Exploring User Retention in WeChat E-Commerce for SME Retailers: Perspective of Perceived Quality and Privacy Calculus

Abstract

The rapid development of e-commerce has brought unprecedented opportunities for small and medium-sized enterprises (SMEs), particularly those in the retail sector. WeChat e-commerce, which combines the advantages of social media and e-commerce, offers SME retailers an accessible and efficient e-commerce solution. As market competition intensifies, user retention has become crucial for their success in WeChat e-commerce. This study extended the Expectation Confirmation Model (ECM) by incorporating perceived quality and privacy calculus to examine users’ continuance intention towards WeChat e-commerce for SME retailers. A total of 694 valid responses were collected from users with prior experience using WeChat e-commerce offered by SME retailers, and the proposed model was validated using structural equation modeling. The results indicated that the trade-off between privacy concerns and perceived benefits significantly affected users’ intention to continue using WeChat e-commerce. Moreover, the information quality and service quality dimensions were found to directly or indirectly influence this continuance intention towards WeChat e-commerce via privacy concerns, perceived benefits, and confirmation. In conclusion, this study provides insights for further research on continuance intention in WeChat e-commerce and suggestions for SMEs to formulate e-commerce strategies.

1. Introduction

With the rapid development of the digital economy, e-commerce has become an essential part of modern business [1]. As an Internet-based commercial activity [2], e-commerce not only restructures business models but also profoundly influences consumer behaviors [3]. Small and medium-sized enterprises (SMEs) play a crucial role in the national economy [4], but they are often overshadowed by large enterprises due to limitations in financial resources, brand recognition, and technological capabilities [5]. Research suggests that e-commerce has significant potential to promote the growth of SMEs [6], allowing them to overcome traditional barriers, access new markets, increase market share, and boost brand awareness [7]. Especially during the COVID-19 pandemic, physical retail stores faced severe restrictions, making e-commerce essential for SME retailers to sustain and grow their businesses [8,9]. In this context, numerous SME retailers have shifted to online operations, engaging in business activities through third-party e-commerce platforms or self-built e-commerce sites [10].

In recent years, social commerce, which integrates social media and e-commerce, has emerged as a new trend in the development of e-commerce [11,12]. As China’s leading social platform, WeChat boasts a substantial user base and comprehensive business functionalities [13,14], serving as an efficient channel for retailers to promote products and maintain customer relationships [15]. Specifically, SME retailers can engage in diversified e-commerce activities such as product display, online sales, marketing campaigns, and customer management through WeChat features like official accounts, mini-programs, and video accounts [16]. Although many scholars have studied the application of e-commerce in SMEs, most of them focus on firm-level adoption, performance, and strategy [2,17,18], and research from the consumer perspective remains limited; existing studies mainly focus on traditional e-commerce platforms [9], with insufficient attention paid to social commerce. Therefore, this study explores users’ continuance intention towards WeChat e-commerce, providing insights for SMEs’ sustainable e-commerce development.

To ensure the retention of SME retailers’ WeChat e-commerce users, it is necessary to identify the key factors influencing their continuance intention. Prior studies have shown that perceived quality influences various behavioral outcomes, including loyalty [19], satisfaction and trust [20], and usage intentions [13]. In the e-commerce context, quality perception has been found to significantly affect purchase intentions [21], while different quality dimensions are crucial for e-commerce success [22]. Hence, perceived quality dimensions can serve as critical determinants of users’ continuance intention in WeChat e-commerce. Furthermore, the virtual transaction environment constantly exposes consumers to inherent risks, especially monetary loss and privacy disclosure [23]. To navigate these risks, users engage in privacy calculus—a cognitive process involving cost-benefit assessments that reflect a trade-off between potential gains and losses [24]. While privacy calculus has been well-established in online contexts involving information disclosure [25], its mechanisms among WeChat e-commerce users remain underexplored. Based on the above analysis, this study speculates that perceived quality and privacy calculus can significantly impact users’ continuance intention in WeChat e-commerce. The following research questions are proposed: (1) Does perceived quality affect users’ continuance intention in WeChat e-commerce for SME retailers? And by what? (2) Does privacy calculus affect users’ continuance intention in WeChat e-commerce for SME retailers? And by what?

In information systems (IS) research, the Expectation-Confirmation Model (ECM) is recognized as a prominent theoretical framework for examining post-adoption behavior [26]. Originally proposed by Bhattacherjee [27], ECM explains users’ continuance intention towards IS through four core constructs: confirmation, perceived usefulness, satisfaction, and continuance intention [28]. Given the limitations of perceived usefulness [29], this study replaces it with the broader construct of perceived benefits. Previous studies have further extended this model by incorporating context-specific variables such as habit [30], social factors [31] and trust [32]. Therefore, this study integrates perceived quality, privacy calculus, and ECM to construct a research model to explore users’ continuous intention in SME retailers’ WeChat e-commerce.

For better readability and coherence, the paper is organized as follows: Section 2 introduces the theoretical foundations, and Section 3 outlines our hypotheses. Section 4 details the methodology and research design, while Section 5 presents the data analysis and hypotheses testing results. Section 6 provides a comprehensive discussion of our findings. Section 7 outlines the implications of this study and concludes with a summary of the limitations and suggestions for future research directions.

2. Theoretical Background

2.1. E-Commerce for SMEs

E-commerce encompasses buying and selling transactions conducted via electronic means such as the Internet [2,33]. It not only facilitates purchasing and selling online but also promotes the optimization of the entire transaction process [23]. Compared to traditional commerce, e-commerce offers many advantages, including greater flexibility, enhanced market expansion, lower cost structure, faster transactions, richer product selection, increased convenience, and opportunities for customization [7]. As a crucial driver for global economic development, SMEs play an irreplaceable role in fostering technological innovation, offering jobs, and maintaining social stability [34]. However, due to their small scale and limited resources, SMEs not only face pressure from large enterprises in fierce market competition [35] but also exhibit a distinct vulnerability to public crises, as was evident during the COVID-19 pandemic [36]. Moreover, with consumers increasingly shifting towards online shopping [9], retail companies must proactively embrace digital transformation to enhance their competitive edge.

E-commerce has generated unprecedented opportunities for SMEs, enabling them to quickly enter new markets, expand their customer base, and compete equally with large enterprises [7,23]. It not only helps SME retailers disseminate product information and enhance brand visibility but also precisely matches consumers’ demands by analyzing their purchasing activities and demographics [3]. SME retailers often choose to enter third-party e-commerce platforms such as Amazon and Taobao to quickly obtain traffic and orders at a low cost, while enterprises with more substantial capital and technical strength typically develop independent e-commerce websites or platforms as a long-term strategy [10]. In recent years, social commerce, combining social media and e-commerce, has emerged as a new trend in the development of e-commerce [12,37]. WeChat e-commerce, as a prominent example, provides a low-barrier and efficient e-commerce solution for SME retailers. Harnessing WeChat tools such as official accounts, mini-programs, WeChat groups, and video accounts [38], SME retailers can closely interact with consumers and provide comprehensive e-commerce services, including marketing promotion, product display, transaction payment, and after-sales service.

Recognizing the enormous potential benefits that E-commerce can bring to SMEs, many scholars have paid considerable attention to its study within this context. Existing research has primarily examined enterprise-level aspects, such as e-commerce adoption [2], performance impact [17], diversification strategy [18], and success factors [39]. However, the consumer perspective remains relatively underexplored. Furthermore, research has prominently focused on general e-commerce research across industry sectors, often lacking consideration of specific industries [2]. Meanwhile, most studies focused on traditional e-commerce platforms or websites [9], with little research on social commerce [40]. However, in the highly competitive environment of WeChat e-commerce, retaining users is still a significant challenge for SME retailers. Therefore, this study explored users’ continuance intention in SME retailers’ WeChat e-commerce from the consumer’s perspective.

2.2. Perceived Quality

Perceived quality is crucial for the success of any firm [41,42], including SMEs engaged in e-commerce. It refers to consumers’ evaluation of a service’s overall excellence or superiority [43] and focuses more on perception than actual quality, reflecting reliability and dependability [44]. Prior studies have established its significant impact on various behavioral outcomes, including loyalty [19], satisfaction and trust [20], usage intentions [13], and purchase decisions [41]. In e-commerce environments lacking physical interaction, perceived quality becomes particularly critical in shaping consumer behaviors. Based on this understanding, this study examines how perceived quality influences users’ continuance intention towards WeChat e-commerce for SME retailers.

Perceived quality encompasses more than just “product quality”; it involves subjective assessments of specific product or service offerings [45], necessitating a disaggregated approach for detailed analysis. Previous research examining website adoption intentions employed a framework comprising information quality (IQ), service quality (SQ), and system quality (SysQ) [46]. Given the two key characteristics of e-commerce, (1) information technology and (2) commercial service, this study focuses on information quality and service quality to evaluate perceived quality in SME retailers’ WeChat e-commerce. Prior scholars have highlighted the importance of information quality [47,48]. Especially in online context, information quality is a vital component of high-quality e-commerce services and positive user experiences [22,37]. Besides that, service quality serves as a crucial concept in customer service [49] and is consistently related to enhancing users’ continuance intention [50]. Recent IS scholars have further confirmed the importance of service quality [42,51], recognizing it as a key factor in e-commerce success [52]. Based on the relevance and importance of information quality and service quality, they were incorporated in our model to better explore users’ continuance intention towards WeChat e-commerce for SME retailers.

2.3. Privacy Calculus

With the rapid development of new technology, the privacy of Internet users can be compromised in various ways [53], while a company’s success and its service quality largely depend on its ability to collect and analyze personal information [54]. In this context, the privacy calculus theory has become a fundamental framework for understanding users’ privacy-related behaviors [14,55]. Privacy calculus refers to individuals’ cost-benefit analysis when weighing the trade-offs between the costs of providing personal information and the benefits of information disclosure [56,57]. Research has established that a consumer’s decision to disclose personal information is based on the privacy calculus, which includes both drivers and disincentives that influence privacy decisions and helps investigate how competitive factors interact [25]. Individuals often rely on experience, intuition, or perception to assess these costs and benefits [24].

The privacy calculus theory has been widely used in diverse online contexts involving self-disclosure and privacy protection [58]. For instance, Li et al. [59] applied this theory to understand information disclosures and privacy sensitivity on short-form video platforms, and Gouthier et al. [60] explored the relationship between privacy concerns, information sensitivity, and personalized services in e-commerce. The core of this theory lies in individuals’ cost-benefit trade-off regarding personal information, where disclosure occurs when perceived benefits outweigh privacy costs [61]. Given the inherent difficulty in precisely quantifying privacy costs, researchers commonly use privacy concerns—reflecting users’ fears of potential losses from disclosure or lack of control—as a practical proxy [24,62]. In the context of WeChat e-commerce, users are willing to disclose information and continue using the platform if their perceived benefits outweigh their privacy concerns.

2.4. Expectation-Confirmation Model

While initial adoption is crucial for an information system (IS), long-term success ultimately depends on users’ continuous use [63]. Building on the Expectation Confirmation Theory (ECT) [64] and the Technology Acceptance Model (TAM), Bhattacherjee [27] proposed the Expectation-Confirmation Model (ECM) to explain IS continuance intention. Unlike traditional technology acceptance theories (e.g., TAM and UTAUT) that primarily focus on initial adoption, ECM systematically investigates post-adoption continuance through an iterative expectation-confirmation process [26]. ECM comprises four core constructs: confirmation, perceived usefulness, satisfaction, and continuance intention [27]. Confirmation refers to users’ assessments of the congruence between their actual performance and initial expectations regarding an IS and services [65], while satisfaction represents the post-use evaluation of the IS [66]. Considering the limitations of perceived usefulness [29], this study replaces it with the broader concept of perceived benefits, which is defined as users’ assessments of benefits associated with using WeChat e-commerce. Drawing on Huang’s [67] definition, this study conceptualizes continuance intention as users’ willingness to continue using SME retailers’ WeChat e-commerce.

The effectiveness of the ECM has been validated across various online contexts, including social media [68], e-learning [69], wearable technology [70], smart fitness [71], and mobile health [72]. While the model has also been applied to explain user behavior in e-commerce and online transactions [26,73], its application to continuance intention in social commerce, particularly WeChat e-commerce, remains underexplored. Prior studies have incorporated various context-specific extended variables into the ECM, such as habit [30] in IS continuance, social factors in Web 2.0 apps usage [31], trust in crowdsourcing contest platforms [32], and perceived risk in online impulse buying [74]. Herein, several variables were incorporated into the ECM, such as perceived quality and privacy factors, and perceived usefulness was replaced with perceived benefits to better analyze users’ continuance intention towards WeChat e-commerce for SME retailers.

3. Hypothesis Development

3.1. Information Quality and Service Quality

In the context of e-commerce, information quality refers to users’ overall assessment of website content [22], including core business information (e.g., product details) and supplementary derivative information (e.g., user reviews). Academic research has identified various dimensions of information quality [75], with completeness and accuracy recognized as core elements [76]. Specifically, completeness reflects the extent to which comprehensive information is provided to satisfy users’ needs [46]. Within WeChat e-commerce, systematically presenting critical elements such as products, services, and user reviews can reduce information asymmetry and enhance decision-making efficiency. Meanwhile, accuracy is defined as the extent to which information is correct and unambiguous [77]. The accurate descriptions in WeChat e-commerce help consumers avoid fraudulent messages and accurately evaluate products and services [37]. This study prioritized these two dimensions given their critical role in shaping information quality in WeChat e-commerce. Dağhan and Akkoyunlu [78] integrated IQ with the ECM in their analysis of online learning services, demonstrating positive correlations between IQ and both confirmation and perceived value. Other researchers suggest that IQ can reduce perceived risks and uncertainties when taking advice from unfamiliar sources [48] and mitigate privacy concerns [77]. In SME retailers’ WeChat e-commerce, accuracy and completeness are closely related to ECM and privacy concerns. Thus, we proposed the following hypotheses:

H1a. 

Accuracy negatively affects privacy concerns regarding WeChat e-commerce for SME retailers.

H1b. 

Accuracy positively affects perceived benefits of WeChat e-commerce for SME retailers.

H1c. 

Accuracy positively affects confirmation of WeChat e-commerce for SME retailers.

H2a. 

Completeness negatively affects privacy concerns regarding WeChat e-commerce for SME retailers.

H2b. 

Completeness positively affects perceived benefits of WeChat e-commerce for SME retailers.

H2c. 

Completeness positively affects confirmation of WeChat e-commerce for SME retailers.

Service quality refers to customers’ overall evaluations of the excellence of service provided via a website [46], with this study focusing on WeChat e-commerce services offered by SMEs. Among the various models for measuring online service quality, the E-S-QUAL framework [79] stands out with its four core dimensions: efficiency, system availability, fulfillment, and privacy. While other scholars have proposed different dimensions, efficiency remains the core element [80], which we prioritized in this study. Considering the context of WeChat e-commerce, this study innovatively introduces another dimension, procedural justice, which highlights users’ perceptions of fairness in processes [81,82]. Within WeChat e-commerce, users rely on procedures to discover, acquire, and evaluate goods/services, making procedural justice vital for service quality perception. This expansion meets users’ demand for process transparency and enhances the evaluation system for service quality in social commerce.

Previous studies have confirmed the rationality of combining SQ with ECM, indicating the positive impact of SQ on confirmation [83] and perceived value [50]. Since efficiency has been shown to directly enhance e-government services [84], it is reasonable to infer that efficiency also positively influences the perceived value of WeChat e-commerce. Furthermore, users’ perceptions and behaviors are partly influenced by procedural and outcome justice, particularly in decision-making and conflict contexts [81]. Procedural justice can amplify users’ sense of gain by shaping their perception of fairness toward the process [82]. Moreover, perceptions of justice often negatively impact perceptions of privacy [85]. Efficient services can reduce uncertainties [86], alleviating privacy concerns in WeChat e-commerce transactions. Thus, we proposed the following hypotheses:

H3a. 

Procedural justice negatively affects privacy concerns of WeChat e-commerce for SME retailers.

H3b. 

Procedural justice positively affects perceived benefits of WeChat e-commerce for SME retailers.

H3c. 

Procedural justice positively affects confirmation of WeChat e-commerce for SME retailers.

H4a. 

Efficiency negatively affects privacy concerns of WeChat e-commerce for SME retailers.

H4b. 

Efficiency positively affects perceived benefits of WeChat e-commerce for SME retailers.

H4c. 

Efficiency positively affects confirmation of WeChat e-commerce for SME retailers.

3.2. Privacy Concerns and Perceived Benefits

Previous research has indicated that perceived benefits significantly influence the adoption of technologies and related services [87,88]. It represents the “benefit” component of the “privacy calculus” concept [61]. In this study, the perceived benefits of WeChat e-commerce include tangible advantages like product diversity and intangible benefits like improved user experience. The conceptualization of perceived benefits is broader than the perceived usefulness [29], which is conducive to explore customers’ willingness to continue using WeChat e-commerce. Current evidence suggests that perceived benefits positively affect attitudes toward online food delivery services [89]. WeChat e-commerce also offers convenience and cost savings, which are factors likely to increase users’ satisfaction. Some scholars have demonstrated that perceived benefits significantly affect the adoption of e-commerce technology [90]. When consumers experience various benefits from WeChat e-commerce, they exhibit a greater propensity to continue using it. Thus, we proposed the following hypotheses:

H5a. 

Perceived benefits positively affect satisfaction with WeChat e-commerce for SME retailers.

H5b. 

Perceived benefits positively affect continuance intention towards WeChat e-commerce for SME retailers.

Privacy concerns arise from the potential negative consequences associated with the misuse of personal information that service providers may intentionally collect, disclose, transmit, or sell without the consumers’ knowledge [91]. Privacy concerns reflect an individual’s apprehension about sharing and exchanging personal information [92]. The role of privacy concerns in the online environment cannot be overstated [53], constituting a crucial factor influencing many online behaviors [55]. This is especially pronounced on the WeChat e-commerce platform, where users frequently provide personal data for online transactions [62]. The principle of “privacy calculus” suggests that, when consumers prioritize privacy, their decision-making in WeChat e-commerce will be influenced, affecting their perception of benefits [93]. In addition, current evidence suggests that privacy concerns negatively impact the intention and actual disclosure of personal information [58], and consumers with high privacy concerns are more likely to boycott companies that do not protect their data [54]. Within WeChat e-commerce, inadequate disclosure of information may hinder service provision, ultimately affecting users’ continuance intention. Hence, we posited the following hypotheses:

H6a. 

Privacy concerns negatively affect perceived benefits of WeChat e-commerce for SME retailers.

H6b. 

Privacy concerns negatively affect continuance intention towards WeChat e-commerce for SME retailers.

3.3. ECM and Continuance Intention in WeChat E-Commerce

Within the ECM framework, confirmation significantly impacts perceived usefulness and satisfaction and is considered an important motivation for users’ intention to continue using systems [4]. Mamun et al. [29] highlighted the positive relationship between user confirmation and the perceived benefits of IS usage. Wu et al. [72] found that confirmation of initial expectations for mobile health services exhibits a positive correlation with satisfaction. With WeChat e-commerce, customer satisfaction and perceived benefits are likely to increase if the actual service performance meets their expectations. Satisfaction is the central variable in ECM [78] and strongly predicts continuance intention [71]. Previous studies have indicated that satisfaction positively influences customers’ willingness to continue using e-commerce, including community-based O2O e-commerce [94] and e-retail environments [95], and satisfied customers demonstrate a strong willingness to continue using services [73]. Thus, we proposed the following hypotheses:

H7a. 

Confirmation positively affects perceived benefits of WeChat e-commerce for SME retailers.

H7b. 

Confirmation positively affects satisfaction with WeChat e-commerce for SME retailers.

H8. 

Satisfaction positively affects continuance intention towards WeChat e-commerce for SME retailers.

Therefore, based on the hypothesis presented above and previous studies, we developed the following conceptual model of the study (Figure 1).

4. Methodology

4.1. Measurements

All measurements were developed from the previous literature and used a seven-point Likert scale ranging from “strongly disagree = 1” to “strongly agree = 7”. The original English questionnaire was translated into Chinese. To enhance its accuracy and validity, we invited two experts to review and improve it, resulting in the initial questionnaire used for data collection.

The questionnaire consists of two parts. The first part collects the basic information of WeChat e-commerce users of SME retailers, including gender, age, education level, duration, and frequency of using WeChat e-commerce. The second section comprises nine constructs with a total of 39 items. In this study, questionnaire items were adapted from previous studies to ensure contextual relevance of WeChat e-commerce for SME retailers to measure accuracy [77,78], completeness [77,96], procedural justice [97], efficiency [84], privacy concerns [98,99], perceived benefits [100,101], confirmation [74,102], satisfaction [74], and continuance intention [13].

Both pretest and pilot tests were used to validate the questionnaire. Firstly, three postgraduate students provided suggestions for improvement through a pretest. Subsequently, 50 respondents with rich experience using SME retailers’ WeChat e-commerce were selected for a pilot test. Based on their feedback, further adjustments were made to the question content to ensure respondents better understand and answer the questionnaire.

4.2. Data Collection and Sampling

Data were collected from Sojump (www.sojump.com), one of China’s most popular online survey companies, to facilitate the recruitment of qualified respondents from their vast database. To avoid data collection bias, we selected users with rich experience using SME retailers’ WeChat e-commerce as subjects. WeChat e-commerce for SME retailers utilizes diversified forms, including mini-program malls, official account promotions, and WeChat group/moment ads. Although these forms have their characteristics regarding function and user experience, they collectively foster the development of WeChat e-commerce for SMEs. The survey was administered between May 2022 and August 2022, yielding a total of 730 responses. A total of 34 were excluded due to incomplete or identical responses. Finally, 696 valid questionnaires were obtained for subsequent data analysis.

Table 1. Demographics of respondents.

Characteristics	Frequency	Percent (%)
1. Gender
Male	207	29.7%
Female	489	70.3%
2. Age
Under 18	7	1.0%
18–25	255	36.6%
26–33	296	42.5%
34–41	90	12.9%
Over 41	48	6.9%
3. Education level
Certificate or below	14	2.0%
High school	45	6.5%
Undergraduate degree	565	81.2%
Master	64	9.2%
Doctor degree	8	1.1%
4. Daily time spent on WeChat
Under 2 h	68	9.8%
2 h–3 h	159	22.8%
3 h–4 h	243	34.9%
4 h–5 h	139	20.0%
5 h–6 h	71	10.2%
Over 6 h	16	2.3%
5. Frequency of using WeChat
Under 4 times a day	20	2.9%
4–6 times a day	101	14.5%
7–9 times a day	131	18.8%
10–12 times a day	141	20.3%
Over 13 times a day	303	43.5%
6. Experience of using WeChat
Under 2 years	8	1.1%
2–3 years	37	5.3%
3–4 years	88	12.6%
4–5 years	100	14.4%
Over 5 years	463	66.5%
7. Frequency of using WeChat e-commerce
Under 2 times a day	70	10.1%
2–3 times a day	214	30.7%
4–5 times a day	182	26.1%
6–7 times a day	79	11.4%
Over 7 times a day	151	21.7%
8. Experience of using WeChat e-commerce
Under 2 years	47	6.8%
2–3 years	133	19.1%
3–4 years	155	22.3%
4–5 years	143	20.5%
Over 5 years	218	31.3%

presents a detailed overview of the demographics of the respondents. Out of the 696 respondents, 29.7% were male, and 70.3% were female. Analysis of the age distribution revealed that 36.6% were in the 18–25 age group, and 42.5% were in the 26–33 age group, collectively representing 79.1% of the total sample. This demographic composition indicated that the study findings primarily reflect the preferences and intentions of young female users. In terms of WeChat usage, over 90% of the respondents reported having more than 3 years of experience with WeChat. Notably, 74.1% of respondents reported having more than 3 years of experience in using WeChat e-commerce of SME retailers. Regarding the frequency of use, nearly 90% of users reported using WeChat e-commerce for SME retailers at least 2–3 times daily.

5. Data Analysis and Results

We employed partial least squares structural equation modeling (PLS-SEM) to test our research model. Compared to covariance-based SEM, PLS does not require special data distribution or multivariate normality for large samples [103]. It is more advantageous in incremental studies, i.e., the construction of new metrics and structural paths, especially in IS research [104]. Thus, PLS-SEM is appropriate for this study, and data analysis was conducted using Smart PLS 3.0.

5.1. Measurement Model

The measurement model underwent rigorous testing for reliability and validity. First, reliability was assessed by testing Cronbach’s α coefficient. As shown in

Table 2. Consistency and reliability test.

Constructs	Indicator Reliability	Consistency Reliability		Convergent Validity
	Outer Factor Loading	Cronbach’s Alpha	Composite Reliability	Average Variance Extracted (AVE)
Accuracy	0.832~0.870	0.812	0.889	0.727
Completeness	0.796~0.805	0.720	0.842	0.640
Procedural justice	0.811~0.824	0.838	0.892	0.673
Efficiency	0.810~0.818	0.746	0.855	0.663
Privacy concerns	0.909~0.948	0.947	0.962	0.863
Perceived benefits	0.809~0.829	0.755	0.860	0.672
Confirmation	0.766~0.820	0.808	0.874	0.635
Satisfaction	0.804~0.854	0.852	0.900	0.693
Continuance intention	0.792~0.823	0.820	0.881	0.649

, the value of Cronbach’s α coefficient ranged between 0.720 (completeness) and 0.947 (privacy concerns), greater than the set threshold of 0.7, indicating that the measurement model had good reliability. The reliability was further validated by the fact that the composite reliability (CR) for all constructs was greater than 0.8, which exceeded the minimum threshold of 0.7 [105]. According to the recommendations of Hair et al. [106] and Kline [107], in structural equation modeling (SEM), particularly for higher-order models, the ideal threshold for outer factor loadings should not be lower than 0.7. Based on this criterion, seven items (PJ5, PJ6, PC1, PB3, SA1, CI1, and CI3) with factor loadings below this threshold were removed. As indicated in Table 2, the factor loadings of all variables were significant after optimization. Regarding convergent validity, AVE values were calculated [108]. Convergent validity requires that AVE values should be greater than 0.5. As shown in Table 2, the AVE values ranged from 0.635 (confirmation) to 0.863 (privacy concerns), all of which exceeded the suggested threshold, indicating good convergent validity.

Previous research has found that discriminant validity can be assessed using the Fornell–Larcker criterion [105] and the Heterotrait-Monotrait (HTMT) ratio [109], both of which have proven its strength in research. Therefore, the Fornell–Larker criterion and the HTMT ratio were chosen to validate discriminant validity in this study. As shown in

Table 3. Discriminant validity of constructs: Fornell–Larcker.

	AC	CO	PJ	EF	PC	PB	CON	SAT	CI
AC	0.853
CO	0.670	0.800
PJ	0.658	0.644	0.820
EF	0.592	0.540	0.557	0.814
PC	−0.213	−0.246	−0.207	−0.062	0.929
PB	0.577	0.501	0.548	0.621	−0.022	0.819
CON	0.649	0.593	0.658	0.595	−0.285	0.638	0.797
SAT	0.671	0.659	0.708	0.626	−0.229	0.684	0.734	0.833
CI	0.676	0.652	0.628	0.595	−0.173	0.657	0.687	0.749	0.806

Note 1: AC = Accuracy; CO = Completeness; PJ = Procedural justice; EF = Efficiency; PC = Privacy concerns; PB = Perceived benefits; CON = Confirmation; SAT = Satisfaction; CI = Continuance intention. Note 2: Diagonal and bold elements are the square root of AVE between the constructs and their indicators. Non-diagonal elements are correlations between constructs.

, the square root of the AVE for each construct was greater than the correlation with any other construct [110], satisfying the requirement for discriminant validity.

In addition, when using the HTMT ratio for the discriminant validity test, it is required that the HTMT ratio is below the threshold of 0.85 (or 0.9) [109,111]. As shown in

Table 4. Discriminant validity of constructs: HTMT.

	AC	CO	PJ	EF	PC	PB	CON	SAT	CI
AC
CO	0.876
PJ	0.798	0.828
EF	0.761	0.732	0.702
PC	0.241	0.298	0.230	0.074
PB	0.737	0.673	0.686	0.826	0.030
CON	0.801	0.776	0.797	0.765	0.324	0.816
SAT	0.804	0.839	0.836	0.781	0.256	0.848	0.881
CI	0.829	0.848	0.757	0.759	0.193	0.833	0.843	0.894

Note 1: AC = Accuracy; CO = Completeness; PJ = Procedural justice; EF = Efficiency; PC = Privacy concerns; PB = Perceived benefits; CON = Confirmation; SAT = Satisfaction; CI = Continuance intention.

, the maximum HTMT value was 0.894, lower than the threshold of 0.9, indicating good discriminant validity. Therefore, all the measurement models in this study are validated by the above results.

5.2. Structural Model

In the structural model analysis, we examined the significance of the path coefficients and the R² variance of the associated construct. This study also evaluated the t-statistics of the standardized path coefficients to verify whether each hypothesis was supported. Figure 2 reveals all the paths of the structural model and the coefficients for each relevant structure.

A bootstrap analysis of 5000 resamples of the structural model estimated by PLS-SEM was conducted, with the statistical results summarized in

Table 5. Hypothesis testing results.

Hypothesis		Path	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics	Path Coefficient	Result
H1	H1a	AC → PC	−0.110	−0.112	0.056	1.956	−0.110	Not
	H1b	AC → PB	0.142	0.142	0.056	2.539	0.141	Supported
	H1c	AC → CON	0.228	0.229	0.046	4.962	0.229	Supported
H2	H2a	CO → PC	−0.207	−0.208	0.052	3.949	−0.208	Supported
	H2b	CO → PB	0.034	0.034	0.047	0.730	0.035	Not
	H2c	CO → CON	0.108	0.109	0.036	2.969	0.107	Supported
H3	H3a	PJ → PC	−0.106	−0.107	0.050	2.109	−0.106	Supported
	H3b	PJ → PB	0.088	0.089	0.059	1.505	0.089	Not
	H3c	PJ → CON	0.335	0.335	0.044	7.610	0.333	Supported
H4	H4a	EF → PC	0.172	0.171	0.055	3.127	0.172	Supported
	H4b	EF → PB	0.271	0.270	0.050	5.405	0.270	Supported
	H4c	EF → CON	0.210	0.209	0.040	5.200	0.211	Supported
H5	H5a	PB → SAT	0.363	0.361	0.039	9.215	0.364	Supported
	H5b	PB → CI	0.286	0.285	0.038	7.493	0.286	Supported
H6	H6a	PC → PB	0.154	0.150	0.029	5.390	0.154	Supported
	H6b	PC → CI	−0.052	−0.052	0.026	2.014	−0.051	Supported
H7	H7a	CON → PB	0.348	0.345	0.054	6.396	0.349	Supported
	H7b	CON → SAT	0.502	0.504	0.033	15.078	0.502	Supported
H8		SAT → CI	0.541	0.541	0.038	14.322	0.541	Supported

Note: AC = Accuracy; CO = Completeness; PJ = Procedural justice; EF = Efficiency; PC = Privacy concerns; PB = Perceived benefits; CON = Confirmation; SAT = Satisfaction; CI = Continuance intention.

. Accuracy significantly affected perceived benefits (t = 2.539, p < 0.05) and confirmation (t = 4.962, p < 0.001), supporting H1b and H1c. However, accuracy (t = 1.956, p > 0.05) did not significantly affect privacy concerns, failing to support H1a. Completeness significantly influenced both privacy concerns (t = 3.949, p < 0.001) and confirmation (t = 2.969, p < 0.01), supporting H2a and H2c. Conversely, H2b, which posited a significant effect of completeness on perceived benefits, was not supported (t = 0.730, p > 0.05). The effects of procedural justice on privacy concerns (t = 2.109, p < 0.05) and confirmation (t = 7.610, p < 0.001) proposed by H3 were both supported, but the relationship proposed in H3b between procedural justice and perceived benefits was not confirmed (t = 1.505, p > 0.05). Besides that, the effects of efficiency on privacy concerns (t = 3.127, p < 0.01), perceived benefits (t = 5.405, p < 0.001), and confirmation (t = 5.200, p < 0.001) were all statistically significant, thus supporting H4. Finally, H5 examined the relationship between perceived benefits, satisfaction, and continuance intention; H6 assessed the relationship between privacy concerns, perceived benefits, and continuance intention; H7 examined the relationship between confirmation, perceived benefits, and satisfaction; H8 posited a positive effect of satisfaction on continuance intention. All the above hypotheses were supported.

Next, the explanatory power of the structural model was assessed by the R² value. The recommended thresholds for R² are categorized as strong (0.67), medium (0.33), and weak (0.19) [110]. In the present study, the R² values for perceived benefits, confirmation, satisfaction, and continuance intention were 0.531, 0.553, 0.615, and 0.6, respectively, indicating that the model achieved considerable explanatory power [112]. Although the predictive validity of privacy concerns was low (0.078), the significance of the coefficient p-value was still supported.

5.3. Common Method Variance

Common method variance (CMV) represents a potential concern for this study, given that data were collected through a questionnaire where respondents simultaneously answered measurement items on both explanatory and dependent variables and were likely to give consistent answers to unrelated questions [113]. Two methods were employed to examine CMV: Harman’s one-factor test and the latent method factor (LMF). First, the results of Harman’s one-factor test showed that the maximum variance explained by a single factor was 38.835%, below the 50% threshold [114]. Therefore, no single factor could explain most of the variance, indicating a low likelihood of CMV.

Second, following the methodology reported by Lindell and Whitney [115], we performed an analysis using the marker variable technique, incorporating a theoretically irrelevant variable as a marker variable. The potential threat of CMV to this study was validated by evaluating the correlation of marker variables with the study model. The results showed that the primary constructs of the study had a low mean correlation with the marker variable. After excluding marker variables, none of the previous significant correlations became insignificant, indicating that CMV did not affect the data in this paper. Therefore, based on the results of both tests, CMV does not pose a threat to this study.

6. Discussion

Our study explored users’ continuance intention towards SME retailers’ WeChat e-commerce by constructing a research model that integrated perceived quality, privacy calculus, and ECM perspectives to examine their cognitive-emotional trade-offs. The results indicated the strong explanatory power of our research model, with most hypotheses being supported. Firstly, perceived benefits, confirmation, and satisfaction significantly influenced continuance intention, consistent with the literature [68,71]. Given the limitations of perceived usefulness [29], we replaced it with a broader conceptualization of perceived benefits. The results confirmed the rationality and feasibility of the substitution. The WeChat e-commerce services provided by SME retailers tend to be more diversified and personalized, prompting consumers to consider not just basic functions but also overall benefits. Perceived benefits is closely related to satisfaction and affects users’ intention to continue using the service.

Secondly, two dimensions of perceived quality were identified: information quality and service quality, with information quality measured by accuracy and completeness. The results showed that accuracy significantly affected perceived benefits and confirmation, while completeness significantly impacted privacy concerns and confirmation. However, no clear relationships were found between accuracy and privacy concerns or between completeness and perceived benefits. These interesting findings can be attributed to the unique context of WeChat e-commerce. On the one hand, the trust mechanism built upon familiar social networks reduced users’ risk sensitivity to information errors [116]. Furthermore, the small transaction amounts and its high frequency, characteristics of retail purchases, may lead to quicker decision-making, diminishing the impact of information accuracy on privacy concerns. On the other hand, information completeness did not invariably equate to perceived usefulness, and excessive details can cause information overload [117]. Customers often seek supplementary information through online customer service and community interactions, diminishing their reliance on completeness. The limited ability of SMEs to optimize information presentation while ensuring completeness further weakens the relationship between completeness and perceived benefits.

Thirdly, within the service quality dimensions, efficiency and procedural justice were selected as key factors. Efficiency has long been studied as a key dimension of service quality [80], and this study confirmed its strong predictive power. Our results showed significant relationships between efficiency and privacy concerns, perceived benefits, and confirmation. Procedural justice is also closely related to privacy concerns and confirmation, which validates the rationality of choosing procedural justice as a dimension of WeChat e-commerce service quality. However, the relationship between procedural justice and perceived benefits has not been confirmed, attributed to the following reasons: the transaction process design of WeChat e-commerce (especially for SME retailers) is relatively simplified, which makes consumers prioritize immediate outcomes, such as price and delivery speed, over process fairness. Furthermore, when pursuing instant and convenient shopping experiences, consumers often passively accept the platform’s rules. Accordingly, it can be challenging for consumers to recognize how procedural justice impacts their specific benefits unless actual disputes occur.

Finally, grounded in the assumption of consumer rationality in decision-making, we incorporated the privacy calculus perspective to study users’ continuance intention to use WeChat e-commerce for SME retailers. The results corroborated the hypotheses of privacy concerns and perceived benefits, fully validating the rationality of privacy calculus in our research model. Our results demonstrated that privacy concerns significantly affect perceived benefits and continuance intention, suggesting that, if SME retailers neglect privacy protection, heightened consumer privacy concerns will reduce their perceived benefits and negatively impact their intention to continue using the platform. Furthermore, the present study confirmed that perceived benefits were positively correlated with satisfaction and continuance intention, consistent with the literature [89,90]. Overall, customers are more likely to keep using WeChat e-commerce when they perceive the benefits outweigh privacy concerns.

7. Conclusions

7.1. Theoretical Implications

This study offers several theoretical implications. Firstly, it adopts a consumer perspective to explore user retention in WeChat e-commerce for SME retailers, addressing gaps in consumer behavior research related to SME e-commerce. Previous research has mainly focused on the enterprise perspective, examining aspects such as e-commerce adoption, performance, and success factors [2,17,39], with less attention given to the consumer perspective, particularly regarding user retention. Besides that, prior studies centered more on general or traditional e-commerce platforms [9,10], while research on specific industries or social commerce scenarios is relatively limited. This study investigated the mechanisms that influence users’ intention to continue using SME retailers’ WeChat e-commerce, providing new insights into user retention in e-commerce. This study extended the ECM model by incorporating perceived quality and privacy calculus and substituting perceived usefulness with perceived benefits, enhancing its suitability for the context of WeChat e-commerce. This contributes to the enrichment of the ECM framework and expansion of its application in social commerce research.

Secondly, analyzing the impact of perceived quality modules on continuance intention is beneficial for future research in e-commerce and other IS fields. Previous studies often approached this topic from a narrow perspective, treating perceived quality as a whole concept [118] or focusing solely on service quality [84]. Considering the two key attributes of e-commerce (information technology and commercial service), this study explored users’ continuance intention from both the information quality and service quality perspectives. Within the specific context of WeChat e-commerce, this study further refined these dimensions: information quality was measured by accuracy and completeness, while service quality was assessed by efficiency and procedural justice. The findings demonstrate the essential role of perceived quality in e-commerce and the obvious advantages of its segmentation. Our research provides a reference for scholars studying quality segmentation and encourages further exploration of the dimensions of information quality and service quality.

Thirdly, this study innovatively applied privacy calculus theory to the WeChat e-commerce domain, a theory that has been widely used in online contexts involving information disclosure [58,59]. With the development of e-commerce and the enhancement of consumer privacy awareness, this theory provides a powerful tool for understanding consumers’ decision-making processes under privacy risks [55]. We have noticed that consumers carefully weigh costs and benefits when making decisions, and this trade-off also occurs when deciding whether to continue using WeChat e-commerce. Therefore, this study adopted privacy concerns as a cost factor and perceived benefits as a benefit factor within the privacy calculus framework, which will provide inspiration for the choice of trade-off factors. Besides that, it not only validated the applicability of the privacy calculus theory in e-commerce but also revealed the uniqueness of user decision-making in social commerce, offering valuable insights for future research.

7.2. Practical Implications

The study also provides practical guidance for operators and developers of WeChat e-commerce for SME retailers. Firstly, understanding the influence of the quality module in WeChat e-commerce is essential for SME retailers to make informed decisions and enhance the quality of their WeChat e-commerce. Information quality is the core dimension of perceived quality, with completeness and accuracy being critical. Therefore, WeChat e-commerce operators should ensure comprehensive coverage of essential business information, including product descriptions, transaction processes, and service guarantees. Meanwhile, developers should minimize information errors by optimizing review and retrieval to deliver precise and reliable information. Moreover, service quality is pivotal in the success of WeChat e-commerce for SME retailers. To improve perceived service quality, developers should optimize system performance, reduce wait times, and improve service delivery efficiency. Operators must also ensure fair service access by establishing uniform standards, promoting process transparency, and providing complaint channels to enhance user fairness.

Secondly, our results emphasize the significance of privacy concerns in hindering behavioral intention and reveal the positive impact of perceived benefits on continuance intention. Service providers should recognize this and adopt strategies to reduce privacy concerns and increase perceived benefits. SME operators should provide consumers with more perceived benefits. For example, they can offer coupons, discounts, and free returns to enhance consumers’ perception of tangible benefits. SME operators can cooperate with various service providers to integrate more services, like logistics, to create one-stop shopping and enhance intangible benefits. Establishing online brand communities and hosting interactive events can also bring additional benefits to customers. Furthermore, privacy concerns remain a significant issue for consumers, emphasizing the need for developers to design diverse privacy settings that allow consumers to adjust their privacy rights and enhance their sense of control. Meanwhile, operators should develop clear privacy policies and inform consumers about how their personal information is collected and used, which can alleviate privacy concerns.

7.3. Limitations and Future Work

Although our research provides certain theoretical and practical contributions, there are still some limitations and future research directions that need to be discussed. Firstly, the generalizability of findings may be affected by limitations in participant recruitment. Due to the constraints of online surveys and the characteristics of the retailer’s e-commerce user group, the respondents were mainly young women, which limits the applicability of the conclusions to male users and older user groups. Furthermore, given the limitations of the research conditions, this survey mainly focused on Chinese users. As WeChat becomes a global social media platform, its e-commerce functionalities are increasingly adopted by users from different cultural backgrounds. Future research could optimize the sample structure through stratified sampling and incorporate cross-border user groups to enhance the generalizability of conclusions.

Secondly, there is room for optimization in the dimension division of perceived quality. While we divided perceived quality into information quality (accuracy and completeness) and service quality (efficiency and procedural justice) based on WeChat e-commerce characteristics, this classification primarily addresses traditional social commerce. The emergence of new formats, such as live-streaming and short-video commerce, has transformed users’ perceptions of quality dimensions, potentially necessitating a reassessment of the composition of information quality and service quality. For instance, information quality might need to incorporate dimensions such as timeliness, clarity, or relevance, and service quality might need to include indicators such as interactivity and responsiveness. Future research should broaden the dimensions of quality perception in response to these emerging scenarios to provide insights into the continued use of social commerce platforms.

Finally, the boundary conditions and moderating effects of the model require further exploration. This study mainly focused on the relationships among core variables, with little emphasis on contextual factors like cultural background and user differences. Indeed, these factors may moderate the strength of the relationships among core variables by influencing users’ privacy attitudes and e-commerce behaviors. Given that privacy concerns and e-commerce behaviors are influenced by culture, future research should explore the impact of factors such as Confucian values or China’s data governance framework on the research results. Moreover, individual characteristics, such as user experience and brand trust, may have significant moderating effects. Subsequent studies could incorporate these factors to build a more comprehensive theoretical model, thereby enhancing the explanatory power of the findings in various contexts.