
==== Front
Heliyon
Heliyon
Heliyon
2405-8440
Elsevier

S2405-8440(24)13658-4
10.1016/j.heliyon.2024.e37627
e37627
Research Article
Antecedent research on cross-border E-commerce consumer purchase decision-making: The moderating role of platform-recommended advertisement characteristics
Xu Yiwei
He Di
Fan Mingyue fanmy@ujs.edu.cn
⁎
School of Management, Jiangsu University, Zhenjiang, 212013, China
⁎ Corresponding author. fanmy@ujs.edu.cn
07 9 2024
30 9 2024
07 9 2024
10 18 e3762724 11 2023
31 8 2024
6 9 2024
© 2024 Published by Elsevier Ltd.
2024

https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Relying on the development of big data, internet technology, and digital trade, consumers are seeking cross-border e-commerce platforms increasingly. Platform-based advertising communication, characterized by low cost and wide coverage, is notable as one of the marketing strategies favored most within general cross-border e-commerce platforms. It furnishes consumers with abundant shopping options while it exerts influence on their purchasing decisions concurrently. Using Behavioral Reasoning Theory to explore in-platform advertisements' effect on consumer purchasing decisions in cross-border e-commerce platforms, this study introduces advertising features as moderating variables. From the perspectives of self-efficacy, behavioral rationality, and inclusiveness, a behavioral reasoning model for consumer purchasing decisions is constructed. This paper focuses on cross-border e-commerce platform consumers as the research object, and uses Partial Least Squares Structural Equation Modelling (PLS-SEM) to analyze the data and validate the model. The results show that self-efficacy and behavioral rationality can have a direct effect on consumers' purchase decision-making behaviors, while self-efficacy plays a significant mediating role between behavioral reasonableness and purchase decision-making behaviors. Inclusiveness has a significant direct effect on self-efficacy and behavioral reasonableness; moreover, advertisement characteristics have a significant negative moderating effect on inclusiveness and behavioral reasonableness-adoption. This study revealed the influencing mechanisms of consumers' purchasing decision behavior on cross-border e-commerce platforms, and expands the scope of Behavioral Reasoning Theory's application. Further, it provides valuable insights for the development of cross-border e-commerce platforms that have both theoretical and practical significance.

Keywords

Cross-border e-commerce platform
Advertising recommendation behavior
Purchasing decision behavior
Behavioral reasoning theory
==== Body
pmc1 Introduction

As an open, free, inclusive, and universal trade platform, cross-border e-commerce platforms connect businesses and consumers from around the world closely, and have gained increasing popularity among a wide range of consumers [1]. To stimulate consumer purchase intentions and promote buying behavior, cross-border e-commerce platforms use big data and machine algorithms to distribute advertisements related to stores or products to consumers actively [2]. The platform's information flow advertisements have achieved personalization and precision in delivering ads, enhanced the reach of platform advertisements, and increased consumer conversion rates significantly [3]. Kim [4] argued that novel marketing models can connect with audiences more effectively, and thus elicit more positive behaviors and attitudes. Despite these emerging advertising formats' effectiveness in circumventing the drawbacks of traditional internet advertising, such marketing strategies still encounter certain issues that may influence consumers' purchasing behavior on cross-border e-commerce platforms profoundly. Consequently, the imperative for cross-border e-commerce platforms to enhance their advertising strategies to garner more consumers and achieve higher conversion rates is underscored.

Compared to the distribution of ordinary e-commerce platforms' advertising, cross-border e-commerce platforms have distinct characteristics related to both the audience and the types of advertisements. With respect to the advertising audience, ordinary e-commerce platforms target local consumers mainly, while cross-border e-commerce platforms have an audience spread across different regions globally that differ in language, culture, and customs [5]. This situation has led to differences between the two platforms' advertising types. When facing local consumers, ordinary platforms use bid-based advertisements primarily that highlight their products' affordable and cost-effective nature through multi-tagged titles and eye-catching promotional prices with the goal to motivate consumers to make purchases. However, when targeting diverse and cross-cultural audiences, cross-border e-commerce platforms rely primarily on product display advertisements [6]. They use images, videos, and other means to showcase products in contextualized promotion scenarios, and increase unique products' visibility in different countries and regions thereby. The audience's diversity implies that cross-border e-commerce platforms must adopt different types of advertisements to meet the needs of a global audience. However, this global characteristic exacerbates the occurrence of privacy security issues and instances of misleading advertising to some extent.

In fact, advertising recommendations' effect on consumer purchase decisions in cross-border e-commerce platforms is multifaceted. The pervasive bombardment of advertising within cross-border e-commerce platforms raises not only issues concerning consumer privacy and information security, but also concerns about the product information's quality and authenticity. On the one hand, personalized advertising based upon shopping habits, search history, and purchasing behavior poses risks to consumers' privacy and increases the likelihood of data breaches, which leads to decreased trust in cross-border e-commerce platforms [7]. This may lead consumers to reduce or avoid cross-border online shopping altogether. On the other hand, the presence of infringing goods and subpar products within the platform, which are not subjected to rigorous scrutiny, can be promoted to consumers through advertisements. This not only diminishes their shopping experience, but also affects legitimate businesses' revenue. Further, it generates negative word-of-mouth for the platform and reduces consumers' willingness to use it [8]. Clearly, the array of issues triggered by platform advertisements poses significant challenges to cross-border e-commerce platforms' sustainable development and consumers' experience. Therefore, investigating the mechanism of consumers' purchase decision-making from the perspective of platform ad recommendations holds substantial value.

Current research on consumer decision-making in cross-border e-commerce focuses primarily on different product types [9], marketing strategies [10,11], and the influencing factors or mechanisms of consumer purchase intentions from various theoretical perspectives [[12], [13], [14]]. Studies that have examined platform advertising recommendations specifically concentrate primarily on different types of advertisements [15,16], and various machine algorithms' effect [17,18] on consumers, businesses, or social development. However, little research has explored the relation between these two areas. Behavioral Reasoning Theory (BRT) is employed often to elucidate the antecedent factors that influence individual intentions and behaviors [19]. Models such as the Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology Model (UTAUT) focus primarily on external conditions' influence on consumers' behavior, while the BRT theory delves into consumers' psychological factors from the perspective of acceptance and rejection to evaluate their effect on behavior [20]. This theory has been applied to explore and validate the factors that influence household energy-saving intentions and behaviors. Based upon the perspective of BRT, Ashfaq [21] found that consumers' attitudes toward Ant Forest influence their intention to continue using it significantly, and promote offline green consumption behavior thereby. However, BRT has not been used fully in the field of cross-border e-commerce, and few scholars have explored consumers' behavior on cross-border e-commerce platforms based upon this theory as a precursor. Hence, this study conducts an antecedent exploration of consumers' behavior on cross-border e-commerce platforms based upon BRT, which is crucial to understand the mechanisms of consumers’ purchase decision-making and optimize advertising recommendations.

To provide a comprehensive understanding of the research topic, this study focuses on consumers' purchase decision behavior on cross-border e-commerce platforms. Based upon BRT and taking platform advertising reach as the research context, this study combines the existing literature to analyze specific behavioral reasons and reveal the antecedents and mechanisms of consumers' purchase decision behavior on cross-border e-commerce platforms in the context of advertising reach systematically. This research not only enriches the theoretical understanding of cross-border e-commerce consumer behavior, but also provides a rational explanation for the formation of consumers’ purchase decision behavior on cross-border e-commerce platforms. Moreover, it can serve as a practical reference to optimize cross-border e-commerce platforms.

2 Literature review

2.1 Consumer Purchase Decision Behavior

Consumer Purchase Decision Behavior (PDB) has no clear and unified definition. According to Engel [22], it refers to the process by which, under certain psychological motivations and various influencing factors, consumers make purchase decisions and engage in purchasing behaviors based upon the information that they have obtained. The consumer decision-making process involves different stages. The classic model of purchase decision behavior has been applied widely in research on consumers’ online purchasing behavior, and includes five stages mainly: need recognition; information search; pre-purchase evaluation; purchase, and post-purchase evaluation [23,24].

Currently, research on consumers' online purchases can be categorized broadly into two main branches. One group of scholars analyzes consumer psychology from the perspective of consumers' motivations, focuses on identifying those motivations [23], and examines the way that different types of motivations influence consumers' purchasing behavior [25,26]. Through a literature review and qualitative research, Martínez-López [27] identified ten dimensions for utilitarian motivations and eleven for hedonic motivations. Further, both hedonic and utilitarian values influence consumers’ behavior strongly, and both motivations contribute significantly to their intention to make online purchases. In addition, word-of-mouth and social value moderate their effects positively [28].

Another group of scholars focuses on external factors, such as marketing activities, influencer characteristics, and website features to explore their effect on consumers' behavior. Xu [29] study revealed that digital inclusiveness influences consumers' green consumption behaviors positively. Zhu [30] conducted a study on the mechanism of online marketing methods' effect on consumers' purchase intention based upon the Stimulus-Organism-Response theory (S-O-R). The results demonstrated that perceived value and perceived usefulness enhanced consumers' willingness to make online purchases significantly. Xu [31] explored the relation between e-commerce influencer characteristics and irrational consumer behavior and found that the more prominent e-commerce influencers' characteristics, the stronger their influence on consumers, which thus promotes irrational consumer behavior. Pan [32] established a relevant research model based upon cross-border e-commerce websites and consumers' traits. The results indicated that cross-border e-commerce website information and service quality affect consumer purchase behavior directly. Moreover, information quality affects consumers’ purchase intention significantly through consumer trust.

Cross-border e-commerce platform advertising recommendations refer to offering store advertisements or product recommendations to consumers proactively based upon certain rules and mechanisms, with the goal to stimulate consumers' purchase intentions and facilitate purchasing behavior [2]. Compared to traditional platform advertising, personalized advertising recommendations can meet consumers' needs, achieve precise targeting, and reduce consumers' resistance better [33]. In Zimmermann [34] research, the author highlighted that personalized recommendations can enhance online consumers' purchasing experience and increase their intention to make a purchase thereby. Further, different types of personalized advertising recommendations and social media marketing have varying effects on consumers' decision-making processes with respect to online green purchases [35]. They have no significant effect during the consideration stage, but have a significant effect during the payment and purchasing stage [15]. However, existing research on platform advertising recommendations’ effect on cross-border e-commerce consumer purchase decisions has focused primarily on studying positive influencing factors, while studies on negative influencing factors are relatively scarce. This is the research gap that this paper intends to address.

2.2 Behavioral Reasoning Theory

Behavioral Reasoning Theory (BRT) is used widely to explain and predict consumers' behavior. Building upon traditional behavioral theories, BRT distinguishes between rational reasons and personal motivations and explores antecedent variables of intentions and behaviors further [36]. BRT suggests that the reasons for opposing behavior differ often from those for supporting behavior. Compared to traditional psychological variables such as attitudes, values, and beliefs, rational reasons have a more direct effect on consumers' decision-making behavior. In addition, as distal factors that influence consumers’ behavior, values have long-term and fundamental effects [37]. BRT bridges values with motivations, intentions, and behaviors through rational reasons, and creates a scientific model to explain behavioral antecedents, as shown in Fig. 1.Fig. 1 Behavioral reasoning theory model.

Fig. 1

Currently, scholars apply the BRT widely in research fields such as online service systems and consumers' purchase intentions. Sobia [38] explored consumers' willingness to purchase suboptimal food products based upon the BRT model. The results indicated that values, behavioral rationality, and motivations influence consumers' purchase intentions directly or indirectly. Chieh [39] integrated the BRT with the Protection Motivation Theory to investigate the mechanism of consumers' decisions to adopt mobile physiotherapy systems, and emphasized health consciousness's significant effect on consumers' adoption decisions.

Ryan and Casidy [20] contend that when applying BRT to measure the rationality of behavior, adoption and rejection behaviors are not simply inverse relations. Rather, they are two distinct viewpoints derived from analyzing consumers' specific behaviors and intentions. However, existing research uses theories such as the Technology Acceptance Model (TAM) [40], Theory of Reasoned Action (TRA) [41], and Theory of Planned Behavior (TPB) predominantly [42,43] to analyze consumers' behavior based upon their reasons for adoption. For instance, factors like flow experience, perceived usefulness [44], and perceived relevance [18] are explored commonly. However, there is scarce exploration from the perspective of reasons for rejection behavior. Moreover, scholars have conducted relatively limited research on consumer behavior's antecedents in cross-border e-commerce shopping. The application of BRT not only validates the applicability of platform-recommended advertisements in promoting consumer purchase behavior in the context of cross-border e-commerce, but also facilitates the exploration of the reasons that platform-recommended advertisements inhibit consumer behavior.

In conclusion, the BRT model can be applied to research on consumers' decision-making behavior, and reveal the mechanism of individual behavior generation, characterized by “values and behavioral rationality - motivation and intention - behavioral performance”. This study's purpose was to conduct antecedent reasoning and validation of purchase decision-making behavior among consumers of cross-border e-commerce platforms. The BRT theory and related research contribute to understanding, predicting, and evaluating consumers' decision-making processes by exploring the dimensions of rationality in cross-border e-commerce consumer behavior. Moreover, this study can expand the application of the BRT theory further and deepen research on cross-border e-commerce consumers.

3 Hypotheses development

3.1 Self-efficacy and purchase decision behavior

Social cognitive theory suggests that self-efficacy is individuals' judgment of their efficacy in interacting with the environment, and individuals with high self-efficacy tend to show interest and engagement in activities [45]. In the context of cross-border e-commerce, platform use self-efficacy refers to consumers' confidence in using cross-border e-commerce platforms for shopping. Based upon the live-streaming marketing model, Apiradee and Nuttapol [46] argued that online marketing through live-streaming can create a good interaction between hosts and viewers, stimulate viewers' interest in watching and making purchases, enhance viewers' self-efficacy, and thus increase their purchase behavior. Hale [47] developed and validated a scale to measure consumers' self-efficacy in making service decisions, and indicated that higher self-efficacy in decision tasks leads to consumers' more efficient decision-making. However, cross-border e-commerce platform consumers may be influenced by factors such as the nationality, language, and cultural habits of the platform operating entity [48], which can decrease their interest in using the cross-border e-commerce platform and undermine their confidence in using the platform for shopping. Therefore, stimulating consumers’ purchase intention and subsequent purchase decision behaviors is more likely when they have sufficient confidence in their ability to use the cross-border e-commerce platform for shopping.H1 Consumers' platform use self-efficacy influences their purchase decision behavior positively.

3.2 Behavioral rationality, self-efficacy and purchase decision behavior

In the BRT model, behavioral rationality is considered a subjective factor used to predict human behavior [49]. Therefore, in exploring the antecedents of consumers' purchase decision behavior, the reasons for adoption or rejection are analyzed from the perspective of consumers’ subjective factors.

3.2.1 Reasons for

This study proposes that the primary reasons for consumer adoption behavior include perceived usefulness and perceived novelty.

Based upon the TAM, perceived usefulness is a key factor that influences consumers’ acceptance of new technology. The stronger the perception of usefulness, the more likely it is that consumers will have positive emotions, which leads to greater acceptance and increased use of new technology [50]. Therefore, when consumers perceive that the platform or its products is highly useful while using a cross-border e-commerce platform, it enhances their confidence in using the platform to fulfill their shopping needs and promotes their purchase decision behavior.

Perceived novelty refers to consumers' perception that the shops and products offered on the cross-border e-commerce platform are novel [51]. Compared to traditional e-commerce platforms, products on cross-border e-commerce platforms often originate from countries different from the consumer's own, which allows consumers to experience different cultures and aesthetics [5]. Novel and unfamiliar products can stimulate and satisfy consumers' curiosity, and motivate them to make purchases thereby. Therefore, perceived novelty is also one of the factors that enhances consumers' self-efficacy and determines their purchase decision behavior.

Based upon the discussion above, the following hypotheses are proposed.H2a Reasons for adoption influence consumers' self-efficacy in using the platform positively.

H2b Reasons for adoption influence consumers' purchase decision behavior positively.

3.2.2 Reasons Against

Innovation Resistance Theory (IRT) suggests that there are five reasons or barriers to engage in a specific task or behavior: use; value; risk; image, and tradition [50]. Considering cross-border e-commerce platforms’ characteristics, this study proposes that the reasons for rejection include primarily perceived intrusiveness, perceived costs, and usability barriers.

Perceived intrusiveness refers to the psychological response to advertisements that disrupt consumers' cognitive processes [18]. Cross-border e-commerce platforms often offer product advertisements based upon consumers' shopping habits and history [52], which not only disrupts their browsing and shopping processes [3] but also makes them more aware that they are targeted, which leads to a sense that their privacy is invaded or insecurity about personal information [53]. This can result in dissatisfaction with, and mistrust of, platform recommendations, which reduce consumers’ confidence in making purchases on the cross-border e-commerce platform.

Perceived costs refer to consumers' perception of economic and time costs overall [54]. When consumers encounter platform advertising recommendations, they need to discern and filter information, which lengthens their shopping browsing time and increases time costs [55]. This can affect their shopping mood, increase mental costs, and lead to interruptions or premature termination of the shopping process. As a result, consumers may develop resistance to platform advertising recommendations, adopt poor attitudes toward the platform, and have a negative shopping experience overall. In addition, compared to domestic e-commerce, cross-border e-commerce involves processes such as international logistics and customs procedures, which lengthen consumers' waiting time. The clearance process and after-sales mechanisms increase the risks associated with cross-border shopping further [56], and weakens consumers’ confidence in shopping on the platform and reduces purchase behavior thereby.

Usability barriers refer to the inconveniences that using the cross-border e-commerce platform cause, which become obstacles for consumers [57]. Because of the different nationalities of the entities that operate cross-border e-commerce platforms and the consumer base, there are differences in language localization, platform web design, shopping processes, and payment methods compared to consumers' domestic e-commerce platforms [58]. These differences may not be consistent with consumers' use habits to some extent, and result in usability barriers. As these barriers increase, consumers’ use emotions decrease, erode their confidence in shopping on the platform gradually, and reduce purchase decisions.

Based upon the discussion above, the following hypotheses are proposed:H3a Reasons for rejection have a negative effect on consumers' self-efficacy in using the platform.

H3b Reasons for rejection have a negative effect on consumers' purchase decision behavior.

3.3 The mediating role of self-efficacy

According to the Expectancy Theory, the degree of an individual's expectation of achieving a goal is influenced by such factors as interest. Having a strong interest in cross-border e-commerce platforms makes consumers more likely to experience pleasure from using the platform [59], which enhances their self-efficacy and promotes purchase decision behavior. Perceived usefulness and perceived novelty (reasons for adoption) satisfy consumers' actual needs and stimulate their latent needs [51], which increases their interest in shopping on cross-border e-commerce platforms greatly. On the other hand, perceived intrusiveness, perceived costs, and usability barriers (reasons for rejection) pose difficulties for consumers during the browsing, payment, and waiting processes [60], and reduce their interest in shopping on cross-border e-commerce platforms thereby. Therefore, different aspects of behavioral rationality can influence consumers' shopping interests, which in turn affects their self-efficacy and influences their purchase decision behavior subsequently.

Based upon the discussion above, the following hypotheses are proposed:H2c Self-efficacy mediates the relation between reasons for adoption and purchase decision behavior.

H3c Self-efficacy mediates the relation between reasons for rejection and purchase decision behavior.

3.4 Tolerance and self-efficacy

With respect to the values component of the BRT, Sahu [61] argued that the values in the theory's framework are inherently contextual, and allows researchers to measure values based upon different research contexts freely. Based upon the schema theory, people's existing cognitive experiences influence their judgment of somethings' value, which influences their subjective behavior in turn [62]. Users of cross-border e-commerce platforms form judgments about platform advertising's value based upon their existing cognitive influences, which results in different levels of tolerance of platform advertising. Therefore, in this study, “tolerance of cross-border e-commerce platform advertising recommendations” is used as a proxy for values, and refers to the degree to which individuals are willing to be exposed to and understand the recommendations. Opening screen advertisements are the first point of contact for most consumers when they access cross-border e-commerce platforms [63]. For example, platforms like Amazon and Wish display popular product advertisements to new consumers on their homepages using big data analysis to stimulate consumers' purchasing desires. When consumers have a higher tolerance of cross-border e-commerce platform advertising recommendations, they accept the platform more, which makes it easier for them to maintain or increase their interest and willingness to use the platform, and increases their confidence in using the platform thereby. Therefore, as tolerance of cross-border e-commerce platform advertising recommendations increases, consumers are more likely to gain confidence in using the platform.

Based upon the discussion above, the following hypothesis is proposed.H4 Tolerance of platform advertising recommendations increases consumers' self-efficacy.

3.5 Tolerance and behavioral rationality

The Theory of Planned Behavior (TPB) [64] and Attribution Theory [36] suggest that values can influence consumers' reasons for adopting or rejecting specific behaviors. In this study, “tolerance of cross-border e-commerce platform advertising recommendations” is used to represent values. Su [65] demonstrated that individuals with a higher tolerance of nature perceive more benefits from engaging in environmentally friendly behavior and have a greater willingness to participate in such behavior. Based upon this, it can be inferred that when consumers have a higher tolerance of cross-border e-commerce platform advertising recommendations, they are more likely to perceive factors in the platform's recommendations that are beneficial to their shopping behavior, such as usefulness and novelty. At the same time, a high level of tolerance can lead consumers to overlook the adverse effects of unfavorable factors on their shopping behavior, such as perceived intrusiveness, perceived costs, and usability barriers to some extent, which influences consumers' reasons for rejecting cross-border e-commerce platform shopping behavior thereby. Therefore, the following hypotheses are proposed.H5a Tolerance of cross-border e-commerce platform advertising recommendations has a positive effect on reasons for adoption.

H5b Tolerance of cross-border e-commerce platform advertising recommendations has a negative effect on reasons for rejection.

3.6 The moderating effect of advertising characteristics

Traditional e-commerce platform advertising recommendations have the characteristics of high frequency, distributed placement, and comprehensive information [66], and cross-border e-commerce platform advertising is no exception. To meet platform merchants' promotional needs and recommend more products to consumers, cross-border e-commerce platforms often have highly frequent and widespread updates of advertisements across various pages, such as the homepage, search page, and payment completion interface, all of which are filled with advertising recommendations [67]. In addition, to cater to consumers' demands, the advertising titles are often loaded with all relevant keywords to increase click-through rates. However, when the update frequency is too high, consumers may find it difficult to capture relevant advertising information promptly [68]. Further, densely placed advertisements and titles with excessive information not only fail to stimulate consumers' desire to make purchases, but also increase the likelihood of information overload and confusion. Therefore, compared to consumers of cross-border e-commerce platforms with weaker advertising features, consumers’ lack of tolerance of platforms with prominent advertising features may inhibit the formation of reasons for adoption. Based upon this, the following hypothesis is proposed.H6 The platform's advertising features have a negative moderating effect on tolerance and reasons for adoption.

Based upon the analysis above, the theoretical model is constructed as shown in Fig. 2.Fig. 2 Research model.

Fig. 2

4 Methodology

This study employs a questionnaire to measure all of the variables. Ethical approval for this survey instrument was obtained from the Ethics Committee of management school at Jiangsu University (Approval Number:JU-ECMS-2023-003). Before participating in the survey, all participants provided informed consent, and the consent process was approved by the Evidence-based Research Center for Educational Assessment.

4.1 Sample and data collection

This study used online questionnaires to collect the data. A professional questionnaire platform, Credamo (https://www.credamo.com/home.html#/), was used to design, distribute, and collect the questionnaires. The Credamo platform offers professional solutions for data collection and serves over 1800 research institutions worldwide and thousands of corporate consumers. In recent years, an increasing number of scholars have chosen to conduct surveys and experiments through Credamo [[69], [70], [71]], as numerous international journals recognize the data's quality.

This study employs an empirical approach to determine the minimum sample size. Generally, it is acknowledged that structural equation modeling analysis is considered reliable when the sample size is at least ten times the number of questionnaire items [72]. This study involved a total of six latent variables and 25 specific items, and the effective number of questionnaires surpassed the theoretical critical threshold significantly (250 samples), and met the requirements for data analysis thereby.

The sampling method used for the online questionnaire was simple random sampling [73]. A total of 498 responses was collected. Among the responses collected, duplicate and invalid questionnaires were removed. According to the pre-survey results, a reasonable time to complete the questionnaire is approximately 40–400 s. Therefore, we excluded questionnaires with excessively short response times (less than 40 s) and excessively long response times (greater than 400 s). In addition, a preliminary screening was conducted on the responses to scale questions (Q7∼Q35). Questionnaires with identical responses to all items or those that demonstrated a clear pattern in item responses (such as responses like “45454545” or “123123123”) were identified. We consider that such responses have a potentially adverse effect on the questionnaire's reliability and validity, as well as on the research results. Consequently, these questionnaires were deemed invalid and excluded. After screening, a total of 427 valid questionnaires remained, for a response rate of 85.7 %. The demographic characteristics of the participants who completed the valid questionnaires were analyzed.

As shown in Table 1, the male-to-female ratio of the sample participants is relatively similar. The majority of respondents are young adults, and a significant proportion are employed in corporations. In addition, most of the sample participants have received at least an undergraduate education, indicating a higher level of education. Overall, the distribution of the sample is reasonable and representative.Table 1 Sample characteristics (N = 427).

Table 1Index	Distributions	Number	%	
Gender	Male	188	44.0	
Female	239	56.0	
Age	18∼25	114	26.7	
26∼35	221	51.8	
36∼45	63	14.8	
46∼55	12	2.8	
>56	17	4.0	
Job	Civil servant	15	3.5	
Education practitioner	14	3.3	
Financial, securities industry practitioner	6	1.4	
Enterprise staff	298	69.8	
Student	65	15.2	
Others	29	6.8	
Education	High school or below	14	3.3	
Junior college	31	7.3	
Bachelor degree	318	74.5	
Master's degree or above	64	15.0	

4.2 Measures

Multiple indicators on a five-point Likert scale were used in this study to measure the constructs of purchase decision behavior, self-efficacy, tolerance, and behavioral rationality. The scale ranged from 1, indicating “completely disagree,” to 5, indicating “completely agree”. The measurement items were adapted from existing relevant literature.

To ensure the questionnaire's reliability, a pre-survey was conducted before formal data collection with 80 experienced consumers of cross-border e-commerce platforms. This pre-survey helped identify any issues of unclear language expressions or inappropriate item design in the questionnaire. Adjustments were made to certain items based upon the pre-survey's results and led to the questionnaire's final version. The measurement items for each construct are presented in Table 2.Table 2 Measurement items and sources.

Table 2Variable	Item	Item setting	Sources	
Purchase decision behavior on CBEC platforms	PDB1	When I need to buy imported goods, I will first choose CBEC platforms.	Adapted from Mark [74]	
PDB2	I am willing to share with others my experience of using CBEC platforms.	
Tolerance of CBEC platform advertising recommendation	Tol1	I accept ads sent to me by CBEC platforms.	Adapted from Tandon [75], Daeseong [76]	
Tol2	I have followed/will follow the ads sent to me by CBEC platforms.	
Self-efficacy of CBEC platforms use	SE1	I have the knowledge and skills to use CBEC platforms for shopping.	Adapted from Rob [77]	
SE2	I am confident that I can use CBEC platforms to shop even if I only have simple instructions to follow.	
SE3	I am confident that I can use CBEC platforms for shopping even without help.	
Reasons for	Perceived usefulness	PU1	I think the products on the CBEC platforms meet my needs.	Adapted from Bonn [78]	
PU2	I think it's faster to buy my favorite products on the CBEC platforms.	
PU3	I think the CBEC platforms have increased my choice of shopping categories.	
Perceived novelty	PN1	I think the content of the CBEC platforms is what I'm interested in.	Adapted from Ki [79]	
PN2	I think the products on CBEC platforms are novel and unique.	
Reasons against	Perceived intrusiveness	PI1	The CBEC platforms made me feel that my privacy and account security were at risk.	Adapted from Sharma [80]	
PI2	I consider the CBEC platforms to be an invasion of my privacy (shopping habits, shopping history, etc.).	
Perceived cost	PC1	Because of the long customs clearance time and long logistics distance, I think shopping on the CBEC platforms will take more time.	Adapted from Rashidin [7]	
PC2	Because of the long logistics distance, I think shopping on CBEC platforms needs to spend more on logistics costs.	
PC3	Compared with other platforms, I think shopping on CBEC platforms takes more energy.	
PC4	I think the price of products on CBEC platforms is higher than that on domestic shopping platforms (or purchasing agents).	
Use disorder	UD1	I think using CBEC platforms makes it more difficult for me to shop online.	Adapted from An [9], Sreen [57]	
UD2	Because the logistics information is not transparent, I think it is inconvenient to use CBEC platforms for shopping.	
UD3	Because of the complexity of after-sales services such as returns and exchanges, I think it is more troublesome to use CBEC platforms for shopping.	
UD4	I don't think the CBEC platforms fit my use habits.	
Advertising characteristics of CBEC platforms	PAC1	I think the frequency of updates recommended by the CBEC platforms is reasonable.	Adapted from Ki [79]	
PAC2	I think the title information of the ads recommended by the CBEC platforms is complicated.	
PAC3	I think the distribution of advertising on the CBEC platforms is reasonable.	

Based upon Mark [74] study, purchase decision behavior was measured according to the selection stage and post-purchase stage. Drawing from Tandon [75] and Daeseong [76] research, self-efficacy was measured with respect to ability and confidence. The measurement of tolerance was adapted from the scales that Rob [77] developed, which examined the willingness to accept and the degree of attention. The measurement of advertising features referred to the scales Ki [79] used that focused on update frequency, information presentation, and page distribution.

Behavioral rationality was divided into two factors: reasons for adopting purchase decision behavior and reasons for rejecting purchase decision behavior. Borrowing from Bonn [78] and Ki [79] studies, reasons for adopting purchase decision behavior were measured based upon perceived usefulness and perceived novelty.

Perceived intrusiveness was measured with respect to consumer privacy infringement, drawing inspiration from the scale that Sharma [80] developed. The measurement of perceived cost referred to Rashidin [7] research that considered time, economic, and mental costs. Drawing from the scales that An [9] and Sreen [57] developed, use barriers were measured with respect to difficulty and convenience. Reasons for rejecting purchase decision behavior were measured from the perspectives of perceived disruptiveness, perceived cost, and use barriers.

4.3 Data analysis methods

This study used the partial least squares structural equation modeling (PLS-SEM) method. PLS-SEM is an approach used commonly for causal modeling. Compared to traditional structural equation modeling (SEM), PLS-SEM has greater flexibility with respect to sample size, data distribution, and complexity. It can manage a wider range of problems and is more suitable for developing theory [81]. As this study's goal is to build and develop a model, and the hypotheses proposed are in the exploratory stage of theory development, an adequate sample size was chosen. However, the data collected in this study deviate from a complete normal distribution, and the model established in this study is relatively complex. Therefore, PLS-SEM was employed for analysis [82]. The data obtained in this study were organized using SPSS v.26 and SmartPLS v.4.0 software was used for data analysis. The bootstrap method with 5000 iterations, employing a two-tailed test and a 95 % confidence interval, was used to assess the significance of the paths based upon p-values.

4.4 Reliability and validity

In this study, Cronbach's alpha (α) and composite reliability (CR) were used to assess the measurement scales' reliability. As shown in Table 3, the α values for all variables ranged from 0.702 to 0.900, and the CR values ranged from 0.839 to 0.921. These values are above the threshold of 0.7 that indicate good or above-average reliability [83]. Thus, the measurement scales demonstrate high structural reliability.Table 3 Reliability and construct validity.

Table 3Variables	Items	Loading	α	CR	AVE	
Purchase decision behavior	PDB		0.702	0.870	0.771	
PDB1	0.876				
PDB2	0.880				
Self-efficacy	SE		0.719	0.842	0.640	
SE1	0.809				
SE2	0.771				
SE3	0.819				
Tolerance	Tol		0.705	0.871	0.772	
Tol1	0.867				
Tol2	0.890				
Advertising characteristics	AC		0.712	0.839	0.635	
AC1	0.808				
AC2	0.766				
AC3	0.816				
Reasons for	RF		0.784	0.921	0.625	
Perceived usefulness	PU			0.874	0.777	
PU1	0.804				
PU2	0.777				
Perceived novelty	PN			0.872	0.772	
PN1	0.747				
PN2	0.787				
Reasons against	RA		0.900	0.860	0.607	
Perceived intrusiveness	PI					
PI1	0.738				
Perceived cost	PC			0.903	0.756	
PC1	0.799				
PC2	0.780				
PC3	0.809				
Use disorder	UD			0.888	0.726	
UD1	0.762				
UD2	0.799				
UD3	0.843				
Note: Cronbach's alpha (α) represents the Cronbach coefficient, CR represents composite reliability, AVE represents the average variance extracted.

This study used the average variance extracted (AVE) and external loadings to assess the measurement scales' convergent validity, which indicates the questionnaire's structural validity. AVE values greater than 0.5 [81] and external loadings above 0.708 are considered satisfactory. As shown in Table 3, the AVE values for each latent variable ranged from 0.607 to 0.777, and the range of external loadings was between 0.738 and 0.890. These values exceed the thresholds recommended and indicate the measurement scales' good construct validity.

This study measured the measurement scales' discriminant validity according to Fornell and Larcker's method. Based upon this method, if the square root of AVE for each variable is greater than its correlation with other variables, it indicates good discriminant validity [84]. As shown in Table 4, the diagonal elements represent each variable's square roots of AVE, while the off-diagonal elements represent the correlations between variables. The diagonal values are all greater than the off-diagonal values in their respective columns, and thus satisfy the criteria of Fornell and Larcker's method. This indicates that the measurement scales have good discriminant validity.Table 4 Correlation coefficients and discriminant validity.

Table 4	Variables	1	2	3	4	5	6	
1	Purchase decision behavior	0.878						
2	Self-efficacy	0.538	0.800					
3	Tolerance	0.600	0.528	0.879				
4	Reasons for	0.533	0.492	0.539	0.779			
5	Reasons against	−0.405	−0.392	−0.507	−0.422	0.791		
6	Advertising characteristics	0.517	0.483	0.683	0.593	−0.602	0.790	
Note: The value on the diagonal is the square root of AVE, and the value on the non-diagonal is the correlation coefficient between the variables.

Overall, the measurement scales demonstrate high reliability and validity based upon the assessments conducted.

4.5 Common method bias

This study employed self-reported survey data to mitigate common method bias. To ensure anonymity and prevent duplicate responses, the survey questionnaire was designed to be answered anonymously. In addition, an exploratory factor analysis using Harman's single-factor test was conducted to examine the presence of common method bias in the data [74]. The results revealed that there were three factors with eigenvalues greater than 1, and the largest factor accounted for 42.718 % of the variance, which is less than 50 %. Hence, there is no severe common method bias in this study.

Further, we assessed the model's goodness of fit. As scholars have not reached a unified opinion on the model fit indicators for PLS-SEM, we referred to Structural Equation Modeling (SEM) and selected certain indicators for examination [85,86]. Among them, X2/df serves as an alternative test value for Chi-square, which mitigates the sample size's influence on fit assessment. The Normed Fit Index (NFI) reflects the difference between the hypothesized model and the independent model [87]. The Goodness of Fit Index (GFI) is used to assess a model's adequacy in fitting the observed values within a sample dataset [88]. The Standardized Root Mean Square Residual (SRMR) evaluates the model's adequacy by computing the difference between the predicted and actual measurement values [89]. As shown in Table 5, the results indicate a good fit of the model.Table 5 Overall fitting coefficient of the model.

Table 5Index	X2/df	NFI	GFI	SRMR	
Judging standard	<3	>0.9	>0.9	<0.1	
Value	2.368	0.902	0.917	0.061	

5 Results

5.1 Direct effects analysis

In this study, path coefficients (β), p-values, t-values, and F2 were used to indicate the direction, strength, and significance of the relations hypothesized. As shown in Table 6, the results indicate that all direct paths (H1-H5b) have significant coefficients, with support at the p < 0.001, <0.01, and <0.05 levels. In addition, the F2 values range from 0.20 to 0.346, indicating moderate to large effects [90].Table 6 Direct effect test.

Table 6	Path	β	SD	t-value	p-value	Results	
H1	Self-efficacy → PDB	0.328	0.059	5.550	0.000c	Support	
H2a rowhead	RF → Self-efficacy	0.269	0.069	3.898	0.000c	Support	
H2b rowhead	RF → PDB	0.310	0.060	5.129	0.000c	Support	
H3a rowhead	RA → Self-efficacy	−0.114	0.057	1.984	0.047a	Support	
H3b rowhead	RA → PDB	−0.146	0.046	3.156	0.002b	Support	
H4	Tolerance → Self-efficacy	0.326	0.062	5.218	0.000c	Support	
H5a rowhead	Tolerance → RF	0.159	0.059	2.677	0.007b	Support	
H5b rowhead	Tolerance → RA	−0.507	0.037	13.649	0.000c	Support	
Note.

a p < 0.05.

b p < 0.01.

c p < 0.001.

The results of the study provide several key findings. First, self-efficacy demonstrates a significant positive effect on PDB (β = 0.328, t = 5.550, p < 0.001), thereby supporting H1. This suggests that as consumers’ self-efficacy in using the platform increases, they are more likely to engage in purchase decision-making behaviors. Secondly, RF exerts a positive and significant influence on self-efficacy (β = 0.269, t = 3.898, p < 0.001), while RF also has a positive and significant effect on PDB (β = 0.310, t = 5.129, p < 0.01), respectively, and supports H2a and H2b. This indicates that the reasons that prompt consumers to accept advertising recommendations can enhance their self-efficacy and purchase decision-making behaviors.

The study reveals that RA has a significant negative effect on self-efficacy (β = −0.114, t = 1.984, p < 0.05). Similarly, RA demonstrates a significant negative influence on PDB (β = −0.146, t = 3.156, p < 0.01), which provides support for H3a and H3b, respectively. This implies that the stronger consumers’ rejection of advertising recommendations, the less likely they are to develop self-efficacy in using the platform and engaging in purchase decision-making behaviors.

Tolerance has a significant positive effect on self-efficacy (β = 0.326, t = 5.218, p < 0.001), and supports H4 thereby. This indicates that consumers' self-efficacy increases as their tolerance increases. Moreover, tolerance exhibits a significant positive effect on RF (β = 0.159, t = 2.664, p < 0.01), which supports H5a. Conversely, inclusiveness has a significant negative effect on RA (β = −0.507, t = 13.649, p < 0.001), and thus supports H5b. This suggests that consumers' acceptance and rejection of advertising recommendations vary with tolerance. The greater consumers’ tolerance of advertising recommendations, the more likely they are to accept such recommendations, while the reasons for rejecting advertising recommendations decrease, and the extent of the decrease becomes more pronounced.

5.2 Indirect effects analysis

5.2.1 Mediation effects testing

This study used the Bootstrap method (with a subsample size of 5000, two-tailed test, and a 95 % confidence interval) to examine self-efficacy's mediating role. The results presented in Table 7 show the direct relations between perceived behavioral control, self-efficacy, and purchase decision behavior. The results indicate that self-efficacy plays a significant positive mediating role in the relation between RF and purchase decision-making behavior (β = 0.088, t = 3.429, p < 0.01, 0.038 < BootCl <0.139), which supports H2c. Further, self-efficacy plays a significant negative mediating role in the relation between RA and purchase decision-making behavior (β = −0.037, t = 2.055, p < 0.05, −0.073 < BootCl < −0.002), which provides support for H3c. This suggests that the reasons that consumers accept advertising can promote their purchase decision-making behavior by increasing their self-efficacy. Conversely, consumers' rejection of advertising recommendations decreases their self-efficacy, and reduces their purchase decision-making behavior thereby.Table 7 Total effect and mediation effect testing.

Table 7	Path	β	t-value	p-value	95 % BootCl	Result	
LLCl	ULCl	
Total effect	RF→PDB	0.398	6.347	0.000c	0.275	0.521		
RA→PDB	−0.183	3.885	0.000c	−0.275	−0.091		
H2c rowhead	RF → Self-efficacy → PDB	0.088	3.429	0.001b	0.038	0.139	Support	
H3c rowhead	RA → Self-efficacy → PDB	−0.037	2.055	0.040a	−0.073	−0.002	Support	
Note.

a p < 0.05.

b p < 0.01.

c p < 0.001.

5.2.2 Moderation effect testing

In this study, latent variables’ moderation effects were tested directly using the Smart-PLS command within the SEM framework. The moderator variable used in this study was “advertising features”, and the Bootstrap method (with a subsample size of 5000, two-tailed test, and a 95 % confidence interval) was employed to examine the moderation effects.

The results of the moderation effect analysis are presented in Table 8. The study found that advertising features have a significant negative moderating effect on the relation between inclusiveness and adoption reasons (β = −0.153, t = 2.712, p < 0.01), and support H9. This indicates that as advertising features become more prominent, inclusiveness's influence on consumers' reasons for adoption weakens. The moderation effect plot is depicted in Fig. 3.Table 8 Moderation effect testing.

Table 8	Path	β	SD	t-value	p-value	Result	
H6	AC × Tol → RF	−0.153	0.056	2.712	0.007**	Support	
Note: *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 3 Advertising features' moderating effect on tolerance and adoption reasons.

Fig. 3

For a comprehensive understanding of the mechanism of consumers’ purchase decision behavior in cross-border e-commerce platforms, please refer to Fig. 4 in the research results.Fig. 4 Mechanism of Purchasing Decision-making Behavior of Cross-border e-commerce Platform Consumers.

Fig. 4

6 Discussion

Based upon the TRA, this study investigates the antecedents that influence cross-border e-commerce consumers’ purchase decision behavior and reveals the mechanism of their decision-making process in the context of exposure to advertising on cross-border e-commerce platforms.

The results indicate that consumers' self-efficacy and perceived behavioral control affect their purchase decision-making behavior on the platform significantly. Specifically, self-efficacy has a positive effect on purchase decision-making behavior, which is consistent with the conclusions of previous studies that Fan and Liu [91], Luo [92], and other scholars conducted in different fields, such as artificial intelligence and live streaming, which emphasized the importance of self-efficacy in decision-making behavior. The reasons for adoption behaviors affect decision-making behavior directly and influence self-efficacy indirectly, which promotes consumers' purchase decision-making behavior thereby. Conversely, the reasons for rejection behaviors have a significant negative effect on self-efficacy and purchase decision-making behavior, which support Ashfaq [21] findings and validate self-efficacy's mediating role further, which enriches the research results. Further, this study reveals that the effect of reasons for adoption on consumers' purchase decision-making behavior is greater than the effect generated by reasons for rejection. This finding is consistent with the research that Wang [93] conducted, as he suggested that positive reasons that determine behavior are more likely to influence individuals.

Therefore, enhancing consumers' self-efficacy is an ideal entry point for cross-border e-commerce platforms to improve consumers' purchasing behavior. When these platforms expand their product advantages continuously with respect to categories, uniqueness, and innovation, optimize backend algorithms and related services, and alleviate consumers’ concerns about privacy and time costs, their experience and confidence in the platform will be enhanced greatly and lead to purchase decision-making behavior.

Further, personal values, as the ultimate determinant of consumers’ behavior, can influence purchase decision-making behavior through self-efficacy and perceived behavioral control. This finding corresponds with the results of Chieh [94] study, which found that personal values influence perceived behavioral control significantly, and the effect of reasons for adoption and rejection varies in different contexts. It is worth noting that in this study, “tolerance of cross-border e-commerce platform advertising recommendations” is used to represent personal values, but it does not mean that tolerance is the only substitute variable for personal values. Because personal values can be measured freely according to different research contexts, future studies can choose appropriate variables to represent personal values based upon different research backgrounds and objectives.

Moreover, it is crucial to highlight that this study validates subjective reasons that affect consumers’ purchase behavior based upon two dimensions: reasons for adoption and reasons for rejection. While previous studies have examined the reasons for adoption extensively, this study focuses on exploring and analyzing the reasons for rejection behavior as well. Although efforts have been made to select factors applicable to major cross-border e-commerce platforms, it is important to acknowledge that limitations may still exist.

In addition, this study selects cross-border e-commerce platforms' advertising characteristics as the moderating variable between tolerance and reasons for adoption. However, differences in page design, algorithm settings, platform operators, and other aspects can lead to unique advertising characteristics for different platforms. La [95] also studied other related advertising features, but he focused primarily on advertisements' direct effect on consumer behavior, while this study examines their influence on consumers’ psychology with advertising characteristics as the moderating variable. Therefore, future research can explore the variable of advertising characteristics further to enhance its comprehensiveness.

7 Conclusion and implications

This study investigated the mechanisms that influence consumers' purchasing decision behavior on cross-border e-commerce platforms in the context of advertising recommendations. In this process, the PLS-SEM method was employed to examine the relations among the variables. The research findings are as follows: the path coefficients between self-efficacy and purchase decision behavior, as well as between behavioral rationality and purchase decision behavior were positive and significant, indicating these two factors' direct and crucial roles in influencing consumers' purchasing decision behavior. Moreover, behavioral rationality can promote purchase decision behavior further by enhancing self-efficacy. In addition, tolerance can enhance consumers’ self-efficacy and behavioral rationality, while advertising characteristics moderate tolerance and reasons for adoption.

7.1 Theoretical contributions

This study makes two significant contributions to research on consumer behavior in cross-border e-commerce platforms.

First, it expands on existing research by examining both positive and negative reasons that influence consumer behavior. Previous studies have focused predominantly on positive reasons for adoption behavior, while they have overlooked the reasons for consumers' rejection behavior. While some domestic scholars have explored the influence of shopping intention on different types of consumption behaviors, such as green consumption and online consumption, limited research has delved deeply into the antecedents of consumers' behaviors and intentions. This study addresses this gap by investigating the antecedents and mechanisms of consumers' shopping decision-making behavior on cross-border e-commerce platforms, particularly within the context of intelligent advertising. Understanding these factors is crucial to develop effective advertising strategies and understand consumers’ shopping psychology in cross-border e-commerce platforms.

Second, this study contributes to the development of BRT in the field of cross-border e-commerce research. While some scholars have applied BRT to the study of consumers’ shopping decision-making behavior, its application within the context of cross-border e-commerce and the inclusion of platform advertising as a factor are relatively scarce. By conducting empirical research, this study unveils the influencing factors and mechanisms of purchase decision-making behavior among consumers on cross-border e-commerce platforms. This research fills a gap that scholars have identified with respect to the need for consumer research within the field of cross-border e-commerce.

Overall, this study sheds light on important aspects of consumer behavior in cross-border e-commerce platforms, and provides insights into consumers' decision-making processes and advertising strategies’ role.

7.2 Managerial implications

The application of research on cross-border e-commerce platform advertising has significant implications for cross-border e-commerce's competitiveness and offers constructive solutions for its enhancement. To optimize the effectiveness of cross-border e-commerce platforms' advertising and stimulate consumers' purchasing behavior, strategies and recommendations are proposed from the perspectives of both platform operators and cross-border e-commerce enterprises (stores).

First, it is suggested that cross-border e-commerce platform operators should optimize their advertising strategies from the perspective of consumers to a greater degree. The survey findings indicate that consumers accept advertisements within platforms highly. Therefore, platform operators can consider increasing the intensity of their advertising distribution appropriately. However, attention should be paid to the frequency of distribution notifications and advertisements' distribution on pages. Excessive distribution frequency and densely packed page advertisements are prone to evoke consumers' aversion. Hence, in formulating advertising distribution strategies, platforms should set reasonable thresholds for advertising refresh rates, page distribution, and information presentation to enhance consumers’ favorable responses and increase purchase intention.

For cross-border e-commerce enterprise (store) managers, emphasis should be placed on the advertisements' content and design. Managers should select different products for advertising promotion based upon the diverse consumer groups that different platforms target. In addition, cross-border e-commerce stores' homepage interface elicits consumers' first impressions of the store. Therefore, when designing advertisements for the store's homepage, managers should opt for visually appealing and concise advertisement images and minimize complex textual descriptions. This approach can have a stronger visual effect on consumers and enhance their impression of the store. Moreover, store managers should collaborate actively with cross-border e-commerce platforms in advertising activities. They should provide information about best-selling products, those with distinctive store features, or innovative products, as well as information about products featured at specific times, including images and brief product descriptions. Active participation in platform advertising promotion contributes to increasing store exposure, attracts more consumers to browse and purchase on the store's homepage, and ultimately increases store sales and competitiveness thereby.

7.3 Limitations of the study and Suggestions for further research

As in all research, this study has certain limitations that provide avenues for future research. First, there is scope for improvement in sample selection. Given cross-border e-commerce platforms' diverse categorizations, conducting research without delineating platform types may result in incomplete survey outcomes. As an extension of this study, future investigations could delve deeper into analyzing various categories of cross-border e-commerce platforms. In addition, while the research subjects in this study were exclusively Chinese consumers, the consumer base of cross-border e-commerce platforms is global, such that consumption behaviors and preferences are influenced by diverse educational and cultural backgrounds. Consequently, regional disparities may exist in the research findings. Future endeavors could explore geographical and cultural factors’ effect on consumer behavior.

Second, there is room for further exploration in selecting the research methodology. As this study pertains to causal research on consumer behavior, PLS-SEM was chosen to validate the model. However, limitations such as the model's insufficient explanatory power and rudimentary evaluation metrics persist. Future investigations could employ alternative methodologies such as regression models, fsQCA, etc., to deepen the understanding of consumer behavior in the context of cross-border e-commerce.

Ethical statement

The study was approved by the Ethics Committee of management school at Jiangsu University (Approval Number: JU-ECMS-2023-003). All participants provided written informed consent for the use of their data for scientific research purposes.

Data availability

The data relevant to our study have not yet been deposited into a publicly available database. The data used in this study are available upon request.

Funding

The authors would like to acknowledge the funding support of the 10.13039/501100012456 National Social Science Foundation of China (No. 22GBL102 ); Postgraduate Research & Practice Innovation Program of Jiangsu Province (KYCX23_3798 ).

CRediT authorship contribution statement

Yiwei Xu: Writing – review & editing, Writing – original draft, Methodology, Investigation, Funding acquisition, Data curation. Di He: Writing – review & editing, Funding acquisition. Mingyue Fan: Writing – review & editing, Supervision, Methodology, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Yiwei Xu (Master's student, Jiangsu University) is currently studying in the second year of the School of Management, Jiangsu University. Her main research direction is cross-border e-commerce and consumer behavior. She has participated in many enterprise research activities.

Di He (Master, Jiangsu University) is currently a professor at the School of Management, Jiangsu University. She has been teaching and researching economics and marketing for a long time and is currently the deputy Director of the MBA Education Center of Jiangsu University. Professor He's main research interests are enterprise development strategy management, human resource management, business model innovation, etc. He has published more than 20 economics and management papers in core journals, and participated in the compilation of 2 textbooks.

Mingyue Fan (Ph.D., Jiangsu University) is currently an associate professor in the Department of Information Management and Information System, School of Management, Jiangsu University, and a member of the Economic Management Branch of Jiangsu Automation Society. In recent years, she has been engaged in research on business management, business ethics and information management. Her contributions have been showcased through the publication of more than 30 papers in reputable journals such as the Journal of Retailing and Consumer Services, Environmental Science and Pollution Research, Frontiers in Psychology, Frontiers in Environmental Science, Electronic Commerce Research and Application, Plos One, and other distinguished international journals. Dr. Fan has demonstrated her commitment to the academic community by undertaking roles as a guest deputy editor and reviewer for Electronic Commerce Research and Application, Frontiers in Psychology, Plos One, and various other esteemed international journals.

Appendix A Supplementary data

The following is the Supplementary data to this article:Multimedia component 1

Multimedia component 1

Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.heliyon.2024.e37627.
==== Refs
References

1 Chen J.H. Lan Y.C. Chang Y.W. Consumer behaviour in cross-border e-commerce: systematic literature review and future research agenda Int. J. Consum. Stud. 47 6 2023 2609 2669 10.1111/ijcs.12969
2 Li B. Li J.H. Ou X.J. Hybrid recommendation algorithm of cross-border e-commerce items based on artificial intelligence and multiview collaborative fusion Neural Comput. Appl. 34 9 2022 6753 6762 10.1007/s00521-021-06249-3
3 Aribarg A. Schwartz E.M. Native advertising in online news: trade-offs among clicks, brand recognition, and website trustworthiness J. Market. Res. 57 3 2020 002224371987971 10.1177/0022243719879711
4 Kim E. Kim D. E Z. Shoenberger H. The next hype in social media advertising: examining virtual influencers' brand endorsement effectiveness Front. Psychol. 14 2023 10.3389/fpsyg.2023.1089051
5 Sun Y. Li Y. The impact of risk-aware consumer trust on CB E-commerce platforms and purchase intention J. Global Inf. Manag. 30 2 2022 13 10.4018/JGIM.20220701.oa10
6 Prompijit A. Consumer Online Search Behavior: an Empirical Analysis 2021 University of California Irvine: California, United States
7 Rashidin M.S. Gang D. Javed S. Hasan M. The role of artificial intelligence in sustaining the E-commerce ecosystem: alibaba vs. Tencent J. Global Inf. Manag. 30 2022 1 25 10.4018/JGIM.304067
8 Chang Y. Wong S.F. Libaque-Saenz C.F. Lee H. E-commerce sustainability: the case of pinduoduo in China Sustainability 11 15 2019 10.3390/su11154053
9 An D. Ji S. Jan I.U. Investigating the determinants and barriers of purchase intention of innovative new products Sustainability 13 2 2021 10.3390/su13020740
10 Zhou F. Zhang C. Chen T. Lim M.K. An evolutionary game analysis on blockchain technology adoption in cross-border e-commerce Operations Management Research 2023 10.1007/s12063-023-00382-z
11 Valarezo A. Perez-Amaral T. Garin-Munoz T. Herguera G.L. Lopez R. Drivers and barriers to cross-border e-commerce: evidence from Spanish individual behavior Telecommun. Pol. 42 6 2018 464 473 10.1016/j.telpol.2018.03.006
12 Jian J. Lv L. Wan L. Exploring consumer purchase intention in cross-border e-commerce: evidence from 'belt and road' countries Asia Pac. J. Mark. Logist. 35 3 2023 625 644 10.1108/apjml-12-2021-0934
13 Mou J. Cui Y. Kurcz K. Trust, risk and alternative website quality in B-buyer acceptance of cross-border E-commerce J. Global Inf. Manag. 28 1 2020 167 188 10.4018/jgim.2020010109
14 Zhu W. Mou J. Benyoucef M. Exploring purchase intention in cross-border E-commerce: a three stage model J. Retailing Consum. Serv. 51 2019 320 330 10.1016/j.jretconser.2019.07.004
15 Jeong J. Kim D. Li X. Li Q. Choi I. Kim J. An empirical investigation of personalized recommendation and reward effect on customer behavior: a stimulus-organism-response (SOR) model perspective Sustainability 14 22 2022 10.3390/su142215369
16 Wang J. Xi Y. Zhao Q. Influence of personalized advertising recommendation types on online green purchase decision-making process China Population Resources and Environment 31 3 2021 108 116
17 Liu B.J. Wei L.W. Machine gaze in online behavioral targeting: the effects of algorithmic human likeness on social presence and social influence Comput. Hum. Behav. 124 2021 10.1016/j.chb.2021.106926
18 Van den Broeck E. Zarouali B. Poels K. Chatbot advertising effectiveness: when does the message get through? Comput. Hum. Behav. 98 2019 150 157 10.1016/j.chb.2019.04.009
19 Le-Anh T. Nguyen M.D. Nguyen T.T. Duong K.T. Energy saving intention and behavior under behavioral reasoning perspectives Energy Efficiency 16 2 2023 10.1007/s12053-023-10092-x
20 Ryan J. Casidy R. The role of brand reputation in organic food consumption: a behavioral reasoning perspective J. Retailing Consum. Serv. 41 2018 239 247 10.1016/j.jretconser.2018.01.002
21 Ashfaq M. Zhang Q. Ali F. Waheed A. Nawaz S. You plant a virtual tree, we'll plant a real tree: understanding users ' adoption of the Ant Forest mobile gaming application from a behavior al reasoning theory perspective J. Clean. Prod. 310 2021 127394 10.1016/j.jclepro.2021.127394
22 Engel J.F. Blackwell R.D. Miniard P.W. Consumer Behavior 1968
23 Liu F. Lim E.T.K. Li H. Tan C.-W. Cyr D. Disentangling utilitarian and hedonic consumption behavior in online shopping: an expectation disconfirmation perspective Inf. Manag. 57 3 2020 10.1016/j.im.2019.103199
24 Howard J.A. Bither S.W. Sheth J.N. The theory of buyer behavior J. Market. 35 1 1971 102
25 Li Z. Zhao X. Ou G. Understanding customer satisfaction in curated subscription services: moderating roles of subscription time and information disclosure Inf. Manag. 60 6 2023 10.1016/j.im.2023.103834
26 Lee J. Sadachar A. Investigating the conceptual link between appearance-related self-discrepancies and retail therapy shopping behavior through motivational routes J. Fash. Mark. Manag. 2023 10.1108/jfmm-07-2022-0160
27 Martínez-López F.J. Pla-García C. Gázquez-Abad J. Carlos Rodríguez-Ardura I. Online Consumption Motivations: an Integrated Theoretical Delimitation and Refinement Based on Qualitative Analyses 2014 Springer Berlin Heidelberg
28 Akram U. Junaid M. Zafar A.U. Li Z. Fan M.Y. Online purchase intention in Chinese social commerce platforms: Being emotional or rational? J. Retailing Consum. Serv. 63 5 2021 102669 10.1016/j.jretconser.2021.102669
29 Xu D.D. Guo D.L. Yue P.P. Li M.S. Household green consumption: does digital inclusion matter? Int. Rev. Financ. Anal. 91 2024 16 10.1016/j.irfa.2023.102977
30 Zhu P. Liu Z. Li X. Jiang X. Zhu M.X. The influences of livestreaming on online purchase intention: examining platform characteristics and consumer psychology Ind. Manag. Data Syst. 123 3 2023 862 885 10.1108/imds-07-2022-0430
31 Xu Y.J. Jiang W.Q. Li Y. Guo J. The influences of live streaming affordance in cross-border E-commerce platforms: an information transparency perspective J. Global Inf. Manag. 30 2 2022 10.4018/JGIM.20220301.oa3
32 Pan Y. Wang M. Chen C. Qu H. Research on the influence of web experience on consumers' purchasing intention J. Bus. Adm. Res. 6 2017 8 10.5430/jbar.v6n2p8
33 Li C. An advertising recommendation algorithm based on deep learning fusion model J. Sens. 17 9 2022 10.1155/2022/1632735
34 Zimmermann R. Mora D. Cirqueira D. Helfert M. Bezbradica M. Werth D. Weitzl W.J. Riedl R. Auinger A. Enhancing brick-and-mortar store shopping experience with an augmented reality shopping assistant application using personalized recommendations and explainable artificial intelligence Journal of Research in Interactive Marketing 17 2 2023 273 298 10.1108/jrim-09-2021-0237
35 Sun Y. Li T.S. Wang S.Y. "I buy green products for my benefits or yours": understanding consumers' intention to purchase green products Asia Pac. J. Mark. Logist. 34 8 2022 1721 1739 10.1108/apjml-04-2021-0244
36 Westaby J.D. Behavioral reasoning theory: identifying new linkages underlying intentions and behavior Organ. Behav. Hum. Decis. Process. 98 2 2005 10.1016/j.obhdp.2005.07.003
37 Thøgersen J. Zhou Y. Huang G. How stable is the value basis for organic food consumption in China? J. Clean. Prod. 134 2015 10.1016/j.jclepro.2015.06.036
38 Sobia T.H. Shahid Y.R.M. Mohammed A.A. Sidra R. Usman S.A. Moamen S.R. Consumers' purchase intention of suboptimal food using behavioral reasoning theory: a food waste reduction strategy Sustainability 14 14 2022 10.3390/su14148905
39 Chieh L.J. Li C. Hengrui Z. Exploring the adoption decisions of mobile health service users: a behavioral reasoning theory perspective Ind. Manag. Data Syst. 123 8 2023 10.1108/IMDS-11-2022-0682
40 Cai H. Examining social E-commerce platforms by mediating the effect of perceived usefulness and perceived trust using the technology acceptance model J. Organ. End User Comput. 34 8 2022 10.4018/joeuc.315621
41 Raman P. Understanding female consumers' intention to shop online the role of trust, convenience and customer service Asia Pac. J. Mark. Logist. 31 4 2019 1138 1160 10.1108/apjml-10-2018-0396
42 Trude A.C.B. Ali S.H. Lowery C.M. Vedovato G.M. Lloyd-Montgomery J.M. Hager E.R. Black M.M. A click too far from fresh foods: a mixed methods comparison of online and in-store grocery behaviors among low-income households Appetite 175 2022 10.1016/j.appet.2022.106038
43 Sun Y. Leng K. Xiong H.T. Research on the influencing factors of consumers? green purchase behavior in the post-pandemic era J. Retailing Consum. Serv. 69 2022 10.10.1016/j.jretconser.2022.103118
44 Rieh S.Y. Collins-Thompson K. Hansen P. Lee H.J. Towards searching as a learning process: a review of current perspectives and future directions J. Inf. Sci. 42 1 2016 19 34 10.1177/0165551515615841
45 Compeau D.R. Higgins C.A. Computer self-efficacy: development of a measure and initial test MIS Q. 19 2 1995 189 211 10.2307/249688
46 Apiradee W. Nuttapol A. The role of live streaming in building consumer trust and engagement with social commerce sellers J. Bus. Res. 117 2020 543 556 10.1016/j.jbusres.2018.08.032
47 Hale D. Thakur R. Riggs J. Altobello S. Consumers' decision-making self-efficacy for service purchases: constr uct conceptualization and scale J. Serv. Market. 36 5 2022 637 657 10.1108/jsm-12-2020-0505
48 Ruiz-Mafe C. Chatzipanagiotou K. Curras-Perez R. The role of emotions and conflicting online reviews on consumers' purchase intentions J. Bus. Res. 89 2018 336 344 10.1016/j.jbusres.2018.01.027 aug.
49 Rita K. Roger B. Rohini V. Bettina G. Driverless futures: current non-drivers’ willingness to travel in driverless vehicles J. Market. Manag. 37 15–16 2021 10.1080/0267257X.2021.1963309
50 Hew J.-J. Lee V.-H. Leong L.-Y. Why do mobile consumers resist mobile commerce applications? A hybrid fsQCA-ANN analysis J. Retailing Consum. Serv. 75 2023 10.1016/j.jretconser.2023.103526
51 Togawa T. Ishii H. Park J. Roy R. The temperature of newness: how vision–temperature correspondence in advertising influences newness perception and product evaluation J. Bus. Res. 160 2023 10.1016/j.jbusres.2023.113801
52 Huang X. Liu X.J. Incorporating a topic model into a hypergraph neural network for searching-scenario oriented recommendations Applied Sciences-Basel. 12 15 2022 10.3390/app12157387
53 Alexander G.D. Jorge G.D. A simplified mobile advertising model to study advertising spreading through personal social networks and branded apps J. Promot. Manag. 29 8 2023 10.1080/10496491.2023.2189204
54 Alyahya M. Agag G. Aliedan M. Abdelmoety Z.H. Understanding the factors affecting consumers' behaviour when purchasing refurbished products: a chaordic perspective J. Retailing Consum. Serv. 75 2023 10.1016/j.jretconser.2023.103492
55 Amblee N. Ullah R. Kim W. Do product reviews really reduce search costs? J. Organ. Comput. Electron. Commer. 27 3 2017 199 217 10.1080/10919392.2017.1332142
56 Feng Y. Green progress of cross-border E-commerce industry utilizing random forest algorithm and panel tobit model Appl. Artif. Intell. 37 1 2023 10.1080/08839514.2023.2219561
57 Sreen N. Dhir A. Talwar S. Tan T.M. Alharbi F. Behavioral reasoning perspectives to brand love toward natural products: moderating role of environmental concern and household size J. Retailing Consum. Serv. 61 2021 10.1016/j.jretconser.2021.102549
58 Zhu W. Mou J. Cohen J.F. A cross-continent analysis of the invariance of product information in cross-border electronic commerce J. Global Inf. Manag. 29 6 2021 10.4018/jgim.289654
59 Erkan I. Evans C. Social media or shopping websites? The influence of eWOM on consumers' online purchase intentions J. Market. Commun. 24 6 2018 617 632 10.1080/13527266.2016.1184706
60 Wei X. Ko I. Pearce A. Does perceived advertising value alleviate advertising avoidance in mobile social media? Exploring its moderated mediation effects Sustainability 14 1 2021 10.3390/su14010253
61 Sahu A.K. Padhy R.K. Dhir A. Determinants and barriers of implementing lean manufacturing practices in MSMEs: a behavioural reasoning theory perspective Prod. Plann. Control 33 12 2022 1197 1213 10.1080/09537287.2020.1857449
62 Scheufele D.A. Tewksbury D. Framing, agenda setting, and priming: the evolution of three media effects models J. Commun. 57 1 2007 9 20 10.1111/j.0021-9916.2007.00326.x
63 Han J. Optimization analysis of advertising information resource allocation in view of the dynamic game model J. Math. 2022 2022 10.1155/2022/7445124
64 Pennington N. Hastie R. Explanation-based decision making: effects of memory structure on judgment J. Exp. Psychol. Learn. Mem. Cognit. 14 3 1988 10.1037/0278-7393.14.3.521
65 Su L.J. Hsu M.K. Boostrom R.E. From recreation to responsibility: increasing environmentally responsible behavior in tourism J. Bus. Res. 109 2020 557 573 10.1016/j.jbusres.2018.12.055
66 Li L. Economics of Recommender Systems in Online Marketplaces 2017 The University of Texas at Dallas
67 Duffett R.G. Facebook advertising's influence on intention-to-purchase and purchase amongst Millennials Internet Res. 25 4 2015 498 526 10.1108/IntR-01-2014-0020
68 Han L. Xu F. Xu F. Hu M. Hou G. Inside Airbnb titles: the location choices of listings from Chinese hosts' advertising expressions Tourism Manag. Perspect. 48 2023 10.1016/j.tmp.2023.101151
69 Cai Q. Public trust in the Chinese government and life satisfaction during the early stages of the COVID-19 pandemic Int. Rev. Adm. Sci. 89 4 2023 1165 1185 10.1177/00208523231161297
70 Huang Z. Gan Q. Luo M. Zhang Y. Ge J. Fu Y. Chen Z. Social support and prosocial behavior in Chinese college students during the COVID-19 outbreak: a moderated mediation model of positive affect and parental care Front. Psychol. 14 2023 10.3389/fpsyg.2023.1127194
71 Okojie G. Ismail I.R. Begum H. Ferdous Alam A.S.A. Sadik-Zada E.R. The mediating role of social support on the relationship between employee resilience and employee engagement Sustainability 15 10 2023 7950 10.3390/su15107950
72 Jackson D.L. Revisiting sample size and number of parameter estimates: some support for the N:q hypothesis Struct. Equ. Model.: A Multidiscip. J. 10 1 2003 128 141 10.1207/S15328007SEM1001_6
73 Credamo Platform User Manual: Chapter Four - Questionnaire Publish, Section Two - Release Channels. Available from: https://www.credamo.world/#/help/web/4?page_id=66.
74 Mark A.B. Woo G. Kim Sora K. Meehee C. Purchasing wine online: the effects of social influence, perceived usefulness, perceived ease of use, and wine involvement J. Hospit. Market. Manag. 25 7 2015 10.1080/19368623.2016.1115382
75 Tandon A. Dhir A. Kaur P. Kushwah S. Salo J. Behavioral reasoning perspectives on organic food purchase Appetite 154 2020 10.1016/j.appet.2020.104786
76 Daeseong A. Ji S. Jan I. Ullah, investigating the determinants and barriers of purchase intention of innovative new products Sustainability 13 2 2021 10.3390/su13020740
77 Rob L. Kirsten R. Ben W. Health, vulnerability, and energy: assessing energy markets and consumer agency in New Zealand Energy Res. Social Sci. 19 2016 10.1016/j.erss.2016.05.021
78 Bonn M.A. Kim W.G. Kang S. Cho M. Purchasing wine online: the effects of social influence, perceived usefulness, perceived ease of use, and wine involvement J. Hospit. Market. Manag. 25 7 2016 841 869 10.1080/19368623.2016.1115382
79 Ki C.W. Cuevas L.M. Chong S.M. Lim H.J. Influencer marketing: social media influencers as human brands attaching to followers and yielding positive marketing results by fulfilling needs J. Retailing Consum. Serv. 55 2020 10.1016/j.jretconser.2020.102133
80 Sharma R. Dhir A. Talwar S. Kaur P. Int. J. Hospit. Manag. 96 2021 10.1016/j.ijhm.2021.102977
81 Hair J.F. Ringle C.M. Sarstedt M. Partial least squares structural equation modeling: rigorous applications, better results and higher acceptance Long. Range Plan. 46 1–2 2013 1 12 10.1016/j.lrp.2013.08.016
82 Hair J. Howard M. Nitzl C. Assessing measurement model quality in PLS-SEM using confirmatory composite analysis J. Bus. Res. 109 2020 101 110 10.1016/j.jbusres.2019.11.069
83 Ketchen J. David A primer on partial least squares structural equation modeling (PLS-SEM) Long. Range Plan. 46 1–2 2013 184 185 10.1016/j.lrp.2013.01.002
84 Fornell C. Larcker D.F. Structural equation models with unobservable variables and measurement error : algebra and statistics J. Market. Res. 18 1 1981 39 50 10.2307/3150980
85 Sarstedt M. Ringle C.M. Hair J.F. Partial Least Squares Structural Equation Modeling 2021 1 47
86 Shela V. Ramayah T. Aravindan K.L. Ahmad N.H. Alzahrani A.I. Run! This road has no ending! A systematic review of PLS-SEM applicati on in strategic management research among developing nations Heliyon 9 12 2023 e22476 10.1016/j.heliyon.2023.e22476
87 Lohmöller J.-B. Latent variable path modeling with partial least squares Physica Heidelberg 286 2013
88 Cho G. Hwang H. Sarstedt M. Ringle C.M. Cutoff criteria for overall model fit indexes in generalized structured component analysis Journal of Marketing Analytics 8 4 2020 189 202 10.1057/s41270-020-00089-1
89 Henseler J. Dijkstra T. Sarstedt M. Ringle C. Diamantopoulos A. Straub D. Jr D. Hair J. Hult G.T.M. Calantone R. Common beliefs and reality about PLS: comments on rönkkö and evermann (2013) Organ. Res. Methods 17 2014 182 209 10.1177/1094428114526928
90 Ali F. Kim W.G. Ryu K. The effect of physical environment on passenger delight and satisfaction: moderating effect of national identity Tourism Manag. 57 2016 213 224 10.1016/j.tourman.2016.06.004
91 Fan Y. Liu X. Exploring the role of AI algorithmic agents: the impact of algorithmic decision autonomy on consumer purchase decisions Front. Psychol. 13 2022 10.3389/fpsyg.2022.1009173
92 Luo H. Cheng S. Zhou W. Ieee. The factors influencing sales in online celebrities' live streaming IEEE International Conference on Information Communication and Software Engineering (ICICSE) Sichuan Inst Elect 2021 PEOPLES R CHINA Chengdu
93 Wang Z. Zhu J. Xu Y. Kim D. Williams D. Cross-platform information flow and multilingual text analysis: a comparative study of weibo and twitter through deep learning Computational Communication Research 5 1 2023 1 10.5117/CCR2023.1.8.WANG
94 Chieh L.J. Li C. Hengrui Z. Exploring the adoption decisions of mobile health service users: a behavioral reasoning theory perspective Ind. Manag. 123 8 2023 10.1108/IMDS-11-2022-0682
95 La C.K. Uriarte C. Elhabashy M. Menkevich M. Maxton O. Awadalla J. Tan A.S.L. Moran M.B. Chen-Sankey J. Exploring the Influence of E-Cigarette Ad Features on Perceived Product Appeal and Use Interest Among Young Adults of Varying Tobacco-Use Behaviors 2023 Nicotine & tobacco research : official journal of the Society for Research on Nicotine and Tobacco 10.1093/ntr/ntad150
