006). Fulfilment not only fosters consumer trust (Shankar et al., 2002), but also impacts a company’s financial performance, operational costs, channel satisfaction and loyalty (Barua et al., 2004; Deng et al., 2010; Wollenburg, Hübner, et al., 2018). Thereby, its particular role and impact are contingent on numerous influencing factors, such as assortment size (Y. He et al., 2019), network effects (X. Luo et al., 2021), service technologies (Saldanha et al., 2021), and consumer habits (Zhang et al., 2021).

Owing to the competitive relevance of order fulfilment (Fang et al., 2021), the majority of work in the IS domain and adjacent fields elaborates on cause-effect relationships for service-related attributes. These range from general influencing factors that affect fulfilment performance to different value outcomes in the form of consumer responses to e-fulfilment and the organisational impact of specific e-fulfilment strategies (see overview in Table 1). It illustrates that existing research has primarily focused on individual factors, rather than conceptualising fulfilment combinations (i.e., the interplay of attributes) holistically as a predominant moderator in the value creation process. Moreover, it shows that value creation and its influencing factors in the generic realm of e-commerce have been commonly investigated, whereas dependencies in the context of e-grocery are rather understudied. This is problematic as e-grocery significantly differs from other e-commerce contexts, such as fashion or electronics.

From a consumer viewpoint, grocery shopping involves repeatedly purchasing the same products with a low degree of involvement. In contrast to general online shopping, which produces feelings of enjoyment as customers look for novel or exclusive products (Jiang & Benbasat, 2007), grocery shopping is perceived as a routine task (Dawes & Nenycz-Thiel, 2014). Thus, a major reason for using e-grocery is its convenience in terms of speed, effort and flexibility (Trenz et al., 2020; Zhang et al., 2021), making service attributes such as delivery velocity highly important for competitive differentiation (Anshu et al., 2022). From an operational standpoint, the nature of the goods sold creates unique logistical challenges. EGF needs to ensure the visual and tactile integrity of fresh produce, while adhering to industry-specific regulations like maintaining the cold-chain (Hays et al., 2005). Contrary to traditional store-based fulfilment, e-fulfilment systems process a high number of orders with low product quantities (Bressolles & Lang, 2019). Thus, e-grocers need to mitigate cost pressure (Agatz et al., 2008; Enders & Jelassi, 2009).

Yet, the choice of EGF attributes entails operational and strategic trade-offs, including inventory and delivery costs or productivity and convenience (J. E. Scott & Scott, 2008). Consider, for example, the case of “click-and-pick” e-grocers, which collect groceries from third-party supermarkets and deliver them to the customers. On the one hand, this approach features distinct advantages such as large service areas with low financial requirements (due to the absence of own fulfilment infrastructure) and short lead times (due to a large network of fulfilment points). On the other hand, it also comes with limited control (due to the dependence on third parties), low operational efficiency (as order management routines are not aligned with delivery operations), and a high degree of competitive threat (as partnering supermarkets can choose to replace or discard their affiliates). Facing multiple fulfilment attributes, structural comprehension on an individual and a holistic basis is subject to intense research (Pereira & Frazzon, 2021; von Viebahn et al., 2020). Thus, in line with the unique business setting of e-grocery, there is a fundamental need to disclose and explain value relationships in this sector, rather than adapting insights from other e-commerce contexts.

Based on the extant literature on value relationships in e-commerce, we argue that the combination of fulfilment attributes and associated capacities constitute major contingency factors in the grocery sector. This is because the interdependencies of attributes and strategies are key influencing factors for a grocer’s online performance (Hulland et al., 2007; Otim & Grover, 2006; Qureshi et al., 2009). As an individual fulfilment attribute (e.g., fulfilment point) can influence value relationships in different ways, its impact on value outcomes can only be explained when considering a combination holistically. For instance, depending on the type of order reception, fulfilment from stores can either generate a low (in the case of self-collection) or a high degree of customer value (in the case of home-deliveries). In turn, these value relationships are subject to additional interdependencies and trade-offs (e.g., the competitive focus and organisational assets of an e-grocer).

Whilst the literature on e-grocery is advancing, it is still unclear as to where certain combinations fit better than others and which value outcomes are related to these combinations and associated capacities. Considering the heterogeneity of order fulfilment and related services (von Viebahn et al., 2020), it is important to integrate past findings into a model that explains how certain combinations influence the value outcomes holistically. In doing so, relevant trade-offs can be disclosed and the role of fulfilment for e-grocery success can be explained.

Focusing on the implications of EGF combinations, our research complements IS literature on multi- and omni-channel retail (Chen et al., 2018; Trenz et al., 2020), e-commerce value relationships (Fang et al., 2021; Polites et al., 2018), and digital technologies for order fulfilment (Cotteleer & Bendoly, 2006; Hardgrave et al., 2008). E-grocery value creation is an instance of e-commerce performance and design within a setting that combines IT-enabled customer value with operational elements. Therefore, it requires distinct theorisation within the IS literature. By organising e-grocery contingency factors, we create an understanding of the interrelationship between fulfilment and value outcomes (e.g., channel adoption and costs), which additionally lays the foundation for developing integrated IT solutions to meet the requirements of specific e-grocery business models.

2.2. Capacities and performance outcomes related to e-grocery fulfilment

Although the link between fulfilment and e-commerce success has been investigated in prior studies (Barua et al., 2004; Deng et al., 2010; Fang et al., 2021), there is scant research on value relationships in e-grocery. However, the distinct logistical requirements of grocery items (e.g., cooling) and the unique customer requirements create an environment where fulfilment is at the core of success (Boyer & Hult, 2006; Müller-Lankenau et al., 2005; Wollenburg, Hübner, et al., 2018). To address this, we conceptualise e-grocery value as an outcome that is contingent on EFG attribute combinations as well as the associated focus and assets of an e-grocer. This allows us to explore the potential impact on operational efficiency and customer value and improve the collective understanding of value creation in grocery.

2.2.1. Capacity constructs: focus and assets

Strategic integration refers to the harmonisation between organisational resources and competitive strategies and is vital for a company’s performance (Rivard et al., 2006). Although organisations require distinct assets to pursue a given strategy (Barney, 1991), resources per se do not explain the differences in a company’s performance, but rather the exploitation of these resources based on an e-grocer’s objectives and capacities (Boyer & Lewis, 2002). In e-grocery, resources are directly linked to strategy and business success (Lynch et al., 2000). Accordingly, competitive focus (i.e., objectives) and organisational assets (i.e., resources) need to be assessed collectively to elaborate on the interdependencies between individual EFG attributes and performance outcomes under consideration of attribute combinations and capacities.

In scientific research, the competitive operations focus of a company is commonly operationalised through competitive priorities. According to Krajewski and Ritzman (2002), organisations can choose to prioritise costs, quality, delivery, and flexibility. The individual combination of priorities delineates a company’s logistics capacities and characterises its business strategy (Stock et al., 1998). When adopting a cost focus, e-grocers seek to compete based on low costs, for example, by reducing inventories or increasing productivity to minimise handling and transportation costs (Onstein et al., 2019). In turn, a quality focus describes the emphasis on EGF resources to compete on the basis of operational excellence (Krajewski & Ritzman, 2002). These organisations attach importance to capacities such as automation and staff training (Fernie et al., 2010; Hübner et al., 2016). As operational quality characteristics primarily relate to efficiency metrics (e.g., picking efficiency) in the value relationship between EGF attributes and performance (Boyer & Hult, 2006), we refer to this construct as efficiency focus in our contingency model. A delivery focus reflects an e-grocer’s intention to outperform the competition by offering superior (e.g., velocity) distribution routines (Krajewski & Ritzman, 2002). Nonetheless, with convenience and service level being the main differentiators for customer value in e-grocery (Anshu et al., 2022), we refer to this as service focus to highlight the role of service-related contingencies for value creation. Finally, flexibility focus represents the ability to (re-)deploy fulfilment resources in response to unforeseen changes (e.g., demand fluctuation) (Krajewski & Ritzman, 2002). Correspondingly, flexibility capacities include supply chain collaborations and internal infrastructure (e.g., delivery fleet) to enable an e-grocer to respond to changing exigencies (Fernie et al., 2010).

These competitive objectives can either complement or contradict each other. A pronounced focus on efficiency may help to foster other foci such as greater flexibility or lower costs. Thereby, trade-offs may arise that force e-grocers to prioritise based on the available distribution structures (Onstein et al., 2019). For instance, to minimise costs, e-grocers need to choose optimal order quantities, which jeopardise the availability of products and finally decrease the service focus.

In another capacity, EGF directly relates to physical and financial assets. Although the broad definition of assets additionally encompasses capabilities as intangible assets (Barney, 1991), we adopt a narrower definition for our contingency model. This allows for a clear distinction between assets and focus, which seems particularly suitable to explain contingencies in the case of EGF (Ellis-Chadwick et al., 2007). Physical assets can be defined as “stocks of available factors owned or controlled by the firm” (Amit & Schoemaker, 1993, p. 35). In the case of EGF, relevant physical assets include supermarkets, distribution facilities, inventories, IT systems, and delivery vehicles (Hübner et al., 2016; Prasarnphanich & Gillenson, 2003). Financial assets represent monetary resources to finance investments, capital, and business activities (Barney, 1991). They drive business performance by enabling companies to acquire physical assets to improve their operational capacities concerning costs, efficiency, service, or flexibility (Anckar et al., 2002; Mithas et al., 2016). Moreover, they can be employed to improve operational capacities directly, for example, when they are invested in staff training to drive flexibility. Correspondingly, assets are interrelated with each other and with the focus of an e-grocer.

2.2.2. Outcome constructs: operational efficiency and customer value

The purpose of aligning EGF combinations with competitive focus and organisational assets is to increase efficiency (Wollenburg, Holzapfel, et al., 2018). Different capacities have a direct impact on an e-grocer’s ability to handle orders (Boyer & Hult, 2006). A majority of e-grocers have been concerned with improving inventory, order, delivery, and return management practices to unlock potential efficiency gains and achieve a high degree of operational efficiency (Bressolles & Lang, 2019; Y. He et al., 2019). By considering the organisational value, operational efficiency reflects the internal value creation perspective in our contingency model that depicts the role of EGF archetypes as well as focus and assets for operational performance metrics (e.g., stock turnover rate).

Another higher-order outcome of an e-grocer’s focus and assets, which are related to its EGF combination, is customer value. It is defined as “the trade-off between the quality, or benefits, which the customer receives and the costs, such as monetary, energy, time and psychic transaction costs, which the customer incurs by evaluating, obtaining and using a product [or service]” (Chircu & Mahajan, 2006, p. 901). Because e-grocery service propositions directly depend on EGF attributes and their combinations (e.g., short lead-times require dense fulfilment networks and IT support), customer value is contingent on fulfilment structures (Boyer & Hult, 2006). Companies that can excel on post-purchase services are likely to create superior shopping experiences, which generate customer value (Saeed et al., 2005). In line with its inherent emphasis on market performance, customer value reflects the external value creation perspective in our model. It illustrates the impact of EGF archetypes and capacities on demand-related performance metrics, such as channel adoption, satisfaction, and loyalty.

2.3. Contingency theory

Facing a high number of decision attributes and the highly dynamic market situation, it is challenging to find an “optimal” strategy (Q. He et al., 2020; Zott & Amit, 2008). With the “growing demand for robust theorising and empirical research on new forms of organising ever-more-complex and dynamic situations” (Van de Ven et al., 2013, p. 394), contingency theory has received a resurgence. In the pursuance of contingency theory, organisations need to understand the variety of changing environmental contingencies and their dependencies to effectively align strategies with competitive surroundings (Crișan & Stanca, 2021; Iivari, 1992). There is, however, no single best way to do this because decisions made in some situations may not be successful in others as they are affected by several contingencies (Fiedler, 1964). Contingencies refer to any attribute that moderates the relationship between organisational characteristics and performance, such as organisation size, product variety, or service capacity (Turedi & Zhu, 2019). The theory argues that the most effective management concept depends on the set of circumstances at a particular point in time (Luthans & Stewart, 1977; Weill & Olson, 1989).

Referring to this paper’s context, we see similarities to contingency theory given that digital transformation in the e-grocery sector requires substantial changes, and e-grocers need to handle a variety of strategic options, particularly in the area of fulfilment. Yet, there are limited insights on trade-offs in implementing e-grocery business models and appropriate digitalisation strategies. Given the relevance of understanding the environment and aligning strategies accordingly, the theory has been employed in domains related to our context, including e-commerce (Giaglis et al., 2002), last-mile logistics (Lim et al., 2018), technology strategy (Teo & Pian, 2003), and digital transformation (Crișan & Stanca, 2021). The contingency theory lens offers an expedient framework to evaluate the multiplicity of EGF attributes and explain the corresponding value relationships in the e-grocery domain. In view of the fact that it theorises that organisational strategy and value creation depend on contextual conditions, we consider it as a viable underpinning to create a comprehensive understanding of the (potential) role of EGF archetypes and associated capacities for different value outcomes in e-grocery.

Building on the contingency lens, Figure 1 outlines the conceptual model for this study. We conceptualise value relationships based on attribute combinations from order management, which encompasses attributes that relate to backend operations (inventory handling, order preparation) and delivery management, which describes distribution characteristics (delivery planning, delivery operation, return handling) as well as from the focus and assets of e-grocers. These contingency factors moderate the relationship between individual fulfilment attributes of an e-grocer and the outcome constructs customer value and operational efficiency.

Figure 1. Conceptual model for this study.

3. Research approach

3.1. Identifying e-grocery attributes

As a first step, we designed a taxonomy to disclose and organise e-grocery attributes (Nickerson et al., 2013) (see details in Appendix B). A taxonomy aids in structuring conceptual peculiarities and interdependencies (Wand et al., 1995), thus serving as a reference architecture that is suitable for comparing different conceptual elements such as fulfilment models. We performed two conceptual iterations in which we integrated attributes from prior conceptualisations and six empirical iterations in which we coded EGF instances obtained from the literature (Appendix C) and 111 e-grocers (Appendix D). Accordingly, we designed a conceptual-grounded taxonomy and extended it through an empirical investigation. The empirical focus corresponds to related studies dealing with creating taxonomies for strategic decisions (Malhotra et al., 2005; Segars & Grover, 1999).

In order to investigate the taxonomy’s usefulness, we evaluated its comprehensibility as well as its ability to support the classification of existing and the development of new EGF combinations through interviews and workshops (Kundisch et al., 2021). Here, we found (1) that e-grocers were able to classify their fulfilment strategies as well as (2) multiple statements from e-fulfilment specialists regarding the taxonomy’s high degree of practical usefulness (e.g., “[A] great tool to develop new fulfilment models in the grocery sector.”).

3.2. Developing EGF archetypes

Based on the taxonomy, we applied cluster analysis to extract common attribute combinations in the form of EFG archetypes (e.g., comparable to Vom Brocke & Lippe, 2011). Considering the number of compiled attributes, “clustering is the only method that can adequately construct the necessary typology” (Bailey, 1989, p. 17). We coded each of the final taxonomy’s attributes that was observable for one of the 111 e-grocers with a 1 and a 0 otherwise. As a result, a comprehensive classification of each of the e-grocers was created.

Unlike other studies that have derived clusters based on individual taxonomy dimensions to improve manageability (e.g., Malhotra et al., 2005), we conducted cross-dimensional cluster analyses with all attributes to grasp fulfilment-related trade-offs via useful archetypes. We followed a two-step clustering process, consisting of Ward’s (1963) hierarchical clustering algorithm to identify the required number of clusters and K-means++ to assign e-grocers and attributes to the clusters. This procedure is widely accepted in IS research (Balijepally et al., 2011) and has been applied in similar research contexts with a moderate sample size and an extensive set of attributes (King & Sethi, 1999). Because the decision concerning the number of clusters entails conflicts between the solution’s manageability and the cluster’s homogeneity, we tested the robustness of the three- and six-cluster solution by comparing the means of each solution through K-means++. Although neither solution featured a significant difference for all attributes across all clusters, both showed a high degree of relative discriminatory power with significant differences across clusters (Appendix E.1). To account for the present and future EGF heterogeneity (Martín et al., 2019) and ensure that our empirical analysis represents a representative set of concept instances (Malhotra et al., 2005), our final selection of the six-cluster solution was then based on a priori theory (Ketchen et al., 1993).

To ensure external cluster validity, we conducted multiple evaluation episodes with seven representatives of an omni-channel-grocer, a pure-online-grocer, and three e-fulfilment specialists, who indicated that our findings are congruent with their operating models and likely to reflect common industry standards. Following the three think-aloud sessions and workshops that were conducted to evaluate the EGF taxonomy, we evaluated the archetypes through seven expert interviews and two workshops (Appendix B). First, during seven online interviews lasting between 21 and 45 minutes, the archetypes were introduced to the experts visually and verbally. Following Balijepally et al. (2011), we asked participants to rate the applicability (“I find the archetypes useful”) and validity (“I think the archetypes show a realistic combination of EGF attributes”) on a five-point Likert-scale (1 – strongly disagree, 5 – strongly agree), indicating that the clusters are highly applicable (Ø 4.71) and valid (Ø 4.86). Second, during workshops (lasting between 44 and 96 minutes), which were performed with two representatives at the same time, practitioners were asked to use the archetypes to solve a fictive task. They collectively had to develop a fulfilment model for a grocery organisation planning to establish an online sales channel, while the main points of discussion were documented manually by the research team. Upon completion of the workshop, we collected verbal feedback on the applicability and usefulness of the archetypes for solving the task.

Based on the interview transcripts and the workshop logs, the feedback was consolidated and applied to improve the archetypes. Experts stressed the specific strategic relevance of the attribute combinations; for instance, concerning decision-making (e.g., “[A] great tool to find or refine a fulfilment strategy and establish competitive advantages.”). In other words, the archetypes provide support in the actual choice of EGF combination, identification of trade-offs and analysis of improvement potentials in a sense that provides competitive delimitation.

Afterwards, we employed abductive reasoning to derive the capacity constructs (i.e., focus and asset) for each of the derived archetypes based on the insights from the literature and the empirical analysis of e-grocers.

3.3. Deriving a contingency model

Finally, to theorise the value relationships, outcome constructs were identified by studying three selected e-grocers of each cluster through documentation and surveys. Operationalisation of the constructs was done via questionnaire items (measured on a 5-point Likert scale) that have been derived from literature: namely costs, productivity, stock turnover rates, and picking accuracy in the case of operational efficiency (Bressolles & Lang, 2019; Wollenburg, Hübner, et al., 2018), as well as delivery velocity, transaction costs, convenience, and product quality in the case of customer value (Chircu & Mahajan, 2006; Saeed et al., 2005).

A synopsis of the research process is given in Figure 2 (see Appendix B for a detailed overview of the entire research process).

Figure 2. Research process.

4. E-grocery fulfilment archetypes

The focal centre points of each cluster depend on the attributes of the EGF taxonomy. Figure 3 synthesises the compiled order and delivery management attributes, including the respective occurrence frequency of an attribute across all e-grocers (see Appendix E.1 for individual frequencies per archetype). In a top to bottom sequence, the taxonomy reflects the journey of an order within the EGF process and comprises five meta-dimensions: Inventory handling covers relevant attributes around inventory planning, zoning of storage areas, governance of inventory, allocation of inventory to different business units, and inventory reservation times. Order preparation is composed of warehouse activities in which items are collected from their storage areas to physically prepare a customer order for shipping. It captures infrastructural attributes (e.g., point of fulfilment and degree of picking automation), procedural attributes (e.g., picking procedure and out-of-stock handling), and strategic attributes (e.g., fulfilment point governance and order threshold). The meta-dimension Delivery planning deals with planning tasks and frame conditions that are related to the physical delivery process of a grocery order. It involves, for instance, the delivery area in terms of geographical location and land arrangement, duration of lead-times, order packaging, delivery governance, and routing approaches. Delivery operations relates to the actual process of physical delivery, including information on the fleet that is to be utilised for the delivery process, cooling procedures, delivery intermediaries, tracking activities, and routines employed to hand over an order to a customer, as well as additional delivery services. Finally, Return handling describes the process of order retraction and complaint management. It differentiates attributes that are concerned with the return procedure, recovery for failures, and the integration of returned goods.

Figure 3. Taxonomy of fulfilment attributes in e-grocery.

In the following section, we describe the EGF archetypes to improve the theoretical understanding of attributes, relationships, and trade-offs of different combinations for value creation (see Appendix C for a description of each attribute). An overview of real-world e-grocers for the six archetypes can be found in Appendix E.2.

4.1. Archetype #1 – Pick-up service proponents

These e-grocers have built on the provision of click-and-collect services that allow consumers to order groceries online and have them prepared for collection at store outlets, affiliated facilities or automated lockers. Compared to stationary shopping, this concept offers a high degree of access convenience. However, as multi- and omni-channel-practice, it requires a sophisticated network of physical assets. Typically, dedicated in-store collection facilities or service-oriented drive-in facilities are provided where pre-ordered items are directly loaded into the customer’s vehicle. Concomitant with the store-based reception type, store outlets, pick-run optimised front centres or an operational mix between store-outlets and distribution centres serve as main fulfilment points, with the e-grocer being in charge of inventory, fulfilment point, and delivery governance. The multi- and omni-channel grocers share inventories between offline- and online-channel and plan capacities based on demand forecasts or hybrid routines.

Orders are picked manually or semi-automated from individual store locations by piece or from front/distribution centres by case over a working day. Due to the fact that orders are collected rather than delivered, order thresholds are absent and lead-times between order placement and reception are short. Because of their sophisticated offline infrastructure, pick-up service proponents such as Delhaize in Belgium can offer grocery collection services on a regional and national scale. However, due to the manual, store-tailored operations, the preparation of individual orders for collection is labour-intensive. Hence, e-grocers adopt moderately long time-windows to capitalise on bundling effects and allow sufficient time for order preparation (Agatz et al., 2008). Given the absence of delivery operations, e-grocers frequently desist from charging additional service fees or offering hybrid incentives, whilst shorter lead times or narrower time windows are associated with dynamic pricing mechanisms. To cope with the lack of service value through deliveries, they provide additional services such as storage assistance. In order to ensure convenient collection, goods are wrapped in disposable packages such as plastic bags. With regard to complaint handling and return procedures, customers can directly claim a monetary refund from the collection location.

4.2. Archetype #2 – Parcel champions

Parcel champions rely on unattended deliveries with delivery boxes. They outsource the delivery governance to third-party couriers, which, depending on the respective delivery area and scope, employ light commercial vehicles (LCV) or heavy duty trucks (HDT) to transfer the items to their facilities close to the order recipient. Here, groceries are transshipped to LCVs and delivered to the customer. This archetype is operated by pure-online- and multi-/omni-channel-grocers, whereby the latter plan, utilise and pick inventories individually for offline- and online-channels (offline-channel: entire assortment; online-channel: dry foods). They benefit from reduced packaging requirements and costs as dry foods do not require cooling. In contrast, grocers with a larger product assortment, including fresh produce and chilled goods, need to invest in insulated packaging to maintain the cold chain. To achieve operational efficiency, parcel champions such as Rossmann (Germany) occupy pick-run optimised distribution centres. These are either centralised, decentralised and separated from offline-channels or decentralised and integrated into offline-channel operations. Orders are picked by case or pallet at generic times (perishables need to be prepared at different intervals to avoid spoilage), utilising semi- or fully-automated systems.

In order to cope with the requirements regarding physical (number/size of distribution centres) and technological (picking technology, channel-based order preparation) assets, this approach requires a sophisticated order management infrastructure (Wollenburg, Hübner, et al., 2018). In line with the courier-based nature of this concept, orders are shipped in reusable packages. Whereas courier companies provide estimated or real-time tracking information, e-grocers do not offer value-added delivery services themselves. Utilising the fulfilment infrastructure of delivery partners, parcel champions can benefit from efficient and reliable delivery routines, including dynamic routing, and can offer EGF on a national/international level for urban, sub-urban and rural areas. Nevertheless, relying on external partners to serve a large delivery area also entails drawbacks, because e-grocers are bound to the operational hours of their fulfilment partners and have to accept longer lead times and unspecific delivery windows. To account for the accruing delivery service costs, e-grocers raise dynamic threshold-, location- or subscription-based fees. For customers that opt to claim a refund, they either eschew return procedures and provide a claim-based, monetary compensation or offer returns via a designated courier.

4.3. Archetype #3 – Click-and-pick experts

Organizations in this cluster focus on outsourcing; inventories and fulfilment points of the pure-online-channel-grocers are governed by contractors or third-parties. Due to the implicated lack of control over stocks and fulfilment infrastructure, e-grocers generally refrain from sophisticated planning routines and handle out-of-stock situations by (partial) cancellation. Inventories are reserved at pre-designated time slots and picked from stores. E-grocers act as an intermediary between customers and supermarkets and manually pick online-orders by piece from selected outlets in close proximity to the recipient. Correspondingly, they predominantly serve local or regional areas, where the product reception is attended and items are delivered by cargo-bikes, crowd-shipping or LCVs. For national deliveries, third-party couriers are employed. As click-and-pick experts such as Com4Buy (Germany) are able to pick from a wide variety of third-party supermarkets, they can serve numerous delivery areas.

Although the widespread fulfilment point infrastructure and rigorous out-of-stock handling routines allow for short lead times and time windows, order reservation times, manual picking procedures, the absence of planning routines, and the need to rely on third-party couriers for national deliveries relativise these advantages and result in moderate lead times and delivery time slots. Moreover, click-and-pick e-grocers are dependent on the existing store infrastructure of the respective fulfilment points (Hays et al., 2005). Delivery services are either settled by a fixed service fee, a threshold-based pricing system or a hybrid pricing scheme (e.g., threshold-based/fixed pricing for local/regional consumers). Cooling requirements are met by employing cooling boxes or vehicles (local deliveries only). Regarding complaints and returns, click-and-pick experts either have service agreements with fulfilment partners that allow consumers to return items to the supermarket they have been picked from or refuse returns and provide an instant discount. Following the third-party based governance, return flows are generally separated from the e-grocer’s inventory system.

4.4. Archetype #4 – Collaboration specialists

Collaboration specialists leverage large partner networks and IT to ensure efficient order management routines, enabling geographically widespread deliveries. Depending on regional peculiarities and demand patterns, they operate a regional mix of inventory zoning approaches across a variety of decentralised distribution centres. Process and fulfilment automation require novel warehousing systems and penetrative organisational adaptations (Kembro & Norrman, 2021), whereas the management of large fulfilment networks is subject to mature communication and coordination (Pereira & Frazzon, 2021). Hence, this archetype is represented by pure-online-channel-grocers rather than multi-/omni-channel-grocers with established warehousing structures and a diversity of functional areas. Following the collaboration focus, inventories are reserved at specific, pre-scheduled times and picking processes are automated or semi-automated. Collaboration specialists such as Amazon Fresh and NetGrocer from the USA employ process automation in combination with an extensive delivery network to serve an international customer base with short lead times and moderate time windows.

Moreover, order thresholds and out-of-stock handling routines based on extended lead-times help to decrease handling costs. Because these kinds of operations and networks place a high demand on physical assets, collaboration specialists cooperate with third-party organisations in terms of inventory, fulfilment point and delivery management. Contingent on the geographical area, the availability of resources within that area and the prevailing customer density, the inventory, fulfilment point and delivery governance are either handled by the e-grocer (high customer density, availability of company resources) or an external partner (low customer density, absence of company resources). Delivery fees are charged dynamically or based on hybrid methods (e.g., reduced fees for urban subscribers). In order to further improve the flow of goods across the entire network and ensure short transit times, collaboration specialists employ stationary hubs as delivery intermediaries and transship orders with HDTs, before attended home-deliveries are conducted with LCVs or e-scooters. Cooling and packaging requirements are handled by insulated, reusable boxes or cooled vehicles and temporary transportation cases, which facilitate cross-docking and transshipment processes. Coinciding with their forward-oriented, large-scale distribution network, collaboration specialists do not offer product returns but provide monetary refunds, discounts or physical correction instead.

4.5. Archetype #5 – Store advocates

Omni-channel-grocers representing store advocates employ supermarkets as fulfilment and order management points and provide attended home-deliveries. E-grocers are responsible for inventory, fulfilment point and delivery governance. To leverage synergy effects resulting from integrated inventory handling (Agatz et al., 2008), inventories are shared between offline- and online-channels. Nevertheless, as picking and storing routines are administered manually in a given outlet with shelf-locations, inventories have to be reserved upon order occurrence. Order picking for offline operations is handled by customers themselves, for which reason the online-order picking management is separated from offline picking. Due to the large product assortment available in stores, out-of-stock situations are settled by substituting the respective product(s) with equivalent alternatives.

Compared to EGF archetypes that built on distribution centres, store fulfilment features a rather low degree of efficiency (J. E. Scott & Scott, 2008); this is compensated by high order thresholds and instance-based, threshold-based or hybrid delivery fee incentives. Depending on the scope of the existing supermarket network, store advocates offer local, regional and national deliveries to urban/sub-urban areas within 12 to 36 hours. In contrast to parcel champions, a unique selling point of store advocates such as Intermarche (France) is the provision of fresh produce and perishable items. To ensure excellent product quality, orders are delivered with cooled LCVs and routed dynamically to ensure fast deliveries (Agatz et al., 2008). Customers benefit from estimation-based tracking and, in high order-density regions, various value-added services. In the case of complaints or failures, they can return the respective item(s) through designated store outlets or instantly upon delivery. Failure recovery is handled by monetary refunds, discounts or hybrid strategies (customers keep faulty items and receive a discount).

4.6. Archetype #6 – Automation enthusiasts

Automation enthusiasts describe multi- and omni-channel-grocers and pure-online-players that rely on automated inventory handling and order preparation systems to ensure efficient processing routines. They operate pick-run optimised warehouses that are equipped with semi-automated or fully-automated picking technologies, thus enabling fast handling routines. In accordance with pick-run optimised inventory zones and a high degree of process automation, order picking is done by case and inventories are reserved directly upon order reception. Regarding out-of-stock situations, automation enthusiasts employ systems that automatically increase the lead-time for a given item when an out-of-stock situation is discovered or algorithmically anticipated. Given that automation enthusiasts scale capacity based on expected demand, utilise a high degree of automation, and establish region-based order thresholds, they can economically serve all delivery areas on a national scale (J. E. Scott & Scott, 2008). Contrary to collaboration specialists, automation enthusiasts such as Tesco (UK) do not manage external partner networks, they govern fulfilment points and deliveries themselves, and utilise centralised (highly automated) rather than decentralised distribution centres.

In order to secure fast delivery operations with lead-times under 12 hours, they operate hub-and-spoke networks. HDTs or LCVs with extensive transportation capacities are used to ship orders from the distribution centre to the stationary hub where they are transshipped to LCVs with multiple cooling mechanisms and delivered through attended reception. In line with the given delivery routines, consolidation objectives and the operational requirements of the hub-and-spoke networks, automation enthusiasts implement moderate time windows and dynamic routing procedures to minimise traffic delays. To compensate for the accruing handling costs, they charge delivery fees based on subscriptions, thresholds and lead-times, and countervail failures with monetary refunds or discounts.

5. Contingency model for e-grocery value creation

Figure 4 presents our contingency model, which is based on the archetypes and the associated capacities. The clustering points out that an EGF archetype is characterised by a magnitude of different, partially antithetical attributes. To improve conceptual understanding, we grouped attributes based on their operational scope. The combination of the fulfilment attributes for order and delivery management in each archetype forms the basis of a contingency model that is related to the focus (service, costs, flexibility, efficiency) and assets (physical, financial) of an e-grocer.¹ In turn, archetypes and capacities are collectively linked in a trade-off relationship with operational efficiency and customer value, finally explaining e-grocery value creation. Next, we highlight differences in the manner in which the e-grocers manage fulfilment and explore value relationships based on the derived contingencies.

Figure 4. Contingency model for e-grocery value creation based on fulfilment.

5.1. Cost emphasis by pick-up service proponents

The managerial focus of pick-up service proponents is reducing costs while maintaining a sound level of efficiency and flexibility. For this purpose, they leverage existing store infrastructure and offer order collection rather than home-deliveries to avoid additional logistics costs at the expense of added customer value. Similarly, demand forecasting and internal inventory governance help to choose appropriate order quantities, thus decreasing the relative inventory costs. However, this approach also jeopardises the availability of products, which negatively impacts service level and efficiency. Because they have established several routines fostering operational productivity (e.g., substituting products that are out-of-stock, shared inventory handling), these e-grocers still seem to focus on efficiency and flexibility to a moderate degree. In contrast, service-related attributes such as delivery services or short time windows are generally not implemented, revealing a lack of service focus. Finally, the scope of the delivery area, reception type, and delivery scope require a large network of individual store units, entailing a moderate demand on the physical assets (as fulfilment points are often, at least partially, governed by third parties) and a high demand on the financial assets of the e-grocer. As a consequence, the overall customer value provided by pick-up service proponents is rather low, which is compensated by a moderate degree of operational efficiency, enabling low operating costs.

5.2. Efficiency quest by parcel champions

Parcel champions appear to emphasise a particularly large delivery area. Delivery outsourcing and order management prioritisation allow efficient operations with moderate physical and financial assets. Because these e-grocers utilise integrated, decentralised fulfilment centres with dedicated, partially automated picking procedures and pick-run optimised inventory zones, they can flexibly react to changing demand patterns while maintaining a high level of productivity. Service appears to be a less important concern for parcel champions, which is evidenced by their lead-times, delivery time-windows, and delivery services. Hence, the overall outcome of this archetype is rather in favour of operational efficiency than customer value.

5.3. Purposeful waiver of control by click-and-pick experts

The physical and financial assets of click- and-pick experts appear to be limited, which these e-grocers compensate for by collaborating with local grocers to offer delivery services from facilities that are governed externally. On the one hand, the delivery services and governance model of this approach indicate a moderate degree of service, cost, and flexibility focus. On the other hand, it neither enables the alignment of order management routines with e-grocery operations, nor allows for the planning of inventories or the optimisation of pick-runs. Accordingly, click-and-pick experts appear to exchange operational efficiency for a minimum demand on assets, still allowing these e-grocers to provide a decent level of customer value based on their order and delivery management attributes.

5.4. Pursuit of service quality and efficiency by collaboration specialists

As indicated by their value-adding delivery services (e.g., moderate time-windows, short lead times, attended deliveries, the provision of replacements or refunds without the need for customers to return a faulty item) and productive order management routines (e.g., hybrid inventory planning, region-based inventory zoning, decentralised distribution, hybrid governance), collaboration specialists set the emphasis on high service levels, while at the same time trying to maximise efficiency and minimise costs. To achieve high degrees of operational efficiency and customer value, they cooperate with numerous external partners, ultimately decreasing the demands on physical assets at the cost of increased management and coordination requirements as well as reduced control. As their operations rely on partnerships to a great extent, their focus on flexibility appears to be rather low. In conclusion, the fulfilment characteristics of collaboration specialists necessitate high financial assets (e.g., due to human resource costs required for coping with the operational and managerial efforts) and moderate physical assets (e.g., in terms of fulfilment infrastructure). However, they also allow these e-grocers to simultaneously pursue multiple operational goals and excel on both, operational efficiency and customer value.

5.5. Exploitation of synergies by store advocates

Store advocates leverage a sophisticated store network to cover the targeted delivery area. They opt to capitalise on synergies from integrated order management procedures and emphasise multi- rather than omni-channelling. In contrast to pick-up service proponents, they govern inventories and fulfilment points internally and offer attended home-deliveries to provide additional customer value. Thus, despite a rather local delivery area, distinctive physical and financial assets seem to be necessary to operate this model and cope with the operational requirements. It appears that they have a balanced focus on service, cost, flexibility and efficiency, though in all cases this is moderate as store advocates operate the online business as a complement to the offline business unlike online-only providers such as collaboration specialists. In line with their balanced focus on competitive value creation factors, store advocates achieve a moderate degree of operational efficiency and provide a medium level of customer value.

5.6. Value creation through technological excellence by automation enthusiasts

Finally, automation enthusiasts focus on service, costs, efficiency, and flexibility to an equally large extent. As omni-channel-grocers, they seek to create an integrated shopping experience for their customers across all channels. Hence, they also attach a lot of importance to their online-channel and opt to excel on all relevant competitive priorities. They utilise automated, technically-assisted order and delivery management routines, featuring a high degree of control as well as extensive demands on physical and financial assets. Their fulfilment model, consisting of sophisticated order (e.g., automated picking) and value-adding delivery management attributes (e.g., short lead-times), suggests a high demand on financial assets (e.g., to invest in and maintain automation technology) and physical assets (e.g., storage robots, delivery fleet). Ultimately, their capacities in terms of cost minimisation, maximisation of efficiency and flexibility, as well as provision of additional service value allow automation enthusiasts to deliver salient value outcomes in terms of operational efficiency and customer value.

6. Discussion

Fulfilment is understood as a major factor for value creation in e-commerce. Whilst prior research has tended to examine individual fulfilment attributes in the generic realms of e-commerce, we presented a holistic contingency model that captures the relationships between diverse fulfilment attributes and value outcomes in the e-grocery domain. Next, we discuss the main implications for theory and practice.

6.1. Theoretical implications

Structuring the solution space of EGF decisions attributes. The theoretically-grounded and empirically-refined taxonomy organises relevant attributes for e-grocery. As “theory for analysing” (Gregor, 2006), our findings add descriptive knowledge to several IS research streams, including multi- and omni-channel commerce (Chen et al., 2018; Trenz et al., 2020) and digital technologies for order fulfilment (Cotteleer & Bendoly, 2006; Hardgrave et al., 2008). Our attributes share similarities with, for instance, work on logistics (e.g., delivery times, Lim et al., 2018), supply chains (e.g., supplier coordination, Iyer et al., 2009), and IT (e.g., technologies used, Weill & Olson, 1989). Contrarily, our taxonomy has unique attributes, such as “order cooling”, and more contextualised ones, including “picking technology” as an instance of technology and “delivery lead times” for order handling.

By focusing on the grocery context, our results advance and contextualise the rather isolated landscape of research on generic e-commerce. Context appreciation is important for reflection on digital options (Sandberg et al., 2014) as well as for contingency theory, which argues for considering the specific surroundings (Crișan & Stanca, 2021; Miller, 1981). Moreover, the compiled attributes and archetypes help to establish a shared, cross-disciplinary understanding of EGF, facilitating knowledge sharing and accumulation.

Assembling EGF archetypes. Building upon the taxonomy, we discovered common combinations of EFG attributes and conceptually enriched these combinations with associated capacities. Existing fulfilment classifications include general distribution models (e.g., Agatz et al., 2008) and typologies for a particular function, such as warehousing (Kembro & Norrman, 2021) or last-mile distribution (Hübner et al., 2016). By deriving archetypes that capture attributes across the entire fulfilment cycle, we complement these streams and help to detect interdependencies that can be subject to future investigation. For instance, when designing and implementing digital technologies to support fulfilment, the underlying attributes, capacity constructs, and value outcomes need to be aligned with the situation at hand. Our archetypes conceptualise operational characteristics and activities to identify trade-offs and strategize digital transformation transactions (e.g., Matt et al., 2015) in the grocery sector. As an example, store advocates will require technical order management solutions, such as RFID or ERP systems, to integrate information flows along the e-grocer’s sales touchpoints (Wamba & Chatfield, 2009).

Handling the interplay of attributes in a specific context. Following the underpinning assumptions from contingency theory, the business’s goals need to the aligned with the environment to make appropriate decisions (Crișan & Stanca, 2021). By disclosing relationships between EGF attributes and value creation based on EGF archetypes and capacities, our study advances IS research on e-commerce value relationships and adoption (Fang et al., 2021; Pavlou & Fygenson, 2006; Polites et al., 2018) and aids future research in adopting a more comprehensive view of value creation in e-grocery.

For illustration, we have shown that automation enthusiasts such as Tesco feature a high degree of operational efficiency and customer value, which coincides with the paradigm that order fulfilment performance is often driven by technology (Kembro & Norrman, 2019). Automation enthusiasts possess a logistics automation capacity profile, representing the “ability of interfirm logistics relationships to automate package labelling, schedule pickup/delivery times for shipments, and specify special processing requirements” (Rai et al., 2012, p. 240). Collaboration specialists achieve customer value and operational efficiency by leveraging an extensive partner network but need to be aware of the fact that today’s partners can become tomorrow’s competitors (Etzion & Pang, 2014). As opposed to other archetypes that also use collaboration (e.g., click-and-pick experts), the fulfilment operations of collaboration specialists feature unique exigencies, which can be evaluated via the proposed contingency model. Hence, our results complement existing findings on the interplay of attributes (e.g., Eriksson et al., 2019; Kembro & Norrman, 2021) by presenting archetypes across the entire fulfilment cycle as well as emphasising capacities, potential implications and trade-offs. Thereby, we help generate theory that informs how to support EGF approaches more efficiently.

6.2. Managerial contributions

Create value through EGF combinations. Our study has relevant practical implications for the design and implementation of EGF, especially in the digital era. The archetypes assist in selecting appropriate means of interaction, for instance in terms of partner networks or digital technologies, or deciding on an appropriate degree of automation and autonomy. By uncovering EGF archetypes, managers are enabled to holistically understand how fulfilment-based value creation can be affected by shortcomings in organisational capabilities and a lack of planning. As e-grocers seek to identify suitable fulfilment options, the interplay of attributes and their trade-offs present a major challenge.

Contextualise EGF choices. Our paper provides an outline of available options and helps e-grocers to align online-channels with their contexts. The contingency model can be used to assess the trade-offs between alternative operational setups and to make purposive design choices for EGF in their specific environment. Hence, our findings respond to practical challenges, including the demand for environmentally-friendly deliveries (Milioti et al., 2021).

6.3. Limitations and outlook

Like all research, our study has limitations that open avenues for future research. First, although several similarities between the e-grocery context and contingency theory can be made, the selection is based on own decisions. Despite matching our objective, one might want to rely on other theories, such as configuration theories that argue for multiple conditions and paths to achieve a certain state of performance (McLaren et al., 2011) or strategic choice theory to explore the role of leading groups on decisions (Turedi & Zhu, 2019). Nonetheless, contingency theory offers a valuable view on dynamic situations and is considered as a foundation that can be complemented through additional perspectives, such as configuration theory (Van de Ven et al., 2013). Second, contingency theory does not provide causality (Weill & Olson, 1989), which is why future research can prioritise the impact of certain options. Besides, the archetypes are based on the assumption that only one fulfilment structure may be adopted for a given scenario; in some markets, it is common to concurrently operate multiple fulfilment structures, which requires investigation of the interrelationships between the combinations. Third, our taxonomy represents the status quo of attributes in practice and literature. In terms of knowledge evolution, our results serve as a starting point for further investigation. Fourth, our sample of e-grocers was limited to 111 instances, thus it cannot be claimed to be exhaustive. To overcome this, we used techniques for triangulation that increase the result’s robustness. Nevertheless, future studies with larger sample sizes, country-specific samples, or e-grocers from emerging economies will be needed for additional validation. Also, although our sample only includes extant providers, this industry is highly dynamic, especially since COVID-19. Fifth, our paper was a cross-sectional rather than a longitudinal study, which uncovered archetypes of e-grocery, but did not investigate how and why enterprises switch from one archetype to another.

7. Conclusion

Fast-changing environments and quickly evolving technology-driven phenomena require organisations to consciously adapt decisions to survive in today’s dynamic markets. Although there is no single best option for aligning an e-grocery archetype with its environment, understanding the set of choices and their relations is a prerequisite to making (better) decisions. To foster this, our paper investigates relationships between EFG attributes and value outcomes. Thereby, we disclose promising indications assisting e-grocers to leverage the full potential of the emerging stream of digital transformation in grocery and establish a sustainable source of customer value. This will help to increase the overall popularity of e-grocery in the medium-term and provide worthwhile advantages for society and environment. Moreover, we hope that our study motivates other scholars to adopt and test the proposed archetypes and contingency model to create more holistic, generalisable knowledge on the increasingly important topics of (digital) fulfilment in general and EGF in particular.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. As we derived focus and asset capacities conceptually and did not directly observe or measure their development extent, these items are outlined in a dotted box (Figure 2).

Appendix Appendix A:

Related classifications (literature review)

In the following section, we provide an overview of related classifications on fulfilment attributes for e-commerce in general as well as e-grocery in particular. To ensure a systematic identification with a high degree of comprehensibility, we employed a systematic literature review, conducted in accordance with the guidelines of Webster and Watson (2002). The review was initially conducted in July 2020 and updated in October 2020, using the following search string without any restrictions from the search engine: (“e-grocery” OR “e-commerce” OR “online food”) AND “fulfilment” AND (“classification” OR “typology” OR “taxonomy”). The queries were performed in the ACM, EBSCO, Emerald, IEEE, AIS, Taylor & Francis, ScienceDirect and JSTOR databases, focusing on the search domain title, abstract and keywords.

As a result, Table A1 shows that a paucity of studies has attempted to classify and conceptualize fulfilment attributes or combinations as influencing factors for value creation. However, considering the heterogeneity of order fulfilment and related service features (von Viebahn et al., 2020), there is an immanent need to holistically integrate past findings into a model that captures the relationships between fulfilment combinations and value outcomes.

Appendix B:

Research procedure

In Table B1 we outline the research steps, methods, and outcomes across the three key clusters of activities performed in this study.

Table A1. Related classifications of fulfilment attributes.

Reference	Category	Main Topic	Warehouse management	Fulfilment points	Distribution management	Reception models	Reverse logistics
Bock & Wiener, 2017	EC	Business models	◎	◎	◎	◎	◎
Boysen et al., 2019	EC	Warehouse models	●	◎	◒	◎	◎
Cullen & Webster, 2007	EC	Business models	◎	◎	◎	◎	◎
Davies et al., 2019	EG	E-grocery models	◎	◒	◒	◒	◎
Dubosson-Torbay et al., 2002	EC	Business models	◎	◒	◒	◎	◎
Durand & Gonzalez-Feliu, 2012	EG	Fulfilment models	◒	●	●	◒	◎
Emeç et al., 2016	EG	Vehicle routing	◎	◒	●	◎	◎
Eriksson et al., 2019	EG	Fulfilment models	●	◒	◒	◎	◎
Gevaers et al., 2011	EC	Last-mile fulfilment	◎	◒	●	●	◎
Holland & Gutiérrez-Leefmans, 2018	EC	Business models	◎	◎	◎	◎	◎
Hübner et al., 2015	EC	Warehouse models	●	◒	◎	◎	◎
Hübner et al., 2016	EG	Last-mile fulfilment	●	◒	●	●	●
Jain et al., 2017	EC	Fulfilment models	◒	◎	◒	◎	●
Johnson & Iyamu, 2019	EG	Business Models	◎	◎	◒	◎	◎
Kämäräinen et al., 2001	EG	Fulfilment models	◎	◒	◒	●	◎
Kaplan & Sawhney, 2000	EC	Business models	◎	◎	◒	◎	◎
Kureshi & Thomas, 2019	EG	Business models	◎	◒	◒	◒	◒
Lee & Park, 2009	EC	Business models	◎	◎	◎	◎	◎
Lim et al., 2009	EG	Business models	◎	◎	◒	◎	◎
Lim et al., 2018	EG	Last-mile fulfilment	◎	●	●	◒	◒
Lim & Srai, 2018	EC	Last-mile fulfilment	◎	◒	●	◎	◒
Liu et al., 2016	EC	Business models	◎	◎	◎	◎	◎
Mahadevan, 2000	EC	Business models	◎	◎	◎	◎	◎
Marchet et al., 2018	EG	Fulfilment models	◒	●	●	◒	●
Martín et al., 2019	EG	Research topics	◒	◒	◒	◎	◒
Melacini et al., 2018	EG	Fulfilment models	◒	◒	◒	◒	◒
Mkansi et al., 2018	EG	Fulfilment models	◒	◒	◒	◒	◒◒
Palmer et al., 2000	EG	Business models	◎	◒	◒	◒	◎
Saleem et al., 2018	EG	Customer behaviour	◎	◎	◒	◎	◎
Seitz et al., 2017	EG	Customer behaviour	◎	◎	◒	◎	◎
Spiller & Lohse, 1997	EC	Business models	◎	◎	◎	◎	◒
Urbinati et al., 2017	EC	Business models	◎	◎	◎	◎	◒
Vakulenko et al., 2018	EC	Fulfilment models	◎	◎	◒	◒	◒
Vakulenko et al., 2019	EC	Last-mile fulfilment	◎	◎	●	◒	◒
Weking et al., 2018	EC	Business models	◎	◎	◒	◎	◒
Wen et al., 2001	EC	Website design	◎	◎	◎	◎	◎
Winkenbach & Janjevic, 2018	EC	Last-mile fulfilment	◒	●	●	●	◎
Wollenburg et al., 2018	EG	Fulfilment models	◒	●	◒	●	◎
Yousept & Li., 2005	EG	Business models	◒	◒	◒	◒	◒

Caption: ● = full consideration; ◒ = partial consideration; ◎ = no considerationEC = E-commerce; EG = E-grocery

Table B1. Research procedure.

Steps and methods	Output
Identifying e-grocery attributes
Literature review To ensure a systematic identification with a high degree of comprehensibility, we employed a systematic literature review (SLR), conducted in accordance with the guidelines of Webster and Watson (2002). The review was conducted between July 2020 and January 2021, using the following search string: (“e-grocery” OR “home-delivery” OR “online food”) AND “fulfilment” AND (“strategy” OR “scenario” OR “option” OR “model”). The queries were performed in the ACM, EBSCO, Emerald, IEEE, AIS, Taylor & Francis, ScienceDirect and JSTOR databases, focusing on the search domain title, abstract and keywords. In line with our main objective to identify sources supplying potential dimensions or attributes on different e-grocery concepts, we only included papers that are written in English and explicitly focus on grocery deliveries as well as fulfilment. Hence, articles related to general e-commerce or internet business were excluded from the results. Moreover, we analysed citations from the identified literature to determine less recent publications (backward search) and assessed citation indices to find articles quoting the key publications identified in earlier stages of the review (forward search). In total, we considered 80 scientific sources (4 books, 15 book chapters, 21 conference proceedings and 40 journal publications).	Relevant literature, input into taxonomy design
Web-content analysis Between January and March 2020, we assessed the company websites of 111 e-grocery providers worldwide (see Table 5) as well as several industry reports (e.g., supermarketnews.com) to derive additional dimensions and attributes of EGF or update priory identified approaches from (outdated) literature results.	Real-life/empirical attributes, input into taxonomy design
Taxonomy development Following the taxonomy building approach (Nickerson et al., 2013), we relied on two sources to derive our overview of e-grocery attributes, (1) a systematic literature review (Webster and Watson 2002) and the empirical example of 111 e-grocer instances (see overview in Appendix D). Meta-characteristics: Our target users are practitioners who are interested in developing or adapting an EGF strategy and researchers opting to analyse conceptual e-grocery specifics. We carried out two conceptual-to-empirical (C2E) iterations and six empirical-to-conceptual iterations (E2C). We decided to incorporate empirical objects (i.e., real e-grocery providers) in particular as this is in line with related studies that also investigate strategic decisions and trade-offs (e.g., Kathuria, 2000; Kathuria et al., 2018). In iteration 1 (C2E), we integrated relevant attributes from the research background, relating to general e-commerce fulfilment schemes. In iterations 2–7, we coded operational EGF instances extracted from literature by employing a SLR (Webster & Watson, 2002). The results were coded based on the occurrence of an e-grocery provider instance (i.e., Tesco). Articles featuring at least one provider instance were analysed during iterations 2–6 (E2C), whereas articles focusing on the general, high-level examination of EGF where screened in iteration 7 (C2E). In iteration 8 (E2C), we conducted a qualitative WCA, whereby we assessed the websites of 111 e-grocers worldwide between January and March 2021 to derive fulfilment attributes (see Appendix D). As a result of the iterations, all ending conditions were met and the taxonomy building ended. We adopted the objective and subjective ending conditions of Nickerson et al. (2013), with one exception: To allow retaining attributes that are not yet employed by e-grocers (but have been suggested by research to be viable in the future), we did not apply the condition where “at least one object is classified under every characteristic” (Nickerson et al., 2013, p. 344), as suggested by Szopinski et al. (2019).	Taxonomy of attributes for e-grocery fulfilment, i.e., foundation for archetypes
Taxonomy evaluation In attempting to indicate the taxonomy’s usefulness, we followed Szopinski et al. (2019) and performed evaluation episodes with different subjects, objects and methods. Thereby, we sought to evaluate the understandability (i.e., do users understand the attributes of e-grocery fulfilment?) as well the object-dependent (i.e., can users correctly classify fulfilment strategies?) and object-independent (i.e., can users correctly apply the taxonomy to develop new fulfilment strategies?) applicability of the taxonomy. Concerning the object-dependent applicability, we performed concurrent think-alouds in which three practitioners were asked to classify their e-grocery models based on the taxonomy and two expert interviews with academics to ensure a methodical rigour. For the object-independent applicability, two virtual workshops were held, in which a group of two participants used the taxonomy to solve a fictive task The evaluation process was initiated by employing logical arguments, which refer to “argument[s] with face validity (Peffers et al. 2012, p. 402). Accordingly, the first author, possessing comprehensive theoretical and practical knowledge on the e-grocery domain, as well as the second author of this paper, featuring a sophisticated theoretical background in taxonomy development and evaluation, mapped and compared the taxonomy to (1) e-grocery provider objects that have not been employed for taxonomy building, (2) other taxonomies in the e-commerce domain, and (3) conceptual studies in the field of e-grocery fulfilment. In episodes 1–5, concurrent think-aloud studies and expert interviews were used to evaluate the object-related applicability. We considered both domain and method experts; selected based on their experience. During the think-aloud studies, practitioners from e-grocery retailers were asked to classify their own fulfilment models while immediately reflecting on their thoughts, as this method is viable in “detecting as many usability problems as possible” (Alshammari & Mayhew 2015, p. 55). The successional expert interviews were designed to additionally capture feedback on the structure and methodological appropriateness of the taxonomy. The studies were conducted via Microsoft Teams and recorded with the permissions of the interviewees. The sessions lasted between 51 and 58 minutes (think-aloud) and 32 and 55 minutes (expert interviews) and were transcribed. For the analysis, we created transcripts of each of the sessions with f4transkript (Version 8.1.1) and performed a selective coding, in which we extracted indications for the artefacts’ usefulness and applicability as well as critical feedback that was employed for improvement In episode 6–7, we conducted two workshops using Microsoft Teams to assess the object-independent applicability of the taxonomy. During the workshops with two participants each, practitioners from an organisation specialised in e-commerce fulfilment were asked to use the taxonomy to solve a fictive task. In the course of the 90-minute workshop, they had to develop a fulfilment strategy for an e-grocer planning to establish an online sales channel.	Empirically evaluated taxonomy of attributes for e-grocery fulfilment
Developing EGF archetypes
Coding/classification We coded each of the final taxonomy’s attributes that was observable for one of the 111 e-grocers with a 1 and 0 otherwise. Hierarchical clustering Following Bensaou and Venktraman (1995), we (a) standardised the aggregate scores for each construct, (b) used squared Euclidean distances as similarity measure and (c) employed Ward’s minimum variance method (Ward, 1963) for an initial cluster formation. To determine an appropriate number of clusters, we inspected the resulting dendrogram and the matching coefficient from the hierarchical clustering, indicating solutions with three and six clusters to be most useful. Although the matching coefficient is less complex than other distance measures, research has proven it to be similarly effective, both, in analogue (e.g., Hands & Everitt, 1987) and congeneric research contexts (e.g., Finch, 2005).	Number of clusters for archetype building
K-means ++ As the decision concerning the number of clusters always entails a conflict between the solution’s manageability and the cluster’s homogeneity (Milligan & Cooper, 1985), we tested the robustness of the three- and six-cluster solution by using K-means++ to additionally calculate two-, four-, five-, seven- and eight-cluster solutions and comparing the individual means of each solution. Although no solution featured a significant difference for all constructs across all clusters, the three- and six-cluster solutions showed a high degree of relative discriminatory power with multiple significant differences across clusters (see Appendix E.2).	Individual cluster formation and composition from real-life e-grocers
Evaluating EGF archetypes
To ensure the external validity of the clusters, we conducted several evaluations with representatives of major omni-channel-grocers, e-grocers and fulfilment specialists, who indicated that our findings are congruent with their own operating models and likely to reflect common industry standards. Cluster evaluations were performed with practitioners from the taxonomy evaluation phase by means of expert interviews (see subjects below).	Empirically evaluated archetypes of e-grocery fulfilment
Enrich EGF archetypes with capacities
To create enriched archetypes, we employed abductive reasoning to derive focus and asset capacities based on the respective fulfilment combination from the literature and the empirical analysis of corresponding e-grocers.	Capacity constructs: Focus and assets
Deriving a contingency model
To investigate value relationships, outcome constructs (operational efficiency, customer value) were identified by studying three selected e-grocers of each cluster through documentation and telephone surveys. These e-grocers were selected based on their fulfilment manager’s willingness to participate in the survey and disclose the desired information. The outcome constructs (i.e., operational efficiency, customer value) were measured on a 5-point Likert scale (1= Unsatisfactory, 5 = Excellent) and measured by the following item scales: Operational efficiency - How would you rate your fulfilment model in terms of: Ability to support cost-efficient handling and supply processes (OE1) Ability to meet productivity standards (OE2) Ability to meet stock turnover rate requirements (OE3) Ability to meet picking accuracy requirements (OE4) Customer value – How would you rate your fulfilment model in terms of: Ability to achieve a high degree of delivery velocity (CV1) Ability to maintain low transaction costs for your customers (CV2) Ability to meet convenience requirements of your customers (CV3) Ability to ensure a high product quality upon reception (CV4) The results of the surveys inform the outcome constructs of the contingency model and show low operational efficiency outcomes for click-and-pick experts (Ø 1.83), medium outcomes for pick-up service proponents (Ø 3.48) and store advocates (Ø 2.96), as well as high outcomes for parcel champions (Ø 3.87), collaboration specialists (Ø 4.54), and automation enthusiasts (Ø 4.46). In terms of customer value, the empirical results indicate low outcomes for pick-up service proponents (Ø 1.99) and parcel champions (Ø 2.25), medium outcomes for click-and-pick experts (Ø 3.38) and store advocates (Ø 3.21), and high outcomes for collaboration specialists (Ø 4.85) and automation enthusiasts (Ø 4.16).	Contingency model; value relationships
Evaluating subjects (for taxonomy evaluation, archetype evaluation, and value outcomes)
Practitioners P1: Global Director (Position), Fulfilment Specialist (Company), 37 (Age), Male (Gender), 10 years (EGF experience), Think-aloud (Methodology), 57:48 (Duration), Taxonomy, Archetypes (Objects) P2: Global Fulfilment Manager (Position), Pure-online-channel grocer (Company), 35 (Age), Male (Gender), 4 years (EGF experience), Think-aloud, Survey (Methodology), 51:05 (Duration), Taxonomy, Archetypes, Value Outcomes (#4) (Objects) P3: Senior Manager Operations Development (Position), Omni-channel-grocer (Company), 43 (Age), Male (Gender), 5 years (E-grocery experience), Think-aloud, Survey (Methodology), 54:27 (Duration), Taxonomy, Archetypes, Value Outcomes (#6) (Objects) P4: Vice President (Position), Fulfilment Specialist (Company), 54 (Age), Male (Gender), 20 years (EGF experience), Workshop (Methodology), 44:15 (Duration), Taxonomy, Archetypes (Objects) P5: Project Manager (Position), Fulfilment Specialist (Company), 30 (Age), Male (Gender), 5 years (EGF experience), Workshop (Methodology), 44:15 (Duration), Taxonomy, Archetypes (Objects) P6: Senior Solution Design Manager (Position), Fulfilment Specialist (Company), 37 (Age), Female (Gender), 7 years (EGF experience), Workshop (Methodology), 01:36:52 (Duration), Taxonomy, Archetypes (Objects) P7: Senior Solution Design Manager (Position), Fulfilment Specialist (Company), 51 (Age), Male (Gender), 10 years (EGF experience), Workshop (Methodology), 01:36:52 (Duration), Taxonomy, Archetypes (Objects) P8: Regional Fulfilment Manager (Position), Omni-channel-grocer (Company), 51 (Age), Female (Gender), 12 years (EGF experience), Survey (Methodology), Value Outcomes (#1) (Objects) P9: Global Fulfilment Manager (Position), Pure-online-channel grocer (Company), 35 (Age), Male (Gender), 9 years (EGF experience), Survey (Methodology), Value Outcomes (#4) (Objects) P10: Fulfilment Manager (Position), Multi-channel-grocer (Company), 29 (Age), Female (Gender), 6 years (EGF experience), Survey (Methodology), Value Outcomes (#2) (Objects) P11: Global Fulfilment Manager (Position), Omni-channel-grocer (Company), 44 (Age), Female (Gender), 10 years (EGF experience), Survey (Methodology), Value Outcomes (#6) (Objects) P12: Local Fulfilment Manager (Position), Omni-channel grocer (Company), 36 (Age), Female (Gender), 7 years (EGF experience), Survey (Methodology), Value Outcomes (#5) (Objects) P13: Assistant Fulfilment Manager (Position), Multi-channel-grocer (Company), 28 (Age), Male (Gender), 5 years (EGF experience), Survey (Methodology), Value Outcomes (#1) (Objects) P14: Area Manager (Position), Pure-online-channel grocer (Company), 42 (Age), Female (Gender), 15 years (EGF experience), Survey (Methodology), Value Outcomes (#3) (Objects) P15: Area Fulfilment Manager (Position), Omni-channel-grocer (Company), 41 (Age), Male (Gender), 15 years (EGF experience), Survey (Methodology), Value Outcomes (#5) (Objects) P16: Global Fulfilment Manager (Position), Multi-channel-grocer (Company), 49 (Age), Male (Gender), 11 years (EGF experience), Survey (Methodology), Value Outcomes (#2) (Objects) P17: Fulfilment Manager (Position), Pure-online-channel grocer (Company), 31 (Age), Female (Gender), 6 years (EGF experience), Survey (Methodology), Value Outcomes (#4) (Objects) P18: Head of Fulfilment (Position), Pure-online-channel grocer (Company), 33 (Age), Female (Gender), 5 years (EGF experience), Survey (Methodology), Value Outcomes (#3) (Objects) P19: Fulfilment Manager (Position), Pure-online-channel grocer (Company), 34 (Age), Female (Gender), 3 years (EGF experience), Survey (Methodology), Value Outcomes (#2) (Objects) P20: Fulfilment Manager (Position), Pure-online-channel grocer (Company), 36 (Age), Male (Gender), 8 years (EGF experience), Survey (Methodology), Value Outcomes (#3) (Objects) P21: Regional Fulfilment Manager (Position), Multi-channel-grocer (Company), 45 (Age), Male (Gender), 4 years (EGF experience), Survey (Methodology), Value Outcomes (#5) (Objects) P22: Head of Fulfilment (Position), Omni-channel-grocer (Company), 44 (Age), Female (Gender), 18 years (EGF experience), Survey (Methodology), Value Outcomes (#1) (Objects) P23: Local Fulfilment Manager (Position), Multi-channel-grocer (Company), 42 (Age), Male (Gender), 13 years (EGF experience), Survey (Methodology), Value Outcomes (#6) (Objects) Researcher R1: PhD, PostDoc (Position), 31 (Age), Male (Gender), 9 (Relevant publications), Interview (Methodology), 55:13 (Duration), Taxonomy (Objects) R2: PhD Student (Position), 29 (Age), Female (Gender), 12 (Relevant publications), Interview (Methodology), 32:11 (Duration), Taxonomy (Objects)

Table C1. E-grocery fulfilment attributes (taxonomy dimensions and attributes).

Dimension	Attribute description		Example ref.
Meta-dimension: Inventory handling
Inventory planning	Demand forecast	Mathematical forecasting methods to estimate future customer demands and plan inventories.	Agatz et al. (2008)
	Safety stocks	Accumulating large stock quantities to cope with all customer requests for a specific time frame.
	Inventory rationing	A strategy for heterogeneous markets that reserves some inventory for high margin customers.
	Abdication of planning	Acceptance of orders in chronological order without additional conditions or planning operations.
	Hybrid planning	Utilisation of different planning strategies across various product groups (e.g., safety stocks for fast-moving and forecasts for slow-moving items)
Inventory zoning	Store-like	Inventory display-case layouts that have been aligned with conventional supermarket layouts.	Not available (empirical evidence only)
	Pick-run optimised	Inventory zone layouts that have been optimised based on the individual demand of different products (e.g., fast-moving items are easily accessible, whereas slow-moving items are placed at unfavourable positions.)
	Region-based mix	A mix of store-like and pick-run optimised zoning structures for different regional fulfilment points.
Inventory governance	Contractor	Collaboration with external partners in terms of inventory ownership (e.g., dropshipping).	Agatz et al. (2008); Jindal et al. 2021
	E-grocer	Individual management and ownership of inventories sold via the e-grocers distribution channel.
	Hybrid governance	The combination of external and internal inventory ownership for different product groups.
Channel inventory management	Pure-online-channel picking	Organizations that exclusively distribute grocery inventories online (web-only).	Agatz et al. (2008); Marchet et al., 2018
	Offline- and online-channel picking	Organizations that distribute groceries via an online- and offline-channel (bricks-and-clicks); individual (for each channel) vs. shared (across all channels).
Reservation time	Specific time	Discrete time slots used to reserve inventories based on customer orders.	Kureshi & Thomas, 2019
	Generic time	Continuous inventory reservations over the day.
Meta-dimension: Order preparation
Point of fulfilment	Store	Utilisation of a supermarket for picking, packing and shipping orders with the designated delivery type. The store can be an outlet (normal supermarket), front centre (combination of supermarket and fulfilment centre).	Boyer and Hult, 2005; Dayarian et al., 2020; Haider et al., 2020
	Distribution Centre	Utilisation of a dedicated DC for picking, packing and shipping orders with the designated delivery type. The DC can be centralised (all orders of an entire region), decentralised and separated (orders of a particular, dedicated area, whereby customer orders are assigned to one particular DC), as well as decentralised and integrated (orders of a particular, dedicated area, whereby customer orders are split across DCs).	Agatz et al. (2008); Aktas et al., 2021; Auf der Landwehr et al., 2020; Winkenbach & Janjevic, 2018 Zhu et al., 2021
	Operational mix	Fulfilment of orders via a mix of store outlets, front centres as well as centralised and decentralised DC for the designated delivery type (e.g., based on product category).	Agatz et al. (2008)
Fulfilment point governance	Third-party	Utilisation of fulfilment points that are legally and operationally governed by external parties (click & pick).	Agatz et al. (2008); Anckar et al., 2002
	E-grocer	Individual management and ownership of fulfilment point infrastructure.	Janjevic & Winkenbach, 2020
	Hybrid structures	Combination of external and internal fulfilment point ownership for different channels or products.	Agatz et al. (2008)
Order Threshold	Absent	The relinquishment of minimum customer order values for e-grocery operations.	Anckar et al., 2002; Haider et al. 2020
	Present	The implementation of a specific minimum order value required to claim the e-grocery service.
Picking procedure	Split-case picking	Picking individual products from unique locations and placing them directly into shippable cartons.	Agatz et al. (2008); Vazquez-Noguerol et al., 2020
	Case-picking	Picking products in full case or carton quantities from a variety of storage mediums.
	Pallet-picking	Picking products in pallet quantities from a variety of storage mediums and assisted by machinery.
Picking automation	Person-to-goods	Product/order picking operations handled by human resources without technical equipment.	Hübner et al., 2015; Marchet et al., 2018
	Goods-to-person	Product/order picking operations handled by human resources and assisted by technology (e.g., pick-by-light).	Agatz et al. (2008); Marchet et al., 2018; Winkenbach & Janjevic, 2018
	Mainly auto-mated picking	Implementation of robotic techniques.
Channel picking management	Pure-online-channel picking	Organizations that exclusively align picking operations for online distribution channels (web-only).	Marchet et al., 2018
	Offline- and online-channel picking	Organizations that distribute groceries via an online- and offline-channel (brick-and-click); integrated (combine picking) vs. separated (separate for online/offline) vs. capacity optimised (based on order capacities).	Hübner et al., 2015; Marchet et al., 2018
Out-of-stock handling	Lead-time	The increase of the delivery lead-time by the e-grocer if an order contains an out-of-stock item.	Boyer & Hult, 2005; Tarn et al. 2003
	Substitute product	A strategy, whereby the e-grocer replaces an out-of-stock item with a comparable product	Anckar et al., 2002; Boyer & Hult, 2005
	Cancellation	A situation, where the e-grocer removes the out-of-stock item from the order list or cancels the order.
Meta-dimension: Delivery planning
Delivery area	Local	Fulfilment on a local scale, regional scale, national scale, or an international scale.	Anckar et al., 2002; de Koster, 2002; Marchet et al., 2018; Xing et al. 2010
	Regional
	National
	International
Delivery scope	Urban only	Fulfilment in urban areas, in urban and sub-urban areas, as well as in urban, sub-urban and rural areas.
	Sub-urban areas
	Sub-urban and rural areas
Delivery frame	Weekdays	Deliveries on weekdays only, weekdays and Saturdays, as well as on weekdays, Saturdays and holidays.	Anckar et al. 2002; de Koster, 2002
	Saturdays incl.
	Saturdays and holidays incl.
Delivery lead-time	30 minutes	A time frame of 30 minutes, two hours, same day, 24 hours, as well as 48 hours and more between the reception of a customer order and the final fulfilment.	Agatz et al. (2008); Anckar et al., 2002; Boyer & Hult, 2005; de Koster, 2002; Vinsensius et al., 2020
	Two hours
	Same-day
	Next-day
	Two/more days
Delivery time window	Narrow	30 minutes |60 minutes.
	Moderate	2 hours |3 hours |4 hours.
	Wide	More than 4 hours.
Delivery charges	Free of charge	The absence of additional fees for delivery services. The additional delivery costs of the e-grocer can be …	Agatz et al. (2008)
	Fixed	… a fixed pricing system for delivery charges; instance (individual order) vs. subscription.	Agatz et al. (2008); Aktas et al., 2021; Haider et al., 2020; Marchet et al., 2018; Vinsensius et al., 2020; Zhu et al., 2021
	Dynamic	… the implementation of a dynamic system for delivery charges. Pricing dynamics can base on lead-time, threshold, location, subscription.
	Hybrid charges	… the combination of several dynamic pricing systems such as lead-time and subscriptions.
Order packaging	Absent/temporary	The utilisation of containers or boxes that are temporarily used for the delivery process and not handed over to the customer upon delivery.	Fernie & Sparks, 2018; Tarn et al., 2003
	Reusable	The utilisation of reusable packages such as cartons to store items during the delivery process that are handed over to the customer upon delivery (e.g., delivery boxes).
	Disposal	The utilisation of disposable packages such as (plastic) bags to store items during the delivery process that are handed over to the customer upon delivery.
Delivery governance	Third-party	Collaboration with external partners in terms of delivery services (e.g., logistics service providers).	Aćimović et al., 2020; Jindal et al., 2021; Kureshi & Thomas, 2019; Tadić & Veljovic, 2020; Tarn et al., 2003; Vanelslander et al., 2013
	E-grocer	Individual management of deliveries and ownership of the delivery fleet of the e-grocer.
	Crowd	Management of delivery activities and delivery fleet by private individuals.
	Hybrid deliveries	Combination of external and internal assets.
Delivery routing	Static	Pre-calculated vehicle routes with master route acting as pre-defined routing condition.	Agatz et al. (2008); Anckar et al., 2002
	Dynamic	Flexible computation of delivery routes based on demand as well as traffic and other confounders.
Meta-dimension: Delivery operation
Order cooling	No cooling	Absence of cooling mechanisms during the delivery.	Choi et al., 2019; de Koster, 2002; Haider et al., 2020
	Insulated boxes	Insulated boxes for passive cooling.
	Cooled vehicles	Cooled vehicles for active cooling.
	Hybrid cooling	Insulated boxes and cooled vehicles.
Fleet – Penultimate-mile	Light commercial supply vehicle	Light commercial vehicles such as vans utilised to cross the distance between the point of fulfilment and the delivery intermediary (if present).	Aćimović et al., 2020; Aktas et al., 2021
	Heavy duty truck	Heavy duty trucks such as lorries utilised to cross the distance between the point of fulfilment and the delivery intermediary (if present).	Aćimović et al., 2020; Delaney-Klinger et al., 2003
	Private supply vehicle	Private vehicles utilised for deliveries between the point of fulfilment and the delivery intermediary (if present) in terms of crowd-shipping.	Choi et al., 2019; Winkenbach & Janjevic, 2018
	No transshipment vehicle	Absence of separate vehicles for transshipment processes due to an absence of a delivery intermediary.	Agatz et al. (2008); Janjevic & Winkenbach
Delivery intermediary	No intermediary	Absence of facilities where orders or goods are handled between point of fulfilment and final delivery destination.	Agatz et al. (2008); Janjevic & Winkenbach, 2020; Rotem-Mindali & Weltevreden, 2013; Vanelslander et al., 2013; Winkenbach & Janjevic, 2018
	Hub	A hub or spoke that is used to transship an order to a different vehicle of the delivery fleet. It can be stationary or mobile.
	Urban consolidation centre	Logistical facilities for the consolidation of goods and orders across multiple companies.
	Courier facilities	Infrastructure of a courier or logistics service provider that is used for fulfilling order through delivery boxes.
Fleet – Last-mile	Light commercial delivery vehicle	Light commercial vehicles such as vans utilised to cross the distance between point of fulfilment (if a delivery intermediary is not present) or delivery intermediary and the order destination.	Aćimović et al., 2020; Agatz et al. (2008); Aktas et al., 2021
	Cargo-Bike	Cargo bikes utilised to cross the distance between point of fulfilment (if a delivery intermediary is not present) or delivery intermediary and the order destination.	Aćimović et al., 2020; Choi et al., 2019
	E-Scooter	Very light commercial vehicles with electric engine utilised to cross the distance between point of fulfilment (if a delivery intermediary is not present) or delivery intermediary and the order destination.	Aćimović et al., 2020; Choi et al., 2019
	Autonomous ground vehicle	Autonomous guides vehicles with integrated storage facilities utilised to cross the distance between point of fulfilment (if a delivery intermediary is not present) or delivery intermediary and the order destination.	Liu et al., 2021; Figliozzi, 2020
	Unmanned aerial vehicle	Unmanned aerial vehicles such as drones utilised to cross the distance between point of fulfilment (if a delivery intermediary is not present) or delivery intermediary and the order destination.	Aćimović et al., 2020; Dayarian et al., 2020
	Private delivery vehicle	Private vehicles utilised for deliveries between point of fulfilment (if a delivery intermediary is not present) or delivery intermediary and the order destination in terms of crowd-shipping.	Choi et al., 2019; Winkenbach & Janjevic, 2018
Delivery tracking	No tracking	Abdication of technology to track deliveries.	Jain et al., 2017; Piroth et al., 2020
	Estimation	Calculatory tracking solution to inform the customer about the estimated arrival of the delivery based on routing predictions and historic delivery data.
	Live	Real-time tracking solution to inform the customer about the estimated arrival of the delivery with high accuracy (±10 minutes)
Reception type	Attended home-delivery	Customer must be present when receiving the goods.	Aćimović et al., 2020; Agatz et al. (2008); Amorim et al., 2020; Auf der Landwehr et al., 2020
	Unattended home-delivery	Orders are delivered to the defined destination without requiring physical presence of the customer. Deliveries are stored in shared reception box (fixed reception, such as lockers for a neighbourhood), dedication reception box (installed individually), delivery box (portable delivery boxes), or in-house (typically unlocked remotely by the customer).	Aćimović et al., 2020; Hübner et al. (2016); Marchet et al., 2018
	Customer collection	Online-orders are collected by the customer. The collection happens in-store and indifferent, in-store and dedicated, drive-in (drive-through or parking facility), lockers (automated locker facilities), or affiliates (store facilities such as post offices).	Agatz et al. (2008); Marchet et al., 2018; Pande et al., 2020
Delivery services	Pure delivery	Absence of additional services besides of the delivery.	Agatz et al., 20078; Robinson et al., 2007; Winkenbach & Janjevic, 2018
	Value added-services	Additional services upon delivery such as support (e.g., providing tips on product storage or usage), storage or transfer (e.g., carrying bags, storing products in fridge).
Meta-dimension: Return handling
Return procedure	No returns	E-grocers refuse returns and provide an instant refund.	Agatz et al. (2008); Haider et al., 2020; Kureshi & Thomas, 2019; Marchet et al., 2018; Rotem-Mindali & Weltevreden, 2013; Tarn et al., 2003
	By customer	Transfer of the return responsibility to the customer. To return orders/items, customers can choose between store (consult supermarket outlets), return point (specific return points), or courier (utilise a courier or parcel service).
	By e-grocer	The responsibility for initiating and handling returns of the e-grocer. Available options are instant returns (product inspection upon reception) or pick-up (pick-up service).
Failure recovery	No recovery	Absence of any compensation efforts for service failures.	Forbes et al., 2005
	Monetary	Provision of a monetary compensation for service failures of the e-grocer. Recoveries can be refund or discount.
	Physical	Failure recovery in terms of additional service provision. Recoveries in this context can describe replacement (replace the affected product) or correction (correct the error and provide an additional physical compensation).
	Hybrid recovery	The combination of monetary and physical recovery strategies, such as replacement of a faulty product with additional discount on the order.
Return flows	Integrated in inventory system	The integration of product return flows into the inventory system of the e-grocer.	Agatz et al. (2008); Tarn et al., 2003
	Separated from inventory system	The separation of product return flows from the inventory system of the e-grocer.

Table D1. E-grocery provider objects and attributes.

E-grocer instance	Website	Literature	Web-content
Albert Heijn NL	https://www.ah.nl	X	X
Albertsons US	n.a.	X
Amazon Fresh DE	www.amazon.de/amazonfresh		X
AmazonFresh IN	www.amazon.in/amazonfresh		X
AmazonFresh UK	http://www.amazon.co.uk/amazonfresh		X
AmazonFresh US	www.amazon.com/amazonfresh	X	X
Appiehein NL	https://www.appiehein.com		X
ASDA UK (2000)	n.a.	X
ASDA UK (2010)	https://www.asda.com	X	X
Auchan FR	https://www.auchan.fr	X	X
Aussie Health AUS	https://www.aussiehealthproducts.com.au		X
BİM A.Ş TU	https://www.bim.com.tr/	X	X
Borders US	n.a.	X
Boxed US	https://www.boxed.com		X
Bring24 DE	https://www.bring24.com		X
Bringmeister DE	https://www.bringmeister.de/		X
BringMirBio DE	https://bringmirbio.de		X
Caddyhome BE	n.a.	X
Carrefour FR	https://www.carrefour.fr	X	X
Casino FR	https://www.casino.fr/		X
Catch AUS	https://www.catch.com.au		X
Cold Storage SG	https://coldstorage.com.sg/		X
Coles AUS	https://www.coles.com.au	X	X
Com4Buy DE	https://www.com4buy.de		X
Coop UK	https://shop.coop.co.uk		X
Costco AUS	https://www.costco.com.au		X
Costco US	https://www.costco.com/		X
Courses U FR	www.coursesu.com		X
Delhaize BE	https://www.delhaize.be	X	X
Disco ARG	https://www.disco.com.ar/	X	X
Dunzo IN	https://www.dunzo.com/gurgaon		X
E. LeClerc FR	https://www.leclercdrive.fr/		X
Edeka24 DE	https://www.edeka24.de		X
Einkauf24 DE	n.a.	X
ENSO DE	https://www.myenso.de		X
Esselunga IT	https://www.esselunga.it/		X
Eurospar FI	n.a.	X
Food Direkt DE	https://food.de/		X
Foodie FI	https://www.foodie.fi		X
Foodly DE	https://www.getfoodly.com		X
Foodworks AUS	https://foodworks.com.au		X
Franprix FR	https://www.franprix.fr/		X
FreshDirect US	https://www.freshdirect.com	X	X
FrischeParadies DE	https://www.frischeparadies-shop.de		X
Frosta DE	https://www.frostashop.de		X
GetNow DE	https://www.getnow.com		X
Gorillas DE	https://gorillas.io		X
GreatFood US	n.a.	X
Grocery Gateway CA	https://www.grocerygateway.com/		X
Grofers IN	https://grofers.com/		X
Harris Farm AUS	https://www.harrisfarm.com.au		X
Holland at Home NL	https://www.holland-at-home.com		X
Hollandshop24 NL	https://www.hollandshop24.com		X
ICA SE	https://handla.ica.se	X	X
Iceland UK	https://www.iceland.co.uk		X
IGA AUS	https://igashop.com.au/		X
Instacart US	https://www.instacart.com/	X	X
Intermarche FR	https://www.intermarche.com/		X
JFC AUS	https://www.jfconline.shop		X
Jio Mart IN	https://www.jiomart.com/		X
Jumbo NL	https://www.jumbo.com		X
Kaisers DE	n.a.	X
Kmart US	https://www.kmart.com		X
Kroger US	https://www.kroger.com	X	X
LeShop SUI	n.a.	X
Metro CA	https://www.metro.ca/		X
Migros SUI	https://shop.migros.ch	X	X
Migros TU	https://www.migros.com.tr		X
Monoprix FR	https://www.monoprix.fr		X
Morrisons UK	https://groceries.morrisons.com		X
MyTime DE	https://www.mytime.de		X
MyWays SE	n.a.	X
My-World DE	n.a.	X
Natures Basket IN	https://www.naturesbasket.co.in		X
NetGrocer US	n.a.	X
Nettimarket FI	n.a.	X
Ocado UK	https://www.ocado.com/	X	X
Peapod US	https://www.peapod.com	X	X
Picnic NL	https://picnic.app/		X
Rappi ARG	https://www.rappi.com.ar/		X
Real DE	https://www.real.de/lebensmittelshop/		X
REWE Digital DE	https://shop.rewe.de	X	X
Rossmann DE	https://www.rossmann.de		X
Ruok@net FI	www.ruoka.net	X	X
Ruokamarkkinat Oy FI	https://www.k-ruoka.fi	X	X
Safeway UK	n.a.	X
Sainsbury UK (2000)	n.a.	X
Sainsbury UK (2010)	https://www.sainsburys.co.uk	X	X
Save on Foods CA	https://www.saveonfoods.com/		X
Shoplink US	n.a.	X
Shoprite US	https://www.shoprite.com		X
Shopwings DE	n.a.	X
S-Kanava FI	https://www.s-kanava.fi	X	X
SPUD CA	https://www.spud.ca/		X
Streamline US	n.a.	X
Supermarche Match FR	https://drive.supermarchesmatch.fr/		X
Target US	https://www.target.com		X
Tesco UK (2000)	n.a.	X
Tesco UK (2010)	https://www.tesco.com	X	X
The Fresh Market US	https://www.thefreshmarket.com		X
Thrifty Foods CA	https://www.thriftyfoods.com/		X
Trendyol TU	https://www.trendyol.com/		X
Waitrose UK	https://www.waitrose.com	X	X
Walmart CA	https://www.walmart.ca/		X
Walmart US	https://www.walmart.com	X	X
Webvan US	n.a.	X
Wellcome TWN	http://www.wellcome.com.tw/ENG	X	X
Woolworths AUS	https://www.woolworths.com.au	X	X
Yihaodian CHN	https://www.yhd.com/		X
Ykköshalli FI	n.a.	X
YümmyBazaar US	https://yummybazaar.com		X

Table E1. Cross-tabulator analysis on clustering results.

Taxonomy dimension	Taxonomy attributes (EGF attributes)	EGF archetypes
		#1 (n = 10)	#2 (n = 26)	#3 (n = 18)	#4 (n = 8)	#5 (n = 29)	#6 (n = 20)
Inventory planning	Demand forecast	50%	31%	6%	13%	10%	75%
	Safety stocks	0%	0%	17%	0%	48%	5%
	Inventory rationing	0%	12%	0%	13%	10%	15%
	Abdication of planning	10%	46%	78%	13%	24%	5%
	Hybrid planning	40%	12%	0%	63%	7%	0%
Inventory zoning	Store-like	50%	8%	100%	0%	72%	20%
	Pick-run optimized	10%	88%	0%	13%	14%	65%
	Region-based mix	40%	4%	0%	88%	14%	15%
Inventory governance	Contractor	0%	12%	67%	13%	3%	0%
	E-grocer	100%	85%	33%	38%	90%	90%
	Hybrid governance	0%	4%	0%	50%	7%	10%
Channel inventory mgmt.	Pure-online-channel inventory	20%	46%	61%	88%	10%	20%
	Offline- & online-channel inventory: Individual	20%	38%	6%	13%	31%	30%
	Offline- & online-channel inventory: Shared	60%	15%	33%	0%	59%	50%
Reservation time	Specific time	0%	23%	78%	88%	3%	0%
	Generic time	100%	77%	22%	13%	97%	100%
Point of fulfilment	Store: Outlet	50%	4%	89%	0%	62%	5%
	Store: Front Centre	10%	0%	0%	0%	3%	0%
	Distribution Centre: Centralized	0%	15%	11%	0%	14%	50%
	Distribution Centre: Decentralized & separated	0%	19%	0%	75%	3%	35%
	Distribution Centre: Decentralized & integrated	0%	58%	0%	0%	3%	5%
	Operational mix	40%	4%	0%	25%	14%	5%
Fulfilment point governance	Third-party	0%	15%	67%	13%	3%	0%
	E-grocer	60%	81%	33%	13%	97%	95%
	Hybrid structures	40%	4%	0%	75%	0%	5%
Order threshold	Absent	90%	62%	44%	13%	17%	15%
	Present	10%	38%	56%	88%	83%	85%
Picking procedure	Split-case picking	60%	12%	94%	25%	72%	20%
	Case-picking	40%	62%	0%	63%	28%	80%
	Pallet-picking	0%	27%	6%	13%	0%	0%
Picking technology	Person-to-goods	50%	12%	78%	13%	59%	0%
	Goods-to-person	50%	73%	17%	25%	34%	30%
	Mainly automated picking	0%	15%	6%	63%	7%	70%
Channel picking mgmt.	Pure-online-channel picking	0%	46%	61%	88%	10%	20%
	Offline- & online-channel picking: Integrated	40%	19%	0%	0%	7%	40%
	Offline- & online-channel picking: Separated	30%	12%	39%	13%	66%	30%
	Offline- & online-channel picking: Capacity optimized	30%	23%	0%	0%	17%	10%
Out-of-stock handling	Lead-time	0%	15%	6%	75%	14%	50%
	Substitute product	70%	19%	22%	13%	66%	25%
	Cancellation	30%	65%	72%	13%	21%	25%
Delivery area	Local	0%	0%	39%	0%	21%	10%
	Regional	40%	12%	50%	25%	52%	40%
	National	60%	73%	11%	0%	28%	50%
	International	0%	15%	0%	75%	0%	0%
Delivery scope	Urban only	0%	0%	44%	0%	24%	15%
	Sub-urban areas included	40%	0%	28%	13%	48%	40%
	Sub-urban & rural areas included	60%	100%	28%	88%	28%	45%
Delivery frame	Weekdays	0%	4%	28%	0%	0%	0%
	Saturdays included	70%	96%	56%	100%	83%	65%
	Saturdays & holidays included	30%	0%	17%	0%	17%	35%
Delivery lead-time	Within 30-minutes	0%	0%	6%	0%	3%	5%
	Within two hours	30%	0%	0%	0%	0%	0%
	Same-day	70%	0%	44%	63%	10%	0%
	Next-day	0%	12%	39%	38%	79%	95%
	Two or more days	0%	88%	11%	0%	7%	0%
Delivery time window	Narrow: 30-minutes	20%	0%	6%	0%	7%	5%
	Narrow: 1-hour	10%	0%	6%	0%	3%	15%
	Moderate: 2-hour	30%	0%	44%	63%	69%	55%
	Moderate: 3-hour	40%	4%	17%	13%	17%	20%
	Moderate: 4-hour	0%	8%	17%	0%	0%	0%
	Wide	0%	88%	11%	25%	3%	5%
Delivery charges	Free of charge	40%	0%	0%	0%	0%	0%
	Fixed: Instance-based	0%	0%	33%	0%	21%	0%
	Fixed: Subscription-based	0%	0%	0%	0%	0%	10%
	Dynamic: Lead-time	0%	0%	0%	0%	3%	10%
	Dynamic: Threshold	20%	58%	33%	25%	31%	40%
	Dynamic: Location	0%	12%	0%	0%	7%	5%
	Dynamic: Subscription	0%	27%	0%	13%	0%	0%
	Hybrid charges	40%	4%	33%	63%	38%	35%
Order packaging	Absent/temporary	0%	4%	28%	13%	24%	35%
	Reusable	0%	96%	56%	88%	34%	5%
	Disposable	100%	0%	17%	0%	41%	60%
Delivery governance	Third-party	0%	92%	44%	13%	10%	10%
	E-grocer	100%	4%	56%	13%	90%	85%
	Crowd	0%	0%	0%	0%	0%	0%
	Hybrid governance	0%	4%	0%	75%	0%	5%
Delivery routing	Static	10%	8%	28%	0%	14%	15%
	Dynamic	90%	92%	72%	100%	86%	85%
Order cooling	No cooling	90%	23%	0%	13%	0%	10%
	Insulated boxes	10%	73%	72%	63%	31%	15%
	Cooled vehicles	0%	4%	28%	13%	69%	0%
	Hybrid cooling	0%	0%	0%	13%	0%	75%
Fleet: Penultimate-mile	Light commercial supply vehicle	0%	23%	11%	0%	0%	35%
	Heavy duty truck	0%	69%	6%	88%	7%	65%
	Private supply vehicle	0%	0%	0%	0%	0%	0%
	No transshipment vehicles	100%	8%	83%	13%	93%	0%
Delivery intermediary	No intermediary	100%	8%	83%	13%	93%	0%
	Hub: Stationary	0%	0%	6%	75%	3%	100%
	Hub: Mobile	0%	0%	0%	0%	0%	0%
	Urban consolidation centre	0%	0%	0%	0%	3%	0%
	Courier facilities	0%	92%	11%	13%	0%	0%
Fleet: Last-mile	Light commercial delivery vehicle	0%	100%	61%	88%	97%	95%
	Cargo-bike	0%	0%	17%	0%	3%	0%
	E-Scooter	0%	0%	0%	13%	0%	5%
	Autonomous ground vehicle	0%	0%	0%	0%	0%	0%
	Unmanned aerial vehicle	0%	0%	0%	0%	0%	0%
	Private delivery vehicle	100%	0%	22%	0%	0%	0%
Delivery tracking	No tracking	70%	4%	39%	25%	10%	10%
	Estimation	20%	85%	61%	63%	90%	80%
	Live	10%	12%	0%	13%	0%	10%
Reception type	Attended home-delivery	0%	4%	67%	75%	100%	95%
	Unattended home-delivery: Shared reception box	0%	0%	0%	0%	0%	0%
	Unattended home-delivery: Dedicated reception box	0%	4%	0%	13%	0%	5%
	Unattended home-delivery: Delivery box	0%	92%	28%	13%	0%	0%
	Unattended home-delivery: In-house	0%	0%	0%	0%	0%	0%
	Customer collection: In-store and indifferent	0%	0%	0%	0%	0%	0%
	Customer collection: In-store and dedicated	40%	0%	6%	0%	0%	0%
	Customer collection: In-store and drive-in	50%	0%	0%	0%	0%	0%
	Customer collection: Lockers	10%	0%	0%	0%	0%	0%
	Customer collection: Affiliates	0%	0%	0%	0%	0%	0%
Delivery services	Pure delivery	40%	100%	100%	88%	66%	80%
	Value-added services	60%	0%	0%	13%	34%	20%
Return procedure	No returns	0%	42%	28%	88%	21%	55%
	By customer: Store	100%	15%	56%	0%	45%	15%
	By customer: Return point	0%	0%	0%	0%	0%	0%
	By customer: Courier	0%	42%	6%	0%	10%	10%
	By e-grocer: Instant returns	0%	0%	11%	13%	21%	20%
	By e-grocer: Pick-up	0%	0%	0%	0%	3%	0%
Failure recovery	No recovery	0%	0%	0%	0%	0%	0%
	Monetary: Refund	100%	88%	56%	38%	59%	85%
	Monetary: Discount	0%	8%	44%	13%	24%	10%
	Physical: Replacement	0%	0%	0%	0%	0%	0%
	Physical: Correction	0%	4%	0%	50%	0%	0%
	Hybrid recovery	0%	0%	0%	0%	17%	5%
Return flows	Integrated in inventory system	100%	23%	28%	0%	48%	30%
	Separated from inventory system	0%	77%	72%	100%	52%	70%

#1 – Pick-up service proponents#2 – Parcel champions#3 – Click-and-pick experts#4 – Collaboration specialists#5 – Store advocates#6 – Automation enthusiasts

Appendix C:

E-Grocery attributes

Following the taxonomy building approach (Nickerson et al., 2013), we relied on two sources to derive our overview of e-grocery attributes. First, during the conceptual iterations, we employed a systematic literature review (Webster and Watson 2002). The review was initially conducted in September 2020 and updated in February 2021, using the following search string without any search engine-based restrictions: (“e-grocery” OR “home-delivery” OR “online food”) AND “fulfilment” AND (“strategy” OR “scenario” OR “option” OR “model”). The queries were performed in the ACM, EBSCO, Emerald, IEEE, AIS, Taylor & Francis, ScienceDirect and JSTOR databases, focusing on the search domain title, abstract and keywords. In line with our main objective to identify sources supplying potential dimensions or attributes on different e-grocery concepts, we only included papers that are written in English and explicitly focus on grocery deliveries as well as fulfilment. Hence, articles related to general e-commerce or internet business were excluded from the results. Moreover, we analysed citations from the identified literature to determine less recent publications (backward search) and assessed citation indices to find articles quoting the key publications identified in earlier stages of the review (forward search). In total, we considered 80 scientific sources (4 books, 15 book chapters, 21 conference proceedings and 40 journal publications). Second, during the empirical iterations, we draw on our sample of 111 e-grocer instances (see overview in Appendix D) to refine and extend our taxonomy of EGF attributes.

Table C1 provides an overview of all attributes, a short description, and example references to literature to ensure transparency of the deductive procedure.

Appendix D:

E-Grocery provider

In this appendix, we provide an overview of real-life e-grocery providers. Between January and March 2020, we assessed the company websites of 111 e-grocery providers worldwide to derive additional dimensions and attributes of EGF or update priory identified approaches from (outdated) literature results. The longitudinal web-content analysis (WCA) was conducted based on the corporate websites listed in Table D1 as well as various industry reports (e.g., www.supermarketnews.com) about the e-grocery retailers.

The gathered sample of real-life objects (e-grocers) and literature articles were coded in three distinctive steps. First, we jointly discussed exemplary objects of our sample to get an initial shared understanding of how to classify the objects utilizing the taxonomy. Second, two coders from the author team independently classified a random 10% sample of our total set of e-grocery providers and compared the results afterwards. Overall, an intercoder reliability of 92.06% as percent agreement (Lombard et al., 2002) could be achieved (layer-specific percent agreement: 100% for ‘inventory management’; 87.5% for ‘order management’; 93% for ‘delivery management’; 92.86% for ‘return handling’). Third, based on this substantial agreement among the two coders, one coder continued to analyse the remaining objects of the sample.

Appendix E:

Archetypes –Clustering results

Appendix E.1Cross-tabulator analysis on clustering results

Table E1 states the relative frequencies of each characteristic in a given dimension for the six archetypes among 111 real-life examples of e-grocery providers.

Appendix E.2E-grocery provider instances per archetype

• Archetype #1 – Pick-up service proponents: Delhaize (Belgium), Walmart (Canada), Yihaodian (China), Auchan, Carrefour, Casino, Supermarche Match (France), Esselunga (Italy), Morrisons (UK), Walmart (USA).

• Archetype #2 – Parcel champions: Rappi (Argentina), Aussie Health, Catch, JFC (Australia), SPUD (Canada), E. LeClerc (France), BringMirBio, Edeka24, ENSO, FrischeParadies, Frosta, MyTime, Real, Rossmann (Germany), Jio Mart (India), Appiehein, Holland at Home NL, Hollandshop24 (Netherlands), LeShop (Switzerland), BİM A.Ş TU, Trendyol (Turkey), Boxed, Kmart, Target, Shoplink, YümmyBazaar (USA).

• Archetype #3 – Click-and-pick experts: Caddyhome (Belgium), Nettimarket, Ruok@net, Ykköshalli (Finland), Com4Buy, Einkauf24, Foodly, GetNow, Gorillas, My-World, Shopwings (Germany), Dunzo, Grofers (India), ICA, MyWays (Sweden), Instacart, Peapod, The Fresh Market (USA).

• Archetype #4 – Collaboration specialists: Amazon Fresh (Germany, India, UK and USA), Picnic (Netherlands), GreatFood, NetGrocer, Streamline (USA).

• Archetype #5 – Store advocates: E Disco (Argentina) Coles, Foodworks, Harris Farm, IGA, Woolworths (Australia), Foodie, Ruokamarkkinat Oy (Finland), Courses U, Franprix, Intermarche, Monoprix (France), Bring24, Food Direkt (Germany), Natures Basket (India), Albert Heijn, Jumbo (Netherlands), Cold Storage (Singapore), Coop, Iceland, Safeway, Sainsbury (2000), Sainsbury (2010), Tesco (2000), Waitrose (UK), Albertsons, Borders, FreshDirect, Shoprite (USA).

• Archetype #6 – Automation enthusiasts: Costco (Australia), Grocery Gateway, Metro, Save on Foods, Thrifty Foods (Canada), Eurospar, S-Kanava (Finland), Bringmeister, Kaisers, REWE Digital (Germany), Migros (Switzerland, Turkey), Wellcome (Taiwan), ASDA (2000; 2010), Ocado, Tesco (2010) (UK), Costco, Kroger, Webvan (USA).

Appendix

Aćimović, S., Mijušković, V., & Milošević, N. (2020). Logistics aspects of goods home delivery: The case of Amazon company. Marketing, 51(1), 3-11. https://doi.org/10.5937/markt2001003A

Agatz, N. A., Fleischmann, M., & Van Nunen, J. A. (2008). E-fulfilment and multi-channel distribution: A review. European Journal of Operational Research, 187(2), 339–356. https://doi.org/10.1016/j.ejor.2007.04.024

Aktas, E., Bourlakis, M., & Zissis, D. (2021). Collaboration in the last mile: evidence from grocery deliveries. International Journal of Logistics Research and Applications, 24(3), 227–241. https://doi.org/10.1080/13675567.2020.1740660

Alshammari, T., Alhadreti, O., & Mayhew, P. (2015). When to ask participants to think aloud: A comparative study of concurrent and retrospective think-aloud methods. International Journal of Human Computer Interaction, 6(3), 48–64.

Amorim, P., DeHoratius, N., Eng-Larsson, F., & Martins, S. (2020). Customer preferences for delivery service attributes in attended home delivery. Chicago Booth Research Paper, 20(7), 1-25.

Anckar, B., Walden, P., & Jelassi, T. (2002). Creating customer value in online grocery shopping. International Journal of Retail & Distribution Management, 30(4), 211–220. https://doi.org/10.1108/09590550210423681

Auf der Landwehr, M., Trott, M., & von Viebahn, C. (2020). Simulation-based assessment of grocery shopping in urban areas. Simulation Notes Europe, 30(4), 145–158. https://doi.org/10.11128/sne.30.tn.10533

Bensaou, M., & Venkatraman, N. (1995). Configurations of interorganizational relationships: A comparison between US and Japanese automakers. Management Science, 41(9), 1471–1492. https://doi.org/10.1287/mnsc.41.9.1471

Bock, M., & Wiener, M. (2017). Towards a taxonomy of digital business models: Conceptual dimensions and empirical illustrations. In Proceedings of the 38^th International Conference on Information Systems.

Boyer, K. K., & Hult, G. T. M. (2005). Extending the supply chain: Integrating operations and marketing in the online grocery industry. Journal of Operations Management, 23(6), 642–661. https://doi.org/10.1016/j.jom.2005.01.003

Boysen, N., De Koster, R., & Weidinger, F. (2019). Warehousing in the e-commerce era: A survey. European Journal of Operational Research, 277(2), 396–411. https://doi.org/10.1016/j.ejor.2018.08.023

Choi, Y., Schonfeld, P. M., Lee, Y. J., & Shin, H. S. (2021). Innovative methods for delivering fresh food to underserved populations. Journal of Transportation Engineering, Part A: Systems, 147(1), 04020140. https://doi.org/10.1061/JTEPBS.0000464

Cullen, A. J., & Webster, M. (2007). A model of B2B e‐commerce, based on connectivity and purpose. International Journal of Operations & Production Management, 27(2), 205–225. https://doi.org/10.1108/01443570710720621

Davies, A., Dolega, L., & Arribas-Bel, D. (2019). Buy online collect in-store: Exploring grocery click&collect using a national case study. International Journal of Retail & Distribution Management, 47(3), 278–291. https://doi.org/10.1108/IJRDM-01-2018-0025

Dayarian, I., Savelsbergh, M., & Clarke, J. P. (2020). Same-day delivery with drone resupply. Transportation Science, 54(1), 229-249. https://doi.org/10.1287/trsc.2019.0944

de Koster, R. M. B. (2002). Distribution structures for food home shopping. International Journal of Physical Distribution & Logistics Management, 32(5), 362–380. https://doi.org/10.1108/09600030210434152

Delaney-Klinger, K., Boyer, K. K., & Frohlich, M. (2003). The return of online grocery shopping: a comparative analysis of Webvan and Tesco’s operational methods. The TQM Magazine, 15(3), 187–196. https://doi.org/10.1108/09544780310469334

Dubosson‐Torbay, M., Osterwalder, A., & Pigneur, Y. (2002). E‐business model design, classification, and measurements. Thunderbird International Business Review, 44(1), 5–23. https://doi.org/10.1002/tie.1036

Durand, B., & Gonzalez-Feliu, J. (2012). Urban logistics and e-grocery: have proximity delivery services a positive impact on shopping trips? Procedia-Social and Behavioral Sciences, 39, 510–520. https://doi.org/10.1016/j.sbspro.2012.03.126

Emeç, U., Çatay, B., & Bozkaya, B. (2016). An adaptive large neighborhood search for an e-grocery delivery routing problem. Computers & Operations Research, 69, 109–125. https://doi.org/10.1016/j.cor.2015.11.008

Eriksson, E., Norrman, A., & Kembro, J. (2019). Contextual adaptation of omni-channel grocery retailers’ online fulfilment centres. International Journal of Retail & Distribution Management, 47(12), 1232–1250. https://doi.org/10.1108/IJRDM-08-2018-0182

Fernie, J., & Sparks, L. (2018). Logistics and retail management: Emerging issues and new challenges in the retail supply chain. Kogan Page Publishers.

Figliozzi, M. A. (2020). Carbon emissions reductions in last mile and grocery deliveries utilizing air and ground autonomous vehicles. Transportation Research Part D: Transport and Environment, 85, 102443. https://doi.org/10.1016/j.trd.2020.102443

Finch, H. (2005). Comparison of distance measures in cluster analysis with dichotomous data. Journal of Data Science, 3(1), 85-100. https://doi.org/10.6339/JDS.2005.03(1).192

Forbes, L. P., Kelley, S. W., & Hoffman, K. D. (2005). Typologies of e‐commerce retail failures and recovery strategies. Journal of Services Marketing, 19(5), 280–292. https://doi.org/10.1108/08876040510609907

Gevaers, R., Van de Voorde, E., & Vanelslander, T. (2011). Characteristics and typology of last-mile logistics from an innovation perspective in an urban context. In C. Macharis & S. Melo (Eds.), City Distribution and Urban Freight Transport: Multiple Perspectives (pp. 57–71). Edward Elgar Publishing.

Haider, Z., Hu, Y., Charkhgard, H., Himmelgreen, D., & Kwon, C. (2020). Creating grocery delivery hubs for food deserts at local convenience stores via spatial and temporal consolidation. SSRN Electronic Journal. http://dx.doi.org/10.2139/ssrn.3603062

Hands, S., & Everitt, B. (1987). A Monte Carlo study of the recovery of cluster structure in binary data by hierarchical clustering techniques. Multivariate Behavioral Research, 22(2), 235–243. https://doi.org/10.1207/s15327906mbr2202_6

Holland, C. P., & Gutiérrez-Leefmans, M. (2018). A taxonomy of SME e-commerce platforms derived from a market-level analysis. International Journal of Electronic Commerce, 22(2), 161–201. https://doi.org/10.1080/10864415.2017.13641144

Hübner, A., Holzapfel, A., & Kuhn, H. (2015). Operations management in multi-channel retailing: An exploratory study. Operations Management Research, 8(3), 84–100. https://doi.org/10.1007/s12063-015-0101-9

Hübner, A., Kuhn, H., Wollenburg, J. (2016). Last mile fulfilment and distribution in omni-channel grocery retailing. International Journal of Retail & Distribution Management, 44(3), 228–247. https://doi.org/10.1108/IJRDM-11-2014-0154

Jain, N. K., Gajjar, H., Shah, B. J., & Sadh, A. (2017). E-fulfilment dimensions and its influence on customers in e-tailing: A critical review. Asia Pacific Journal of Marketing and Logistics, 29(2), 347–369. https://doi.org/10.1108/APJML-11-2015-0167

Janjevic, M., & Winkenbach, M. (2020). Characterizing urban last-mile distribution strategies in mature and emerging e-commerce markets. Transportation Research Part A: Policy and Practice, 133, 164–196. https://doi.org/10.1016/j.tra.2020.01.003

Jindal, R. P., Gauri, D. K., Li, W., & Ma, Y. (2021). Omnichannel battle between Amazon and Walmart: Is the focus on delivery the best strategy?. Journal of Business Research, 122, 270–280. https://doi.org/10.1016/j.jbusres.2020.08.053

Johnson, O., & Iyamu, T. (2019). Framework for the adoption of e‐commerce: A case of South African retail grocery sector. The Electronic Journal of Information Systems in Developing Countries, 85(5). https://doi.org/10.1002/isd2.12095

Kämäräinen, V., Saranen, J., & Holmström, J. (2001). The reception box impact on home delivery efficiency in the e‐grocery business. International Journal of Physical Distribution & Logistics Management, 31(6), 414–426. https://doi.org/10.1108/09600030110399414

Kaplan, S., & Sawhney, M. (2000). B2B E-commerce hubs: Towards a taxonomy of business models. Harvard Business Review, 79(3), 97–100.

Kathuria, R. (2000). Competitive priorities and managerial performance: a taxonomy of small manufacturers. Journal of Operations Management, 18(6), 627-641.

Kathuria, R., Kathuria, N. N., & Kathuria, A. (2018). Mutually supportive or trade-offs: An analysis of competitive priorities in the emerging economy of India. The Journal of High Technology Management Research, 29(2), 227-236.

Kureshi, S., & Thomas, S. (2019). Online grocery retailing: Exploring local grocers beliefs. International Journal of Retail & Distribution Management, 47(2), 157–185. https://doi.org/10.1108/IJRDM-05-2018-0087

Lee, S., & Park, Y. (2009). The classification and strategic management of services in e-commerce: Development of service taxonomy based on customer perception. Expert Systems with Applications, 36(6), 9618–9624. https://doi.org/10.1016/j.eswa.2008.08.014

Lim, H., Widdows, R., & Hooker, N. H. (2009). Web content analysis of e‐grocery retailers: A longitudinal study. International Journal of Retail & Distribution Management, 37(10), 839–851. https://doi.org/10.1108/09590550910988020

Lim, S. F. W., & Srai, J. S. (2018). Examining the anatomy of last-mile distribution in e-commerce omnichannel retailing. International Journal of Operations & Production Management, 38(9), 1735–1764. https://doi.org/10.1108/IJOPM-12-2016-0733

Lim, S. F. W., Jin, X., & Srai, J. S. (2018). Consumer-driven e-commerce: A literature review, design framework, and research agenda on last-mile logistics models. International Journal of Physical Distribution & Logistics Management, 48(3), 308–332. https://doi.org/10.1108/IJPDLM-02-2017-0081

Liu, D., Deng, Z., Zhang, W., Wang, Y., & Kaisar, E. I. (2021). Design of sustainable urban electronic grocery distribution network. Alexandria Engineering Journal, 60(1), 145-157. https://doi.org/10.1016/j.aej.2020.06.051

Liu, X., Werder, K., & Mädche, A. (2016). A taxonomy of digital service design techniques. In Proceedings of the International Conference on Information Systems.

Lombard, M., Snyder‐Duch, J., & Bracken, C. C. (2002). Content analysis in mass communication: Assessment and reporting of intercoder reliability. Human Communication Research, 28(4), 587–604. https://doi.org/10.1111/j.1468-2958.2002.tb00826.x

Mahadevan, B. (2000). Business models for internet-based e-commerce: An anatomy. California Management Review, 42(4), 55–69. https://doi.org/10.2307/41166053

Marchet, G., Melacini, M., Perotti, S., Rasini, M., & Tappia, E. (2018). Business logistics models in omni-channel: a classification framework and empirical analysis. International Journal of Physical Distribution & Logistics Management, 48(4), pp. 439–464. https://doi.org/10.1108/IJPDLM-09-2016-0273

Martín, J., Pagliara, F., & Román, C. (2019). The research topics on e-grocery: Trends and existing gaps. Sustainability, 11, 321–336. https://doi.org/10.3390/su11020321

Melacini, M., Perotti, S., Rasini, M., & Tappia, E. (2018). E-fulfilment and distribution in omni-channel retailing: A systematic literature review. International Journal of Physical Distribution & Logistics Management, 48(4), 391–414. https://doi.org/10.1108/IJPDLM-02-2017-0101

Milligan, G. W., & Cooper, M. C. (1985). An examination of procedures for determining the number of clusters in a data set. Psychometrika, 50(2), 159–179. https://doi.org/10.1007/BF02294245

Mkansi, M., Eresia-Eke, C., & Emmanuel-Ebikake, O. (2018). E-grocery challenges and remedies: Global market leaders perspective. Cogent Business & Management, 5, 147–175. https://doi.org/10.1080/23311975.2018.1459338

Palmer, J., Kallio, J., Saarinen, T., Tinnila, M., Tuunainen, V. K., & van Heck, E. (2000). Online grocery shopping around the world: Examples of key business models. Communications of the Association for Information Systems, 4(3). https://doi.org/10.17705/1CAIS.00403

Pande, A., Gupta, A. D., Ni, K., Biswas, R., & Majumdar, S. (2020). Substitution techniques for grocery fulfillment and assortment optimization using product graphs. Proceedings of the 16^th International Workshop on Mining and Learning with Graphs, USA.

Peffers, K., Rothenberger, M., Tuunanen, T., & Vaezi, R. (2012). Design science research evaluation. International Conference on Design Science Research in Information Systems (pp. 398-410). Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-29863-9_29

Piroth, P., R.-M., E., & Bruwer, J. (2020). Digitalisation in grocery retailing in Germany: An exploratory study. The International Review of Retail, Distribution and Consumer Research, 30(5), 479–497. https://doi.org/10.1080/09593969.2020.1738260

Robinson, H., Dall’Olmo Riley, F., Rettie, R., & Rolls-Willson, G. (2007). The role of situational variables in online grocery shopping in the UK. The Marketing Review, 7(1), 89–106. https://doi.org/10.1362/146934707X180703

Rotem-Mindali, O., & Weltevreden, J. W. (2013). Transport effects of e-commerce: what can be learned after years of research?. Transportation, 40(5), 867-885. https://doi.org/10.1007/s11116-013-9457-6

Saleem, M., Khan, M. M., Ahmed, M. E., Neha Shah, S. A., & Surti, S. R. (2018). Online grocery shopping and consumer perception: A case of Karachi market in Pakistan. Journal of Internet and e-Business studies, 1–13. https://doi.org/10.5171/2018.931248

Seitz, C., Pokrivčák, J., Tóth, M., & Plevný, M. (2017). Online grocery retailing in Germany: An explorative analysis. Journal of Business Economics and Management, 18, 1243–1263. https://doi.org/10.3846/16111699.2017.1410218

Spiller, P., & Lohse, G. L. (1997). A classification of Internet retail stores. International Journal of Electronic Commerce, 2(2), 29–56. https://doi.org/10.1080/10864415.1997.11518307

Szopinski, D., Schoormann, T., & Kundisch, D. (2019). Because your taxonomy is worth it: Towards a framework for taxonomy evaluation. Proceedings of the 27th European Conference on Information Systems.

Tadić, S., & Veljović, M. (2020). Home delivery: Concept and characteristics. International Journal for Traffic and Transport Engineering, 10(4), 519-533. http://dx.doi.org/10.7708/ijtte.2020.10(4).10

Tarn, J.M., Razi, M.A., Wen, H.J., & Perez, A.A. (2003). E‐fulfilment: The strategy and operational requirements. Logistics Information Management, 16(5), 350–362. https://doi.org/10.1108/09576050310499345

Urbinati, A., Chiaroni, D., & Chiesa, V. (2017). Towards a new taxonomy of circular economy business models. Journal of Cleaner Production, 168, 487–498. https://doi.org/10.1016/j.jclepro.2017.09.047

Vakulenko, Y., Hellström, D., & Hjort, K. (2018). What’s in the parcel locker? Exploring customer value in e-commerce last mile delivery. Journal of Business Research, 88, 421–427. https://doi.org/10.1016/j.jbusres.2017.11.033

Vakulenko, Y., Shams, P., Hellstrom, D., & Hjort, K. (2019). Service innovation in e-commerce last mile delivery: Mapping the e-customer journey. Journal of Business Research, 101, 461–468. https://doi.org/10.1016/j.jbusres.2019.01.016

Vanelslander, T., Deketele, L., & Van Hove, D. (2013). Commonly used e-commerce supply chains for fast moving consumer goods: comparison and suggestions for improvement. International Journal of Logistics Research and Applications, 16(3), 243-256. https://doi.org/10.1080/13675567.2013.813444

Vazquez-Noguerol, M., Comesaña-Benavides, J., Poler, R., & Prado-Prado, J. C. (2020). An optimisation approach for the e-grocery order picking and delivery problem. Central European Journal of Operations Research. https://doi.org/10.1007/s10100-020-00710-9

Vinsensius, A., Wang, Y., Chew, E. P., & Lee, L. H. (2020). Dynamic incentive mechanism for delivery slot management in e-commerce attended home delivery. Transportation Science, 54(3), 567–587. https://doi.org/10.1287/trsc.2019.0953

Webster, J., & Watson, R. T. (2002). Analyzing the past to prepare for the future: Writing a literature review. MIS Quarterly, xiii–xxiii.

Weking, J., Hein, A., Böhm, M., & Krcmar, H. (2018). A hierarchical taxonomy of business model patterns. Electronic Markets, 30, 447–468. https://doi.org/10.1007/s12525-018-0322-5

Wen, H. J., Chen, H. G., & Hwang, H. G. (2001). E‐commerce web site design: Strategies and models. Information Management & Computer Security, 9(1), 5–12. https://doi.org/10.1108/09685220110366713

Winkenbach, M., & Janjevic, M. (2018). Classification of last‐mile delivery models for e‐commerce distribution: A global perspective. In E. Taniguchi & R. G. Thompson (Eds.), City Logistics 1: New Opportunities and Challenges (pp. 209–229). Wiley.

Wollenburg, J., Hübner, A., Kuhn, H., & Trautrims, A. (2018). From bricks-and-mortar to bricks-and-clicks. International Journal of Physical Distribution & Logistics Management, 48(4), 415–438. https://doi.org/10.1108/IJPDLM-10-2016-0290

Xing, Y., Grant, D. B., McKinnon, A. C., & Fernie, J. (2010). Physical distribution service quality in online retailing. International Journal of Physical Distribution & Logistics Management, 40(5), 415–432. https://doi.org/10.1108/09600031011052859

Yousept, I., & Li, F. (2005). Building an online grocery business: The case of asda.com. International Journal of Cases on Electronic Commerce, 1(2), 57–78. https://doi.org/10.4018/jcec.2005040104

Zhu, S., Hu, X., Huang, K., & Yuan, Y. (2021). Optimization of product category allocation in multiple warehouses to minimize splitting of online supermarket customer orders. European Journal of Operational Research, 290(2), 556–571. https://doi.org/10.1016/j.ejor.2020.08.024