Exploring purchase intention in cross-border e-commerce: a relational view Open Access

Cross-border e-commerce (CBEC) – the online trade between retailers and consumers across national borders – has significantly expanded opportunities for interaction (Samiee, 2020; Do et al., 2023). Current forecasts suggest that the business-to-consumer (B2C) CBEC industry will surpass US$1.21tn by the end of 2025, up from US$562bn in 2019 (Statista, 2025a). Meanwhile, rising tariffs worldwide (Statista, 2025b) and geopolitical tensions in the global business environment (Fan et al., 2025) have increased consumer uncertainty in CBEC purchases. During this period, consumer buying behavior has changed significantly, and consequently, consumer purchase intention in CBEC has gained attention in the international marketing literature (e.g. see Han and Kim, 2019; Mou et al., 2020a; Xu et al., 2020; Do et al., 2023; Wang et al., 2023; Lee and Xiong, 2026). Understanding how consumers develop purchase intentions has been at the forefront, as it enables retailers to improve their marketing strategies (Kim et al., 2017; Ipsmiller et al., 2022) and establish the foundation for long-term international relationships (Cui et al., 2020). In this area of research, most studies have concentrated on consumer purchase intentions related to large, well-known multinational retailers (Luo and Ye, 2019; Mou et al., 2020a; Chen and Yang, 2021; Xu et al., 2023; Lee and Xiong, 2026).

However, studying purchase intentions toward large, well-known retailers provides an inadequate understanding of how consumers develop purchase intentions toward smaller, resource-limited, unfamiliar retailers in CBEC (Hu et al., 2024; Gong et al., 2024; Da Rocha et al., 2024; Lee et al., 2024). This is a critical issue because strategies used by larger retailers to drive consumer purchases may not be feasible for many smaller retailers, as they are often costly and require significant resources (Ipsmiller et al., 2022). Smaller retailers must manage their limited resources and adopt alternative, more cost-effective strategies. One such approach involves developing country-specific websites, also known as partial online internationalization (hereafter POI-retailer) (e.g. see Daryanto et al., 2013; Safari and Yamin, 2016; Schu and Morschett, 2017; Swoboda and Sinning, 2022; Wistedt, 2024). This involves replicating their home-country website for the host country, with some local adaptations (Schu and Morschett, 2017; Wistedt, 2024). While this provides a cost-effective way to address language barriers and the liability of foreignness (Schu and Morschett, 2017; Swoboda and Sinning, 2022), the main drawback is that limited resources hinder the development of a well-known brand. As a result, many consumers they aim to attract remain unfamiliar, which thereby creates consumer uncertainty. This uncertainty, caused by the POI-retailer's unfamiliarity, can impede the retailer's long-term growth by presenting a significant barrier to consumer purchasing (Darke et al., 2016; Sharma and Klein, 2024) and, more importantly, may obstruct relationship building (Cui et al., 2020; Cheng et al., 2022; Lin et al., 2023). Consumers attempt to reduce this uncertainty, but it remains challenging when dealing with unfamiliar retailers (Darke et al., 2016; Sharma and Klein, 2024). Recent reports indicate that nearly seventy percent of consumers prefer to develop relationships and make purchases from large, well-known retailers (e.g. Amazon and Zalando) in CBEC (Statista, 2025c). This clearly suggests that purchasing from POI-retailers is less desirable, yet how these retailers can better attract consumers has not been adequately addressed in scholarly research. We argue that it is vital to explore how POI-retailers can create conditions for building consumer relationships, with the first step being understanding the mechanisms that drive consumers' purchase intentions toward these retailers. Therefore, this paper aims to investigate consumer purchase intentions toward a POI-retailer in CBEC from a relational view.

We build and extend the body of knowledge in CBEC by focusing on consumer purchasing in the context of POI-retailers. In doing so, our study has several goals. First, extant research remains limited to studying consumer purchase intention in the context of POI-retailers. That is, previous studies have focused on purchase intention from large, well-known retailers in CBEC (e.g. Huang and Chang, 2019; Cui et al., 2020; Chen and Yang, 2021) and the POI-retailer's online internationalization journey (e.g. Daryanto et al., 2013; Schu and Morschett, 2017; Swoboda and Sinning, 2022). Although a few studies have shed light on consumer purchasing toward POI-retailers, these studies are either conceptual (Safari and Yamin, 2016) or focus on the technological antecedents (Wistedt, 2024). We wish to establish how consumers form their perceptions and willingness to purchase when interacting with an unfamiliar POI-retailer in CBEC. Therefore, the first goal of this paper is to explore what influences and how consumers develop purchase intentions in the context of POI-retailers. It is important to study this consumer–POI-retailer context because it is linked to a high level of uncertainty, which will help understand how consumers develop purchase intentions under such uncertain conditions.

Second, present research offers limited insights from a relational view when studying purchase intentions under conditions of uncertainty, particularly when the uncertainty arises from the retailer's unfamiliarity. Previous studies have drawn on relational exchange theory and acknowledged the value of a relational view in studying purchase intentions (e.g. Pavlou et al., 2007; Eastlick et al., 2006; Mukherjee and Nath, 2007; Palvia, 2009; Cui et al., 2020; Lin et al., 2023). However, these studies focus on the effects of relational antecedents (i.e. trust and commitment) on consumer purchasing in ongoing relationships with a retailer, where uncertainty is already reduced most of the time (e.g. see Mukherjee and Nath, 2007; Cui et al., 2020; Lin et al., 2023). This overlooks a critical inquiry: how relational antecedents function when consumers do not have an established relationship with the retailer. Accordingly, when consumers are unfamiliar with a retailer, developing trust is troublesome as it requires consumers to first cope with uncertainty (Darke et al., 2016; Sharma and Klein, 2024). Therefore, it is vital to examine this inquiry because relational antecedents may be ineffective for consumers' purchase intentions due to the adverse effect of uncertainty. We build and extend this notion by drawing on consumer attitude literature. Consumer attitude can shape positive perceptions about the retailer and enhance purchase intentions (Patel et al., 2023; Pop et al., 2023). In this light, the role of attitude may strengthen the relational antecedents and foster purchase intention despite the perception of uncertainty. With this in mind, we incorporate attitude into the relational antecedents of trust, commitment and uncertainty to develop a comprehensive framework for studying purchase intentions. Therefore, our second goal is to introduce these concepts into a unified framework to develop a relational framework for studying consumer purchase intention toward POI-retailers from a relational view. We argue that it is essential to use the concept of attitude in this context because it is linked to a high level of uncertainty. While in established relationships, purchasing intention and behavior are influenced by trust and commitment, in the consumer–POI-retailer context, purchase intention under high uncertainty may be derived from consumer attitude.

This paper has theoretical implications for the CBEC literature and managerial implications for managers. To begin with, we shed light on consumer purchase intention toward POI-retailers, specifically the formation of consumer purchase intention under high levels of uncertainty, which has been neglected in the CBEC literature. We provide novel empirical evidence to support our proposed relational framework, based on consumers' encounters with a real-life POI-retailer in CBEC. Theoretically, our research builds and expands the CBEC literature by offering a relational framework for studying purchase intention. Specifically, we connect the relational exchange theory and the literature on consumer attitude into a unified framework, which has, thus far, been overlooked. While this study focuses on consumer purchasing from a specific type of retailer, our proposed relational framework is expected to serve as a pivotal foundation for future research in CBEC, particularly for those exploring consumer–retailer relationships. Especially, we shed light on the earlier phases of relationship building and consumer purchase intention formation under the conditions of high levels of uncertainty. This study contributes to CBEC and relationship marketing research by examining whether core assumptions of relational exchange theory hold in a high-uncertainty, unfamiliar POI-retailer context. Contrary to the dominant literature, our findings show that trust does not directly drive either commitment or purchase intention; rather, consumer attitudes and commitment are the primary mechanisms shaping purchase intentions. By revealing how established relational concepts operate differently under conditions of unfamiliarity and uncertainty, this study identifies important boundary conditions of relational exchange theory in CBEC.

Managerially, our research offers valuable insights for managers with less familiar brands. Given that uncertainty is a significant barrier to consumer purchasing, our study suggests that managers can influence purchase intentions. In doing so, nurturing favorable attitudes is a crucial initial step, as it drives consumer commitment and purchase intentions. Moreover, attitude serves as a mediator of trust's effect on commitment and purchase intention. Accordingly, our proposed relational framework outlines key areas to focus on in creating prerequisites for building relationships with consumers internationally.

The rest of the paper is organized as follows. First, we outline existing research on consumer purchase intention in CBEC and highlight that consumers can purchase from three types of retailers in CBEC. Here, we demonstrate that the literature on purchase intention in CBEC contains three key sub-literatures, each of which we review and discuss. We further review relational exchange theory by discussing its foundation and key concepts, followed by delving into the ideas from the consumer attitude literature. Next, we connect these two ideas and build a relational framework for studying consumer purchase intention in CBEC. We then use this framework to develop hypotheses in the context of POI-retailers, which we test by exposing this study's respondents to a real-life POI-retailer and its country-specific website. The paper further discusses the findings, highlighting that purchase intention toward POI-retailers differs from past research. Finally, we conclude with theoretical and practical implications, limitations, and avenues for future research.

Consumer purchase intention – defined as a consumer's willingness to buy from a retailer – is a central research focus in CBEC (Dodds et al., 1991; Mou et al., 2020a; Do et al., 2023; Wang et al., 2023; Lee and Xiong, 2026). It offers insights into consumer purchasing behavior (Hazarika and Mousavi, 2021; Lee and Xiong, 2026), enables the prediction of future behavioral outcomes (Zerbini et al., 2022) and creates prerequisites for building relationships (Cui et al., 2020; Lin et al., 2023). While CBEC has unlocked remarkable opportunities for parties to interact worldwide easily, significant obstacles remain affecting consumers' willingness to purchase (Han and Kim, 2019; Lee and Xiong, 2026). According to extant studies, these obstacles include language barriers (Alcantára-Pilar et al., 2017), different legal systems (Huang and Chang, 2019), product information asymmetry (Zhu et al., 2020; Lee and Xiong, 2026), payment security, complex deliveries (Kim et al., 2017; Do et al., 2023) and tariffs (Shao et al., 2021). All these factors cause consumer uncertainty, referred to as the difficulty of accurately anticipating the outcome (Pavlou et al., 2007). Uncertainty has especially negative consequences for trust (Pavlou et al., 2007; Lu and Chen, 2021) and consumer purchase intentions in CBEC (Mou et al., 2020b; Lee and Xiong, 2026). Therefore, the success of enhancing consumer purchase intentions depends much on the retailer's resource investments in CBEC to overcome these obstacles and thereby reduce this uncertainty (Ma et al., 2022). In their studies (e.g. Yamin and Sinkovics, 2006; Sinkovics et al., 2013; Safari and Yamin, 2016; Ipsmiller et al., 2022), researchers discuss how retailers can employ three strategies to mitigate uncertainty and enhance consumer purchasing in CBEC. Specifically, they argue that consumers can purchase from three types of retailers in CBEC, which include (I) Default Online Internationalization (DOI) retailers, (II) Active Online Internationalization (AOI) retailers and (III) POI retailers (for a detailed comparison of these retailers, see  Appendix 1). That is, the literature on consumer purchase intention in CBEC can be organized into three sub-literatures, which we review below.

The first sub-literature investigates consumer purchase intention in the context of DOI-retailers (e.g. see Safari, 2012; Safari et al., 2013; Huang and Chang, 2019; Ma et al., 2022; Wagner et al., 2024). In short, DOI-retailers lack interest in CBEC activities as they primarily focus on their domestic market and its consumers (Yamin and Sinkovics, 2006; Sinkovics et al., 2013; Ipsmiller et al., 2022). Although they do not target consumers outside their home country, foreign consumers can purchase from the retailer under two conditions: (1) if they can cope with language barriers, and (2) if the retailer is willing to ship the product overseas (Safari and Yamin, 2016). In this sub-literature, previous research identifies the drivers and obstacles in consumer purchase intentions toward DOI-retailers. While the drivers include trust (Safari, 2012; Wagner et al., 2023), perceived benefits (Wagner et al., 2024) and website design (Huang and Chang, 2019), the obstacles involve perceived costs (Huang and Chang, 2019), and psychic distance stemming from language barriers, different regulations and cultures (Safari et al., 2013; Ma et al., 2022). Next, the second sub-literature investigates consumer purchase intention in the context of AOI-retailers (e.g. see Singh et al., 2006; Luo and Ye, 2019; Mou et al., 2020a; Xu et al., 2023; Lee and Xiong, 2026). Simply put, AOI-retailers are large multinational retailers that invest heavily in CBEC and actively target foreign consumers from multiple foreign markets simultaneously (Ipsmiller et al., 2022). In doing so, they adapt to foreign consumers' preferences by, for example, offering multilingual native and culturally adapted websites, convenient shipping with local warehouses, local payment methods, tailored local product offerings and customer support in consumers' mother tongues (Safari and Yamin, 2016). In this second sub-literature, past studies have identified the significant drivers of purchase intention toward AOI-retailers such as trust (Hsu et al., 2022), commitment (Cui et al., 2020), product involvement, information quality (Mou et al., 2020a; Zhu et al., 2020; Shao et al., 2021; Xu et al., 2023), perceived value (Luo and Ye, 2019; Mou et al., 2020b) and high cultural adaptation (Singh et al., 2006). In contrast to the drivers, research indicates that obstacles to developing purchase intentions toward AOI-retailers include the retailer's opportunistic behavior (Cui et al., 2020) and consumer dissatisfaction (Lin et al., 2018). To summarize, although there is an extensive body of empirical and theoretical knowledge on consumer purchase intentions toward both DOI-retailers and AOI-retailers, we still have a relatively limited understanding of purchase intentions in the context of the third type of retailer, namely POI-retailers.

In this paper, we build and extend the third type of sub-literature in CBEC, namely consumer purchase intention in the context of POI-retailers (see Table 1). POI-retailers are smaller retailers interested in CBEC activities, but they can only make partial investments (Schu and Morschett, 2017; Wistedt, 2024). On the one hand, they develop country-specific websites targeting consumers in a specific foreign country (Safari and Yamin, 2016). To achieve this cost-effectively, they replicate their domestic website for a foreign market by making partial adaptations related to the local language, unique resource locators (i.e. web addresses) and currency (Schu and Morschett, 2017). On the other hand, they cannot afford convenient shipments (i.e. local warehouses), customer support in the consumers' mother tongue and cultural adaptation (Safari and Yamin, 2016). Most significantly, their limited resources prevent them from establishing a well-known brand name in the host market, resulting in most consumers they seek to attract being unfamiliar with them (Wistedt, 2024). This unfamiliarity creates consumer uncertainty when purchasing from them, as they have no prior experience anticipating the outcome (Safari and Yamin, 2016; Wistedt, 2024). Uncertainty related to the retailer and its products becomes strongly tangible and prominent for the consumer due to a lack of familiarity with the retailer (Darke et al., 2016; Sharma and Klein, 2024).

A closer review of the third sub-literature indicates that research in the consumer–POI-retailer context is still in its infancy (see Table 1). Prior studies have primarily focused on the POI-retailer itself, investigating its market selection (Schu and Morschett, 2017), firm growth (Swoboda and Sinning, 2022) and behavioral adoption of country-specific websites (Daryanto et al., 2013). Although few contributions have shed light on consumer purchasing in the context of POI-retailers, either through a conceptual discussion (Safari and Yamin, 2016) or by studying the influence of technological antecedents (Wistedt, 2024), we still know relatively little regarding how consumers develop purchase intentions in the context of these retailers under the condition of uncertainty. Notably, both studies, Safari and Yamin (2016) and Wistedt (2024), highlight the essential role of consumer uncertainty in this purchasing context. Nevertheless, despite these discussions, there is limited empirical evidence on how uncertainty influences the development of consumer purchase intention in this context. Therefore, to extend this idea, we adopt the concept of uncertainty in our relational framework. In doing so, we follow others (Pavlou et al., 2007) and define uncertainty as the consumer's difficulty in predicting future outcomes, including product and retailer uncertainty, forming the consumer's overall uncertainty.

Based on the above, we expand empirical and theoretical knowledge by studying this novel consumer purchasing context in CBEC. We address the lack of research on (1) consumer purchase intention toward POI-retailers, and (2) understanding the impact of consumer uncertainty stemming from the POI-retailer's unfamiliarity. Specifically, we build a relational framework by drawing on the ideas from relational exchange theory by highlighting the role of relational antecedents (i.e. trust and commitment) under conditions of uncertainty. Furthermore, we extend its scope by incorporating ideas from the literature on consumer attitude. It is important to adopt the concept of attitude in conditions of high uncertainty since consumers' attitudinal evaluations are expected to play a pivotal role in affecting consumer purchase intention in situations where uncertainty causes hesitancy and avoidance behavior. Therefore, we seek to introduce these concepts into a unified framework for studying consumer purchase intention under uncertainty stemming from the retailer's unfamiliarity. Next, we discuss the foundation and core concepts of relational exchange theory, followed by an examination of ideas from the consumer attitude literature.

In marketing, there are two foundational theoretical paths to studying consumer purchase intention: Transactional and Relational exchange theories. Thus, all other marketing theories are mid-range sub-theories/perspectives derived from the foundations of these two general theories (Dwyer et al., 1987; Safari and Albaum, 2019). The transaction exchange theory views purchasing as the ultimate goal, focusing on a short-term orientation, rationality, opportunistic behavior, switching behavior and uncertainty avoidance (Lee and Xiong, 2026). The relational exchange theory, on the other hand, views purchase intention as the entrance for developing a relationship, with a focus on long-term orientation, bounded rationality, mutual benefits between buyers and sellers, and that these parties use relationships for coping with uncertainty (Morgan and Hunt, 1994; Parvatiyar and Sheth, 2000; Mukherjee and Nath, 2007; Pavlou et al., 2007; Safari and Albaum, 2019). In this study, the foundation of our proposed relational framework is built upon the ideas from relational exchange theory.

However, there are different variations of relational exchange theory (for example, see Dwyer et al., 1987; Johanson and Mattson, 1987; Morgan and Hunt, 1994; Parvatiyar and Sheth, 2000), with minor differences in perspectives existing among these relational theorists. The theory has historically been used in the business-to-business (B2B) context (Dwyer et al., 1987; Johanson and Mattson, 1987; Morgan and Hunt, 1994), but in recent years has also been confirmed of its use in the B2C online context (Eastlick et al., 2006; Mukherjee and Nath, 2007; Cui et al., 2020; Lin et al., 2023). Despite the different variations of relational exchange theory and the fact that the nature of B2B and B2C contexts differ, scholars acknowledge the shared reasoning on relationships and hence share a common foundation, namely, they are all built on the two core concepts of trust and commitment (Parvatiyar and Sheth, 2000; Safari and Albaum, 2019). In other words, trust and commitment are essential for developing, building and maintaining long-term relationships between parties (Mukherjee and Nath, 2007). In ongoing relationships, trust fosters confidence in one another and is crucial in reducing uncertainty, serving as a pivotal prerequisite for commitment (Eastlick et al., 2006; Pavlou et al., 2007). The commitment reflects the parties' investments in building and maintaining a long-term, mutually beneficial relationship (Safari and Albaum, 2019; Lin et al., 2023). In this light, previous research has confirmed the significant effect of consumer trust on commitment, and their joint impact on consumer purchase intention (Eastlick et al., 2006; Mukherjee and Nath, 2007; Wang et al., 2016; Cui et al., 2020; Lu and Chen, 2021; Lin et al., 2023).

In this paper, we build on these ideas and adopt trust and commitment to develop our relational framework. To begin with, we follow previous studies (Eastlick et al., 2006) and define commitment as the consumer's interest in maintaining interactions with the retailer, as reflected in their support, recommendations and choice to purchase from the retailer over other retailers. Furthermore, in line with others (Pavlou et al., 2007), we define trust as consumers' acceptance of vulnerability based on his/her beliefs that buying from the retailer will meet their expectations. Several studies argue for the multidimensionality of trust (McKnight and Chervany, 2001; Pavlou et al., 2007; Oliveira et al., 2017; Tam et al., 2020). While there are several alternatives to categorize the dimensions of trust, we follow studies in e-commerce and CBEC, which mostly conceptualize trust based on three primary dimensions: benevolence, integrity and competence (McKnight and Chervany, 2001; Pavlou et al., 2007; Oliveira et al., 2017; Tam et al., 2020). Thus, benevolence refers to the likelihood that a retailer keeps the consumer's best interest in mind instead of engaging in opportunistic behavior. In contrast, integrity refers to the retailer acting honestly, reliably and consistently when fulfilling its promises (Tam et al., 2020). The third dimension of trust is related to the retailer's competence; this includes the retailer's ability to keep its promises to consumers (Pavlou et al., 2007; Tam et al., 2020). Accordingly, these three dimensions vary but are interrelated, forming consumers' overall trust in the retailer.

Although previous studies have argued for the use and importance of relational exchange theory in the field of consumer marketing research, particularly in domestic e-commerce (Eastlick et al., 2006; Mukherjee and Nath, 2007; Palvia, 2009; Lin et al., 2023), and have begun to apply it in CBEC research (Safari, 2012; Cui et al., 2020; Cheng et al., 2022). However, these studies have primarily focused on established and ongoing relationships between consumers and retailers, where uncertainty is already significantly reduced (Mukherjee and Nath, 2007; Lin et al., 2023; Cui et al., 2020). In other words, the role of uncertainty on relational antecedents prior to relationship development remains unclear. The relational antecedents may be ineffective in influencing consumer purchase intentions due to the negative impact of uncertainty. Theoretically, relational theorists would argue that developing purchase intentions and thereby building a relationship is challenging, as uncertainty impedes trust and hinders commitment (Parvatiyar and Sheth, 2000; Eastlick et al., 2006; Mukherjee and Nath, 2007; Pavlou et al., 2007; Lin et al., 2023). While we share their view, we also extend this by arguing for the importance of understanding whether there is an alternative way for consumers to develop purchase intentions despite the impediment of uncertainty on trust. Here, we emphasize the importance of exploring the link between attitude and the relational antecedents. Although few scholars discuss consumer attitudes in consumer–retailer relationships (e.g. see Kim and Kim, 2021; Alvarez et al., 2023), they have instead been treated as a “background” concept and, thus far, not properly linked with the relational antecedents into a common framework. We argue that in the early phases of relationship development under conditions of high uncertainty, consumer attitude is expected to play a crucial role in forming purchase intention. Therefore, we draw on the consumer attitude literature, which we discuss in the following section.

The concept of attitude has received significant attention in scholarly research. Specifically, past theorists have acknowledged the importance of an individual's attitude across various academic fields for decades, including information technology (Davis, 1993), social psychology (Fishbein and Ajzen, 1977; Ajzen, 2001) and anthropology (Agheyisi and Fishman, 1970). Thus, extensive research confirms that attitude plays a crucial role in shaping behavioral intention, decision-making and actual behavior, both directly and indirectly (Cabeza-Ramírez et al., 2022; Zerbini et al., 2022; Patel et al., 2023). The concept of attitude has also been acknowledged in marketing research (e.g. see meta-analysis by Zerbini et al., 2022), highlighting that consumer attitude is pivotal in shaping consumer behavior (Hultman et al., 2017; Cabeza-Ramírez et al., 2022; Sun et al., 2022; Patel et al., 2023; Boukis et al., 2024). While a positive attitude can increase favorable perceptions and beliefs about a retailer and drive purchase intentions, negative attitudes can decrease these outcomes (Zerbini et al., 2022). In other words, consumers' favorable and unfavorable attitudes play a crucial role in performing specific behaviors (Hultman et al., 2017; Zerbini et al., 2022; Li et al., 2024). Previous studies have confirmed the significant effect of attitude on consumer purchase intentions in various purchasing contexts, such as e-commerce (e.g. see Patel et al., 2023; Pop et al., 2023) and CBEC (e.g. see Singh et al., 2006; Alcantára-Pilar et al., 2017). Others have recently discussed the importance of favorable attitudes toward a retailer in consumers' development of a relationship with the retailer (Kim and Kim, 2021; Alvarez et al., 2023). Therefore, studying consumer attitudes is essential for understanding and explaining consumer purchase intentions, which can create prerequisites for building consumer–retailer relationships.

However, despite the highlighted importance, attitude remains underexplored in connection to both relational antecedents (i.e. trust and commitment), and particularly how attitude is shaped by consumer uncertainty stemming from a retailer's unfamiliarity. Although few studies have shown a significant effect of trust on attitude (Al-Debei et al., 2015; Akroush and Al-Debei, 2015; Li et al., 2024), it remains unclear how uncertainty impacts attitude and how this, in turn, affects consumers' commitment and purchase intentions. This requires further consideration, as it facilitates the theoretical connection between these concepts and, more importantly, offers better insights into understanding consumer purchase intentions from POI-retailers. Following Bhattacherjee and Premkumar (2004), we define consumer attitude as the extent to which a consumer perceives purchasing from the retailer as favorable. We argue that it is important to incorporate attitude into the relational exchange literature since attitude is expected to play a crucial role in building commitment and influencing purchase intention in the early stages of the relationship due to their lack of experience with the retailer.

This study examines consumer purchase intention in the context of POI-retailers within the CBEC from a relational view. Drawing on relational exchange theory (Eastlick et al., 2006; Mukherjee and Nath, 2007; Pavlou et al., 2007; Safari and Albaum, 2019; Lin et al., 2023) and the consumer attitude literature (Patel et al., 2023; Pop et al., 2023; Boukis et al., 2024), we connect these ideas into a unified model and introduce a relational framework for studying purchase intention. Figure 1 illustrates the proposed conceptual model. The model demonstrates that consumer uncertainty determines trust and attitude. These two, together with commitment, serve as joint antecedents of consumer purchase intention toward the POI-retailer. Our proposed relational framework can thereby provide essential insights into the likelihood of building a relationship with POI-retailers, despite the consumer's initial perceived uncertainty toward the retailer.

Accordingly, we need to clarify that we do not study an ongoing/actual relationship between the consumer and the POI-retailer. This is because our focus is on understanding the effect of uncertainty (arising from the POI-retailer's unfamiliarity) in the initial interaction, before any established relationship. In this early relationship phase, uncertainty can play an essential role in shaping consumer trust and attitude, and thereby influencing the development of commitment and purchase intentions. We specifically develop a relational framework to study purchase intentions under conditions where uncertainty is initially highly present. While previous research has focused on already established consumer–retailer relationships (Mukherjee and Nath, 2007; Cui et al., 2020; Lin et al., 2023), we contrast these studies and thus take a step back. In other words, we would rather shed light on the drivers of purchase intentions, which is the first step in creating prerequisites to build relationships. Therefore, we explore purchase intention by shedding light on how trust, commitment and attitude impact consumer purchasing development under conditions of uncertainty.

Our conceptual rationale is that during the earliest phase of the relationship with a POI-retailer, the consumer is unfamiliar with the retailer and its products, which influences their uncertainty toward the retailer. Consumer uncertainty negatively impacts trust development and consumer attitude. We argue that consumer attitudes are expected to play a crucial role in this early phase of the potential relationship. Specifically, attitude may pave the way for commitment building and purchase intention formation in the initial phase of the relationship. We suggest that attitude functions as an evaluative judgment that enhances consumers' relational behavior when relational cues are limited or absent. Using the concept of attitude helps clarify how consumers shape their perceptions and become willing to engage under conditions of high uncertainty, particularly when there is no prior experience or relational history. This approach extends relational exchange theory to encompass the pre-relationship phase, offering a deeper understanding of how relational exchange evolves in high-uncertainty environments, such as the consumer–POI-retailer setting.

In CBEC, uncertainty is a multidimensional concept that stems from both product uncertainty and retailer uncertainty, forming the consumer's overall uncertainty (Lee and Xiong, 2026). Product uncertainty relates to the discrepancies between the product received and what was promised (Lu and Chen, 2021). Consumers perceive product uncertainty because they cannot touch the product, which makes the product evaluation complicated (Sun et al., 2022). In contrast, retailer uncertainty involves concerns such as fraud, making false promises or hiding the retailer's actual characteristics (Sun et al., 2022). Consumers perceive retailer uncertainty when they are unfamiliar with and lack experience with the retailer, making it challenging for them to evaluate the retailer's quality (Pavlou et al., 2007; Safari and Yamin, 2016). Accordingly, several studies have documented the hampering effect of uncertainty. To begin, previous research has shown that uncertainty negatively affects trust (Pavlou et al., 2007; Safari and Thilenius, 2013; Lu and Chen, 2021). Consumer uncertainty creates explicit doubts about the retailer, which more likely diminishes their trust in the retailer (Safari and Thilenius, 2013). Following this, past research has found that uncertainty negatively affects attitudes (Quintal et al., 2010; Sun et al., 2022). However, a closer look at these studies reveals that they study the impact of product uncertainty on consumer attitudes, and thus, overlook the multidimensionality of uncertainty. Although they contribute to the notion that product uncertainty negatively affects attitude, they ignore the role of retailer uncertainty, an essential dimension of consumers' overall uncertainty that significantly influences how consumers perceive the retailer (Pavlou et al., 2007). Consequently, this provides an incomplete understanding of how overall uncertainty, comprising both product and retailer uncertainty, influences consumer attitude.

Nevertheless, we propose that overall uncertainty will impact consumers' attitudes toward the retailer. As shown in CBEC research, consumers are concerned not only with what they are purchasing (i.e. the product) online but also from whom they are purchasing (i.e. the retailer) (Lee and Xiong, 2026). According to previous research, if consumers are unable to reduce their uncertainty, it leads to doubt and hesitation, which hampers their ability to evaluate with confidence (Pavlou et al., 2007) and causes them to perceive the purchase more negatively (Zerbini et al., 2022). This can consequently hinder their formation of favorable attitudes toward the retailer. This is especially anticipated to be prominent in the context of purchasing from POI-retailers because consumers are unfamiliar with the retailer. This lack of experience complicates consumers' judgment of the POI-retailer and its products, which is consequently expected to negatively affect their attitudes. Building on previous research (Pavlou et al., 2007; Quintal et al., 2010; Safari and Thilenius, 2013; Sun et al., 2022), we hypothesize that uncertainty regarding the POI-retailer is expected to negatively affect both trust and attitude. Hence, the following two hypotheses are proposed:

Trust in a retailer is related to consumers' perceptions that the retailer acts with integrity, benevolence and competence (Pavlou et al., 2007; Oliveira et al., 2017; Tam et al., 2020). Building trust is essential because it assures consumers that the retailer will have their best interests in mind and will not engage in opportunistic behavior (Mukherjee and Nath, 2007; Cui et al., 2020). Research suggests that consumer trust shapes feelings toward a retailer, and these feelings form a favorable or less favorable attitude toward the retailer (Al-Debei et al., 2015; Akroush and Al-Debei, 2015; Li et al., 2024). Consumers who perceive that the retailer has their best interest in mind are more likely to perceive it as a good idea and a positive step to purchase from the retailer, thereby forming a favorable attitude (Oliveira et al., 2017; Yan et al., 2023). Accordingly, previous studies have found a significant effect of consumer trust on attitude in various purchasing contexts (Al-Debei et al., 2015; Akroush and Al-Debei, 2015; Moriuchi, 2021; Yan et al., 2023; Li et al., 2024). In the context of CBEC, some studies have highlighted the importance of developing trust and fostering positive attitudes to enhance consumers' purchase intentions toward retailers (Singh et al., 2006; Cui et al., 2020). However, no previous studies have empirically verified this effect in the context of purchasing from POI-retailers. Trust in the POI-retailer is essential as it fosters a more positive attitude toward the retailer. When trust is formed, consumers are more likely to form a favorable attitude, even in the face of unfamiliarity with the retailer. Thus, building on recent studies (Moriuchi, 2021; Yan et al., 2023; Li et al., 2024), we propose that trust will have a positive effect on consumer attitude. Hence, the third hypothesis:

Commitment is one of the core concepts of relational exchange theory (Morgan and Hunt, 1994; Safari and Albaum, 2019). Consumer commitment contributes to a long-term relational exchange with a retailer and is essential for initiating relationships between buyers and sellers (Lin et al., 2023). Previous studies have emphasized the importance of trust in affecting consumer commitment in online consumer–retailer relationships (Eastlick et al., 2006; Mukherjee and Nath, 2007; Lin et al., 2023). Other recent studies discuss the importance of favorable attitudes toward a retailer for developing and maintaining a long-term relationship (Kim and Kim, 2021; Alvarez et al., 2023). While these studies have empirically verified the positive impact of trust on commitment, the evidence remains dispersed when it comes to understanding attitude's role in shaping consumer commitment. In line with Zerbini et al. (2022), we argue that attitude should positively impact commitment because consumers tend to perform a specific behavior when they perceive positive feelings. This is expected because when consumers have positive perceptions of a retailer, they are more likely to explore the retailer further, leading to a more substantial commitment toward the retailer (Palvia, 2009; Li et al., 2024). Building on previous studies (Safari and Albaum, 2019; Zerbini et al., 2022; Lin et al., 2023), we propose that attitude and trust are essential for driving consumers' commitment toward the retailer. Thus, it is expected that trust and a favorable attitude toward the POI-retailer will positively impact consumer commitment toward the POI-retailer. Hence, the following hypotheses:

In their meta-analysis, Zerbini et al. (2022) show that consumer purchase intention can anticipate consumer purchasing behavior. Developing purchase intention is also essential for developing a long-term relationship (Cui et al., 2020). Numerous studies have empirically shown that trust and commitment are strongly determining factors in consumers' purchase intentions (Eastlick et al., 2006; Wang et al., 2016; Cui et al., 2020; Lin et al., 2023). Additionally, studies also show the significant role of attitude on consumer purchase intention (Alcantára-Pilar et al., 2017; Cabeza-Ramírez et al., 2022; Sun et al., 2022; Patel et al., 2023; Pop et al., 2023; Boukis et al., 2024). Although attitude has often been previously studied with consumer trust (Al-Debei et al., 2015; Yan et al., 2023; Li et al., 2024) and trust in connection to commitment (Wang et al.., 2016; Cui et al., 2020; Lin et al., 2023), it is also essential to examine all three concepts and their influence on consumer purchase intention. This is especially essential in the context of POI-retailers since consumers face uncertainty arising from unfamiliarity (Wistedt, 2024). Building on previous studies (Cui et al., 2020; Cabeza-Ramírez et al., 2022; Lin et al., 2023), it is expected that trust, commitment and attitude will have a positive impact on consumer purchase intention toward the POI-retailer. Hence, the final hypotheses:

This study employed a survey approach to examine the proposed research framework and its hypotheses by exposing participants to a POI-retailer. Before the data collection, it was important to ensure that participants of this study answered the survey questions in the context of a POI-retailer. Therefore, we established the research context by pre-selecting a POI-retailer beforehand, which the participants were exposed to during the survey. This step was crucial because the review of the CBEC literature revealed that consumers can purchase from three types of retailers in CBEC (AOI-, DOI- and POI-retailers). As such, allowing respondents to freely select any retailer would have made it difficult to ensure that purchase intentions toward a POI-retailer were being examined, compromising this study's validity. To address this, we used specific criteria to carefully pre-select a retailer beforehand that utilizes a POI approach to reach consumers in a specific foreign country. Thus, the following criteria were used:

1. The retailer is required to have a country-specific website with country-specific language, domain name and currency to target consumers in a specific foreign country, as these are typical characteristics of a POI-retailer (Schu and Morschett, 2017; Swoboda and Sinning, 2022).

2. The retailer is required to be physically located in a different country (without a warehouse in the host country) to ensure that products from the retailer to the consumer are shipped across national borders, verifying that we study consumer purchasing in a CBEC setting, rather than domestic e-commerce (Mou et al., 2020b).

3. The retailer must be a smaller, resource-constrained retailer (i.e. small and medium-sized enterprise) and thus not benefit from having a well-known brand name. This criterion was established to create a context in which respondents need to cope with uncertainty stemming from the retailer's unfamiliarity (Darke et al., 2016; Sharma and Klein, 2024), as this is a common characteristic of a POI-retailer (Safari and Yamin, 2016; Wistedt, 2024).

Based on the above criteria, we carefully selected a retailer in Germany that uses a POI approach to target consumers in Sweden. The retailer sells products related to fabrics, yarn and creative accessories through its website. We selected this retailer because it represents the typical characteristics of a POI-retailer and thus fulfills all the criteria above. More specifically, they have developed a Swedish website based on the Swedish language, domain name and currency (Criterion I); the retailer is in Germany, and the products are shipped to consumers in Sweden (Criterion II), and is classified as an SME with no established well-known brand name, making it unfamiliar to the participants (Criterion III). We followed the European Commission's (2005) definition of an SME, which includes firms with less than 250 employees and an annual turnover of less than €50m. To ensure the retailer was unfamiliar to the study's respondents, we asked them to confirm that it was their first time encountering the retailer. Based on this, the pre-selected POI-retailer (and its country-specific website) was the research object during data collection.

The next step entailed recruiting respondents for this study after pre-selecting a POI-retailer to obtain the research context. Following Lin et al. (2023), the respondents were recruited from Internet channels because those who use the Internet have a higher likelihood of having experience with purchasing online. This approach was particularly relevant to the purpose of this study, which was to investigate consumer purchase intention toward a POI-retailer in CBEC. The data was collected from Swedish-speaking consumers in Sweden because they have a high degree of online purchasing experience (Statista, 2023) and were of relevance since the German POI-retailer had developed a Swedish website. Interested respondents were informed that their participation in the study was for academic purposes. Anonymity was ensured by disregarding any respondent's identification information. Moreover, participation was voluntary, and they could drop out at any time during the procedure. Once the respondents consented to participate in the study, they were instructed to follow two steps.

The first step entailed making sure that respondents had no previous experience and familiarity with the pre-selected POI-retailer and its country-specific website. This was done by having respondents confirm that it was their first time encountering the pre-selected POI-retailer. Once this was ensured, they were further instructed to visit our pre-selected POI-retailer's website. They were encouraged to see any part of the website they would like, allowing respondents to interact with the website in a way that felt authentic and comfortable for them. This open-ended exploration was designed to mimic a real-world purchasing experience, enabling respondents to form opinions and intentions about the retailer. This approach has been shown to generate more accurate and meaningful data when studying consumer purchase intention in CBEC (e.g. see Shao et al., 2021; Wang et al., 2023). After they had interacted with the POI-retailer and its website, in the second step, we distributed this study's questionnaire, which included our items aimed at measuring the constructs in this study (see operationalization in Table 3). Thus, respondents were assigned to answer the questionnaire based on their perceptions and experiences of visiting the POI-retailer's country-specific website. To ensure the survey accurately reflected the study's research context, it was translated into Swedish. By providing the questionnaire in Swedish for the respondents, we aimed to minimize language barriers that could affect the clarity of the respondents' answers and ensure that data collection was reliable and valid.

We applied two methods to determine the minimum sample size required for this study: (1) a 10-times rule method and (2) a G*power test. First, we applied the 10-times rule method by following the guidelines of Shmueli et al. (2019) and Sarstedt et al. (2021). The minimum sample size was calculated as 270 respondents, multiplying the total number of measurement items (27) by 10 (Sarstedt et al., 2021). Second, we also conducted a G*power test based on the recommendations by Faul et al. (2009). This method has been extensively adopted in CBEC studies (e.g. Lee and Xiong, 2026). Thus, we configured the software parameters using an effect size of 0.15 (the average value), a power level of 0.95 and a maximum allowed error of 0.05. The results of the G*power test indicated that a minimum of 119 samples were required for this study. Considering the above guidelines, the data was collected in Sweden in 2022 with a sample size of 385 consumers. Hence, this exceeds the minimum requirement of the two methods, ensuring a robust sample for the study (Faul et al., 2009; Shmueli et al., 2019; Sarstedt et al., 2021; Lee and Xiong, 2026). We removed 21 respondents because they lacked online purchasing experience. This is because if they generally lack experience with e-commerce, they are not suitable for a CBEC survey. After all, they are usually hesitant to purchase online (Hazarika and Mousavi, 2021). As shown in Table 2, the final sample consists of 364 respondents, including 209 females and 155 males, representing different ages, online purchasing frequencies and educational backgrounds.

In this paper, all measurements of the constructs were retrieved from previously validated measures and were assessed in alignment with their definitions (see Table 3). After modifying the items for the research context of this study, we adopted a seven-point Likert scale (1 = strongly disagree and 7 = strongly agree) to measure the constructs in this study. Specifically, five items from Dodds et al. (1991) were retrieved to measure consumer purchase intention. Commitment was measured with six items adopted from Eastlick et al. (2006), while attitude was measured based on five items retrieved from Bhattacherjee and Premkumar (2004). Seven items for trust and five for uncertainty were retrieved from Pavlou et al. (2007). Moreover, this study included several control variables (i.e. gender, age, education and online purchasing frequency) because previous studies have shown the need to control these when studying consumer purchase intention in CBEC (e.g. see Cui et al., 2020; Mou et al., 2020a).

The measurement model for this study was analyzed using IBM SPSS 29.0 and IBM AMOS 26.0. We initially conducted an exploratory factor analysis for two reasons: (1) to detect cross-loadings, and (2) to ensure that all items are loaded onto their expected factor. Five factors were extracted based on the principal extraction technique and the varimax method, with all items loading onto their expected factor. After confirming that all loadings were aligned with their respective constructs, we also measured the sampling adequacy based on Kaiser–Meyer–Olkin (KMO) to assess how suitable the data were for factor analysis by using the cut-off of 0.5 (Sallis et al., 2021). The value of KMO showed 0.943, hence indicating that the data is highly suitable for confirmatory factor analysis (CFA). Therefore, we further conducted a CFA to assess the convergent validity of the constructs. This involved evaluating the factor loadings and average variance extracted (AVE) values. As shown in Table 4, all factor loadings exceed the threshold of 0.6 (Tabachnick et al., 2013) and the cut-off of 0.6 for AVE (Fornell and Larcker, 1981), confirming the convergent validity of the constructs. The discriminant validity was also examined by comparing the square root of AVE and the inter-construct correlations (Fornell and Larcker, 1981). Accordingly, results show that the square root of AVE was of greater value than inter-construct correlations (see Table 5), confirming the presence of discriminant validity of the constructs (Fornell and Larcker, 1981).

Next, we examined the reliability of each construct by considering Cronbach's alpha (α) and composite reliability (CR). We followed the cut-offs of 0.7 for CR (Fornell and Larcker, 1981) and 0.8 for Cronbach's alpha (Cortina, 1993). As shown in Table 4, all measurements were evaluated as meeting these thresholds: Uncertainty (α = 0.953; CR = 0.948), Trust (α = 0.979; CR = 0.977), Attitude (α = 0.971; CR = 0.969), Commitment (α = 0.929; CR = 0.922) and Purchase intention (α = 0.922; CR = 0.916), thus conforming construct reliability.

Finally, we assessed the model fit for the measurement and structural models (Table 6). All indices of the model fit assessment meet the recommended thresholds (Ullman and Bentler, 2012). Goodness-of-fit (GFI) was slightly below the recommended threshold, yet this is not a concern if the overall model fit indices meet the thresholds (MacCallum et al., 1996). Therefore, the model fit results show that the data is a good fit with the measurement and structural models. Finally, we checked for normal distribution by following the established thresholds from past research (Hair et al., 2010; Byrne, 2016), where skewness values within ±2 and kurtosis values within ±7 are considered acceptable for assessing normality. As shown in Table 4, all values meet the specified thresholds, indicating that non-normality is not a major concern in this study.

In this study, we checked for multicollinearity, common method variance (CMV) and robustness. Consistent with past studies (Dhir et al., 2024), we tested for multicollinearity by checking the Variance inflation factor (VIF) and Tolerance values. As shown in Table 4, the VIF values were less than 10, and all tolerance values exceeded 0.10 thresholds, which confirms the absence of multicollinearity between our variables (Hair et al., 2010). Further, we considered ex ante and post hoc methods to decrease the influence of CMV. For the ex ante method, we followed the recommendations by MacKenzie and Podsakoff (2012). This involved considering the questionnaire's structure, voluntary participation and encouraging respondents to read each unique question carefully. For the post hoc test, we adopted statistical tests to evaluate CMV. To begin, we assessed the correlation between the constructs. As shown in Table 5, the correlation matrix does not indicate any highly correlated constructs, whereby all are below the threshold of 0.9 (Pavlou et al., 2007; Mou et al., 2020b). Following prior studies (Khan, 2023; Mandler et al., 2023), we further tested for CMV by using the marker variable (MV) technique proposed by Lindell and Whitney (2001). We used a theoretically unrelated MV (i.e. “I feel very experienced with buying on the internet”) as a proxy, which was measured on a seven-point Likert scale (1 = strongly disagree and 7 = strongly agree). In doing so, we used the smallest positive correlation (i.e. r = 0.06, p = 0.255) between the MV and other variables to adjust the inter-construct correlations. The results showed that all correlations that were significant prior to the effect of the MV remained significant. Based on these evaluations, we can conclude that CMV is not a major concern in this study.

Further, there are several ways to assess the robustness of a hypothesized model, and the choice of method is typically guided by the normal distribution of the data (Hair et al., 2010). As the normality test confirmed that our data followed a normal distribution (Table 4), this guided the selection of the most appropriate robustness test, namely an analysis of potential nonlinearity (Little et al., 2006). Following the suggestions by Little et al. (2006), we tested quadratic effects post hoc on our hypothesized model to assess its robustness. In doing so, we first created quadratics on all independent variables (Uncertainty, Trust, Attitude and Commitment) by squaring the variable (i.e. Uncertainty², Trust², Attitude² and Commitment²). Then, we used regression analysis in SPSS and added these square variables as independent variables on our dependent variable (Purchase intention) to create residualized quadratic variables to control for multicollinearity. We finally used AMOS and added these residualized quadratic variables to our hypothesized model (Figure 1) by directing all paths from residualized quadratic variables to their respective outcome variables aligned with our hypothesized model. For instance, in our hypothesized model, Uncertainty is directed to affect Trust. Hence, we additionally added Uncertainty² to also affect Trust in order to test the potential nonlinearity in Uncertainty affecting Trust. As a result, all paths of the residualized quadratic variables were found to be insignificant (Table 7), and the model fit of the structural model remained superior compared to the structural model with the quadratics (Table 6). Therefore, we can conclude that our structural hypothesized model is robust (Little et al., 2006).

This study used structural equation modeling in IBM AMOS 26.0 to analyze the structural model and to test the hypotheses. The results of the path coefficients, significant levels (p-value), determination coefficients (R² values), and control variables are presented in Figure 2 and Table 8. The results showed that five out of the eight hypotheses were supported. Uncertainty exhibited a significant negative impact on trust (β = −0.413, p < 0.000) but did not have a significant effect on attitude (β = −0.052, p = 0.178). This indicates that H1 is supported, whereas H2 is not. The effect of trust on attitude was further examined, showing a significant positive impact (β = 0.685, p < 0.000), thus confirming H3 ^[1]. Concerning the effect of trust and attitude on commitment, results reveal that trust did not have a significant impact on commitment (β = 0.028, p = 0.523), while attitude was significantly impacting commitment (β = 0.547, p < 0.000). In other words, while H5 is supported, H4 is not. The results further reveal that trust had an insignificant impact on purchase intention (β = 0.049, p = 0.367). On the contrary, the influence of commitment (β = 0.629, p < 0.000) and attitude (β = 0.145, p = 0.046) showed a positive and significant impact on purchase intention. That is, while H6 is not supported, H7 and H8 are confirmed. Lastly, the proposed model revealed relatively high explanatory power on the endogenous variables. Specifically, the R² values of trust, attitude, commitment and purchase intention yielded 22%, 52%, 55% and 49%, respectively, exhibiting a significant explanatory power (Garson, 2016).

Based on the unexpected results in the hypothesis testing with the insignificant impact of trust on commitment (H4) and purchase intention (H6), we conducted a mediation analysis to better understand the role of trust in influencing these variables. While some argue that significant relationships are required to conduct a mediation analysis (Baron and Kenny, 1986), this study follows the approach by Preacher and Hayes (2008), who argue for the theoretical relevance of investigating mediating effects. Studies have theoretically and empirically supported the mediating effect of commitment (Lin et al., 2023) and attitude (Zerbini et al., 2022). Thus, we considered them as mediators to investigate whether trust has an indirect effect on commitment and purchase intention (Table 9). Following Preacher and Hayes (2008), we adopted a bootstrapping technique in AMOS with 5,000 bootstraps to analyze the specific indirect effects. To test if commitment and attitude could mediate between trust and purchase intention, we first analyzed their single role as mediators and then their joint mediating effect.

Considering the single mediating effect, the results show that the impact of trust on purchase intention is not mediated by commitment (β = 0.018, p = 0.562), whereas the single mediating effect of attitude is significant (β = 0.099, p = 0.046). We further tested the joint mediating effect of attitude and commitment, and the results show that trust has a greater indirect effect on purchase intention when it is jointly mediated by both attitude and commitment (β = 0.236, p < 0.000). Finally, given the lack of direct impact of trust on commitment, the mediation analysis shows that the effect of trust on commitment is significantly mediated by attitude (β = 0.375, p < 0.000).

CBEC has received momentum in recent years, with the critical need to understand consumer purchase intention (Mou et al., 2020a; Do et al., 2023; Wang et al., 2023; Lee and Xiong, 2026). However, much of the previous research has primarily focused on consumer purchase intention in the context of DOI-retailers (e.g. see Safari, 2012; Safari et al., 2013; Huang and Chang, 2019; Ma et al., 2022; Wagner et al., 2024) and AOI-retailers (e.g. see Singh et al., 2006; Luo and Ye, 2019; Mou et al., 2020a; Xu et al., 2023; Lee and Xiong, 2026), leaving a noticeable gap in understanding how consumers develop purchase intentions in the context of POI-retailers. Understanding purchase intention toward these smaller, resource-constrained retailers is crucial, as their disadvantage of being unfamiliar creates uncertainty, which impedes consumers' purchasing (Darke et al., 2016; Sharma and Klein, 2024) and, most significantly, hinders the building of consumer–retailer relationships (Cui et al., 2020; Cheng et al., 2022; Lin et al., 2023). Therefore, this study aimed to explore consumer purchase intention in the context of POI-retailers from a relational view. We introduced and thus tested a proposed relational framework built on the ideas from the relational exchange and consumer attitude literature, a combination previously overlooked. We argued that uncertainty stemming from the retailer's unfamiliarity may decrease the impact of relational antecedents (i.e. trust and commitment) on consumers' purchase intentions, but a favorable attitude can offset this by strengthening commitment and purchase intentions. While more than half of the proposed hypotheses were supported by the results, our findings revealed some interesting insights and, thus, unexpected results.

To begin with, our proposed relational model demonstrates a strong predictive power for explaining the endogenous variables (see R² in Figure 2). Specifically, it explains our dependent variable (i.e. purchase intention) with nearly fifty percent, which substantially contributes to understanding the drivers of purchase intention toward POI-retailers (Garson, 2016). This also represents greater predictive power compared to previous CBEC research examining purchase intention (e.g. see Singh et al., 2006; Han and Kim, 2019; Hsu et al., 2022; Wagner et al., 2024; Lee and Xiong, 2026). Further, our study reveals that consumer uncertainty has only a negative effect on trust and thus does not significantly impact consumer attitude. On the one hand, our findings are consistent with previous studies emphasizing the effect of uncertainty on trust (Pavlou et al., 2007; Safari and Thilenius, 2013; Lu and Chen, 2021), but on the other hand, we differ from studies proposing the role of uncertainty on attitude (Quintal et al., 2010; Sun et al., 2022). Specifically, this study's results suggest that while uncertainty stemming from the POI-retailer's unfamiliarity substantially decreases trust in the retailer, it does not necessarily cause consumers to develop a negative attitude toward the POI-retailer. Therefore, once consumers are uncertain about the retailer's products and the retailer itself, their trust is hampered, not their attitude. In other words, our findings align with previous studies that report a negative effect of uncertainty on trust (e.g. Safari et al., 2013) and add to this by demonstrating that uncertainty does not necessarily impact consumer attitude. Following this, we additionally considered studying the impact of consumer trust on their attitude. As suggested by previous research (Al-Debei et al., 2015; Akroush and Al-Debei, 2015; Moriuchi, 2021; Yan et al., 2023; Li et al., 2024), we also found a significant positive effect of trust on attitude. This suggests that when consumers encounter a POI-retailer, their trust plays a significant role in shaping consumers' attitudes, determining whether consumers will perceive the retailer with a favorable or less favorable attitude.

We further examined the hypothesized effects of trust and attitude on consumer commitment. Interestingly, findings reveal that consumer attitude only affects commitment, with trust having no significant impact. Commitment and trust are the two central concepts of relational exchange theory (Morgan and Hunt, 1994; Safari and Albaum, 2019), whereby extensive research has repeatedly shown evidence for the significant impact of trust on commitment (Eastlick et al., 2006; Mukherjee and Nath, 2007; Lin et al., 2023). However, our results contrast with these studies and thus differ from what relational exchange theory suggests. Specifically, our findings yield surprising insights, showing that consumer commitment is only driven by consumer attitudes toward the retailer, while trust is not a key determinant. In the context of purchasing from POI-retailers, this unexpected finding may indicate that once the consumer wants to initiate their commitment toward POI-retailer, their attitudes primarily determine it. Most studies in CBEC assume that trust increases commitment (Safari et al., 2013; Cui et al., 2020), yet our findings indicate that consumer commitment is shaped more by the extent to which consumers perceive the POI-retailer with a favorable attitude, instead of trust in the POI-retailer. As illustrated in the mediation analysis (Table 9), trust indirectly affects commitment through attitude. Hence, one explanation for the lack of direct impact of trust on commitment relates to the role of consumer attitudes. Although consumers trust the POI-retailer, this trust only matters if it shapes their attitude (favorable or unfavorable). In other words, once consumers form their attitudes toward the POI-retailer, it is this attitude that directly increases their commitment, and hence not necessarily trust itself. Therefore, trust might influence commitment, but only through its impact on the consumer's attitude. This intriguing finding extends previous research in CBEC (Wang et al., 2016; Cui et al., 2020) as it sheds light on the importance of consumer attitude as a significant mediator between trust and commitment.

Moreover, as we focused on consumer purchasing in the context of POI-retailers, our dependent variable was consumer purchase intention. As suggested by previous research, we hypothesized a positive impact of trust, commitment and attitude on consumer purchase intention. However, in a similar vein to the above, findings yielded unexpected results. While most studies in CBEC assume that trust increases purchase intentions (Safari et al., 2013; Huang and Chang, 2019; Cui et al., 2020), our findings challenge this assumption. Particularly, our findings indicate that in the context of unfamiliar POI-retailers, consumer attitude can serve as a greater predictor for purchase intention, compared to trust. In other words, we found that attitude and commitment only affected purchase intention, with trust being an insignificant driver. This suggests that trust in the POI-retailer does not necessarily motivate their willingness to buy from the retailer, which could be explained by the strong negative effect of uncertainty (Pavlou et al., 2007; Darke et al., 2016; Mou et al., 2020a; Sharma and Klein, 2024). Instead, if consumers feel positive about the POI-retailers, thus a favorable attitude, then they are more inclined to follow through with a purchase. Similarly, consumers will solely purchase from the POI-retailer if they want to invest and thus commit to the retailer. Together, attitude and commitment create prerequisites for driving consumers' motivations to conduct a purchase from the POI-retailer. Therefore, while the impact of attitude and commitment on purchase intention supports past studies (Alcantára-Pilar et al., 2017; Cui et al., 2020; Cabeza-Ramírez et al., 2022; Patel et al., 2023; Pop et al., 2023; Boukis et al., 2024), the lack of significant impact of trust on purchase intention contrasts with previous research (Eastlick et al., 2006; Wang et al., 2016; Cui et al., 2020; Lu and Chen, 2021; Lin et al., 2023).

Lastly, our findings differ from previous studies that reported a significant impact of trust on both commitment and purchase intention. This divergence is particularly interesting in the context of POI-retailers, as it reveals that trust may not be a direct driver of commitment and consumer purchase intention, but rather operates through a different pathway. Regarding the effect of trust on commitment, our findings highlight that consumers' attitudes play a central mediating role between the two. As such, our results indicate that trust has an indirect role and hence not a direct role as proposed in previous CBEC studies (e.g. Huang and Chang, 2019; Cui et al., 2020) and others drawing on relational exchange theory (Lu and Chen, 2021; Lin et al., 2023). Regarding the effect of trust on purchase intention, trust has a direct influence on purchase intention when it is jointly mediated by attitude and commitment (see Table 9). As such, our model thus refines and presents an alternative understanding of the relational mechanisms under conditions of uncertainty by suggesting that trust functions as an indirect role for purchase intention through the joint mediating effect of attitude and commitment. This finding challenges the conventional notion of relational exchange theory in CBEC by showing that trust alone is insufficient to drive consumer purchase intention in the context of POI-retailers. Instead, it needs to be transformed into favorable attitudes before it can translate into commitment and purchase intention. Accordingly, our study not only extends relational exchange theory to a new context but also provides an alternative path on how trust operates when consumers face heightened uncertainty.

To clarify, our study does not suggest that trust is irrelevant. Instead, we propose that, given the presence of uncertainty, consumers may face difficulties in developing trust. Therefore, attitude may serve as an alternative path for increasing commitment and purchase intentions. Theoretically, consumer attitude can, at first hand, serve as a temporary substitute for trust in situations where uncertainty strongly impacts trust, thereby strengthening commitment and purchase intentions. This is especially relevant in the initial phase of the relationship when consumers need to cope with uncertainty arising from the POI-retailer's unfamiliarity. Intuitively, as consumers engage more with the retailer, trust becomes increasingly essential and may have a direct impact on the commitments made to maintain a long-term relationship. We interpret this not as a contradiction of relational exchange theory itself, but as a boundary condition. Under increased uncertainty and lack of familiarity, trust may have a secondary or conditional role, especially in the early phases of the relationship. We believe that trust will ultimately be vital for maintaining or even strengthening the relationship.

This study focused on consumer purchase intention in the context of POI-retailers in CBEC from a relational view. Thus, our research has the following theoretical implications. First, it sheds a new light on consumer purchase intention in CBEC. We particularly enrich the understanding of how consumers develop purchase intentions toward POI-retailers. Extensive research has contributed to the body of knowledge on consumer purchase intention in CBEC, yet it has predominantly focused on purchase intention in the context of DOI-retailers (e.g. Huang and Chang, 2019; Ma et al., 2022; Wagner et al., 2024) and AOI-retailers (e.g. Mou et al., 2020a; Xu et al., 2023; Lee and Xiong, 2026). This focus has neglected the third type of retailer consumers can purchase from in CBEC, namely POI-retailers. Although few have shed light on consumer purchasing toward POI-retailers, they have primarily focused on the POI-retailer's online internationalization journey (e.g. Daryanto et al., 2013; Schu and Morschett, 2017; Swoboda and Sinning, 2022), been conceptual (Safari and Yamin, 2016), or focused on technological antecedents (Wistedt, 2024). We fill this research gap and establish a relational conceptual framework for understanding consumer purchase intention toward POI-retailers from a relational view. Accordingly, this study is among the few that explore purchase intention from POI-retailers, offering pivotal insights into how consumers' purchase intentions unfold in this novel purchasing setting.

Second, this study builds on the ideas from the relational exchange theory (Pavlou et al., 2007; Eastlick et al., 2006; Safari and Albaum, 2019; Lin et al., 2023) and the consumer attitude literature (Patel et al., 2023; Pop et al., 2023; Boukis et al., 2024). In doing so, we establish a connection between its theoretical concepts by introducing them into a unified framework, which has, thus far, been overlooked in scholarly research. Hence, we contribute to developing a relational framework for studying consumer purchase intention. This is not the first time purchase intention has been conceptualized based on the ideas from relational exchange theory (e.g. Pavlou et al., 2007; Eastlick et al., 2006; Mukherjee and Nath, 2007; Palvia, 2009; Cui et al., 2020; Lin et al., 2023). Yet, these studies focus on the effects of relational antecedents (i.e. trust and commitment) on consumer purchasing in ongoing relationships with a retailer, where uncertainty is already reduced. In this light, it neglects the notion that consumers need to cope with uncertainty before developing any relationship and, more significantly, overlooks the understanding of what prerequisites are important for relationship building. We build and extend this notion by studying the effects of relational antecedents and extending its scope with consumer attitude literature (Patel et al., 2023; Pop et al., 2023; Boukis et al., 2024). Therefore, the proposed relational framework in this study is an essential contribution, as it provides a foundation for expanding knowledge on consumer purchase intentions from a relational view, helping guide future CBEC research.

Lastly, our proposed relational framework sheds new light on how retailers, especially unfamiliar retailers such as POI-retailers, can create prerequisites to build relationships with consumers despite consumer uncertainty. While past research shows that consumer uncertainty, stemming from the retailer's unfamiliarity, hampers trust and thereby has consequences for relationship building (Darke et al., 2016; Sharma and Klein, 2024), our study offers essential insights into the prerequisites for relationship building. Given that we did not examine an actual/ongoing consumer–retailer relationship, this study shows that retailers, particularly POI-retailers, can still create prerequisites for building relationships despite the initial challenges in fostering trust. They can achieve this by making consumers follow an alternative path to commitment and purchase intention, even when consumers face uncertainty arising from the retailer's unfamiliarity. This is achieved by nurturing a favorable consumer attitude because our results show that it serves as a crucial catalyst for increasing commitment and purchase intentions. Moreover, our results also indicate that attitude could act as a mediator of trust's impact on commitment and purchase intention. Therefore, the relational view developed for this study is a significant contribution since it can be utilized to advance further knowledge about consumer purchase intentions toward POI-retailers, especially CBEC in general. Furthermore, our findings contribute to relational exchange theory by identifying vital boundary conditions under which its core assumptions do not hold. Although trust is widely regarded as the central mechanism driving commitment and purchase intention in relationships, our results show that this logic weakens in high-uncertainty, unfamiliar POI-retailer contexts. Instead, consumer attitude and commitment emerge as the primary drivers of purchase intention, suggesting that relational exchange mechanisms operate differently when consumers lack prior experience and familiarity with the retailer. These findings imply that relational exchange theory should be applied with greater sensitivity to contextual uncertainty, particularly in early-phase CBEC purchasing.

Based on the empirical findings, this study offers significant insights into consumer purchase intentions for managers. Specifically, it has managerial implications, particularly for managers of firms that employ a POI strategy in CBEC. While this approach provides cost-effective opportunities to reach consumers in a specific country, managers should remember that consumers are initially unfamiliar with them. This unfamiliarity creates uncertainty for consumers, which may erode their trust since they are not familiar with the company's product or the company itself (i.e. the POI-retailer). Therefore, these companies and their managers must implement POI strategies effectively and invest resources in developing a website that reduces perceived consumer uncertainty while enhancing favorable attitudes. The website should appear professional and provide detailed product information, retail details and insights into the purchasing process to influence consumer commitment and positive attitudes, thereby increasing the likelihood of purchases from their website. The case company used in this study demonstrates that by utilizing a POI strategy, retailers can attract foreign consumers by creating a country-specific website, which boosts their purchase intentions. More specifically, since trust operates indirectly through attitudes, managers should focus on fostering positive attitudinal responses by using trust-enhancing cues. These include: (1) Third-party certifications such as Trustpilot and ISO standards to make up for the lack of familiarity; (2) Secure payment options and clear refund policies to lower perceived transaction uncertainties; (3) Social proof mechanisms like verified customer reviews, ratings and user testimonials to boost credibility; and (4) Localized content and culturally adapted messaging to bridge institutional and cultural gaps.

Furthermore, while the POI strategy enables an unfamiliar retailer to target and attract consumers from a specific country, it is also advisable that they engage in additional marketing activities to increase consumer awareness of their products, brand and country-specific website. Otherwise, they risk remaining unknown to consumers, and many individuals in their target market may not even be aware of their existence, which could lead them to mistakenly believe that simply developing the website is sufficient while neglecting other marketing efforts. Such marketing activities, including online marketing or partnering with local entities to disseminate information about their presence, can help alleviate or even significantly reduce consumers' perceived uncertainty toward them. However, fostering long-term relationships with consumers involves more than just creating a country-specific website or executing marketing activities; retailers must generate value for consumers. Retailers need to offer quality products, support and services to cultivate favorable attitudes, trust and consumer commitment, all of which play direct and indirect roles in enhancing consumer purchase intentions. Accordingly, this approach increases the likelihood of forming a successful consumer–retailer relationship.

Although this study is essential and has shed light on consumer purchase intention toward POI-retailers, it is not without limitations. This paper studied only one POI-retailer in one specific industry, specifically from Germany. It is, therefore, important that future research investigates other contexts and industries to confirm or challenge the findings of this study. While our study focuses on one pre-selected POI-retailer and one product category, future studies are encouraged to use other examples of POI-retailers. This includes retailers operating in various industries, offering different types of products and originating from different parts of the world. Such diversity would help to generalize the findings and further advance the understanding of consumer purchasing in the context of these retailers. Moreover, our study solely investigated consumers from Sweden. We recommend future research to study consumers from other countries to explore potential cross-cultural differences. It is vital to determine whether consumers are more willing to purchase from POI-retailers from a culturally close country than from their home country or distant country markets. Additionally, this study did not consider other factors that may influence consumer purchase intention or how previous CBEC experience influences trust and purchase intention. It is thus essential to point out that we did not ask about the respondents' product involvement or whether they have purchased online from Germany before. We encourage future research to assess these factors, as it could provide insights into the formation of consumers' purchase intentions.

Further, we built our relational framework on the past relational exchange theory and consumer attitude literature, and our model has strong predictive power in explaining consumer purchase intention toward POI-retailers (see R² in Figure 2). Yet, it is also essential to use our model to explain purchase intention in other purchasing contexts. We hope that our model also has predictive power in explaining actual purchasing and might also be used for developing future conceptual models in describing how consumers and POI-retailers initiate, develop and maintain a long-term relationship. Thus, we both welcome quantitative and qualitative approaches to study relationship development. For example, conducting a process study on how consumers and retailers initiate, develop and maintain a long-term relationship in CBEC would be wise. This could have strong theoretical implications for practitioners in their journey to developing a country-specific website and attracting consumers in CBEC. Finally, although our findings indicate a mediating role of attitude, it is important to note that the design is cross-sectional. Therefore, further attention is needed in future research to explore the role of attitude and trust using other research designs.