The Impact of Service Convenience in Online Food Delivery Apps on Consumer Behavior in the Chinese Market: The Moderating Roles of Coupon Proneness and Online Reviews

Abstract

To enhance the performance of online food delivery (OFD) services, it is essential to strengthen consumers’ intentions to use OFD apps, which are the core interface of this business. Accordingly, this study aims to identify the cognitive mechanisms that shape consumers’ intentions to use OFD apps and explore strategies to encourage their adoption. To achieve this, the study develops a research model that incorporates segmented dimensions of service convenience as key motivational factors, along with variables from the Technology Acceptance Model (TAM). A survey was conducted among OFD consumers in China, and the proposed research model was empirically tested using data from 478 valid responses. The analysis revealed that all six dimensions of service convenience serve as significant motivational drivers of OFD app usage. Furthermore, the study demonstrates that consumers’ coupon proneness and user-generated online reviews have significant moderating effects that reinforce the mechanism by which consumers adopt and use OFD apps. The findings and implications discussed in this study are expected to provide valuable insights and practical guidance for formulating effective strategies to promote more active consumer engagement with OFD apps in the future.

1. Introduction

Smartphones have become a vital component of everyday life in recent years, altering consumer behavior and driving the growth of mobile commerce [1,2]. Among the industries that have benefited from the rise of smartphones, online food delivery (OFD) services have grown particularly rapidly, with smartphones offering consumers greater convenience and flexibility to place orders at any time and from anywhere [3,4]. In this context, OFD has become the major food service consumption model for consumers in China and is now deeply embedded in the daily lives of millions of consumers [1,5]. According to Statista [6], China’s OFD market has evolved into the largest in the world, surpassing 540 million users and generating nearly USD 450 billion in revenue in 2024. This rapid growth is supported by a network of approximately three million affiliated restaurants, contributing to the creation of more than five million jobs [6].

The OFD market in China is predominantly dominated by two major platforms, “Meituan” and “Ele.me”, which together account for more than 90% of the market share by the end of 2024 [7]. Under these circumstances, the OFD market in China is fiercely competitive, and these platforms have made various marketing strategies to attract more customers and increase purchases [8]. To make these strategies work successfully, it is essential for OFD providers to have a deep understanding of the motivations that drive consumers’ ordering behavior through their apps and the cognitive processes that lead to their purchase decisions [9].

A variety of factors have been identified as key factors influencing consumers’ motivation to use OFD apps, including food quality (e.g., freshness, taste, and presentation) [10,11,12], price (e.g., value for money), delivery speed and reliability (e.g., accuracy, packaging, and customer service) [13], and enjoyment and convenience [14]. Among these factors, service convenience, in particular, has been identified as a key determinant in several studies (e.g., [14,15,16]). The concept of service convenience should be examined in several sub-dimensions for further study. A representative study in this regard is the study by Berry et al. [17]. Berry et al. [17] presented five core dimensions of service convenience: accessibility, transaction, decision, benefit, and post-service benefit. Building on this foundation, Seiders et al. [18] and Colwell et al. [19] empirically verified these five dimensions and developed the theory into a practically applicable form. This five-dimensional framework has been further extended and applied in various contexts, notably in OFD-related studies (e.g., [15,16,20]). While these dimensions in the framework are proper for comprehensively capturing the main aspects of convenience, they fail to encompass the aspect of search behavior within the Internet environment. As Engel and Blackwell [21] pointed out in their classical consumer decision-making model, the actual buying process typically includes five distinct stages: problem recognition, information search, alternative evaluation, purchase, and post-purchase behavior. Among these stages, information search is recognized as a separate and essential phase that directly precedes decision-making. In line with this, Moeller et al. [22] also defined shopping convenience as comprising five distinct dimensions—decision, access, search, transaction, and after-sales convenience—each relevant at separate stages of the purchasing process. Among these dimensions, search convenience reflects the ease with which consumers can identify and select desired products, a process that may be significantly hindered by poor information design or product overload [22]. Similarly, Jiang et al. [23] also emphasized that search inconvenience is perceived by consumers as a significant barrier to efficient and convenient online shopping, especially in mobile-based service environments. In addition, Almarashdeh et al. [2] further stressed the significance of search convenience in mobile shopping, noting that customers typically begin their experience on an e-commerce platform by searching for products using search tools. Thus, given the rapid expansion of OFD apps, real-time access to restaurant options, menus, and promotions plays a distinct and critical role in shaping user experience, and this study suggests that the concept of service convenience for OFD apps used in the Internet environment also needs to be expanded to six dimensions, including search convenience.

Meanwhile, to promote the use of OFD apps, it is also necessary to consider a method that utilizes moderating factors in addition to an approach that enhances service convenience, which is related to the intention to use OFD. However, research on this seems to be still insufficient. Several previous e-commerce-related studies have demonstrated that consumer characteristics, such as coupon proneness, play a significant role in enhancing business performance [24,25,26] as they are strongly associated with consumer purchasing behavior. However, research on OFD has not yet proven the moderating effect of coupon proneness. Similarly, online reviews have also been recognized as playing an essential role in purchasing decisions in e-commerce [27]. Regarding OFD services, Yang and Fang [28] argued that online reviews play a crucial role in consumers’ evaluation of food quality and safety, which ultimately are associated with their purchasing decisions. Nevertheless, in OFD research, the moderating role of online reviews on consumers’ purchasing intentions and behaviors has seldom been empirically proven. Thus, it is necessary to consider coupon proneness and online reviews as moderating factors from the perspective of consumer characteristics and marketing promotions, respectively.

To fulfill the identified gaps discussed above and conduct research, this study proposes the following research questions (RQs): (1) What is the cognitive mechanism by which service convenience is associated with consumers’ intentions to use OFD apps? (2) What factors can further promote consumers’ intentions and behaviors to use OFD apps? To answer RQ1, this study first adopts the Technology Acceptance Model (TAM) as a theoretical framework. The TAM is a widely recognized theoretical framework that effectively analyzes users’ cognitive mechanisms for accepting new technologies and has been extensively applied to explain their behavioral intentions and actual usage [29,30,31,32]. Therefore, TAM is highly appropriate for this study, which examines service convenience as a key feature provided by OFD apps, typically driven by new information technologies, and explores how it relates to users’ acceptance (i.e., intention to use). Moreover, to gain a deeper understanding of the mechanism underlying user acceptance of OFD apps, this study also integrates six specific dimensions of service convenience [2,17], namely, access, search, transaction, decision-making, benefit, and post-benefit, into the TAM framework [32]. Meanwhile, to answer RQ2, this study focuses on examining how coupon proneness and online reviews are associated with variations in the observed relationships, by setting them as moderator variables representing customer characteristics and marketing aspects, respectively. The analysis results and implications of this study are expected to provide valuable information and insights for developing effective strategies to enhance the performance of OFD businesses.

The rest of the paper is organized as follows. Section 2 reviews the theoretical background of service convenience. Section 3 presents the research model and hypotheses. Section 4 describes the research method, and Section 5 presents the results. Section 6 discusses the findings, provides practical implications and theoretical contributions, and concludes with a discussion of limitations and directions for future research.

2. Theoretical Background

2.1. Technology Acceptance Model (TAM)

The TAM, initially proposed by Davis [33], has been widely adopted to explain users’ acceptance and use of new technologies. The original version of TAM emphasized two key cognitive beliefs, specifically, perceived usefulness and perceived ease of use, which influence users’ attitudes toward using a system, thereby affecting their actual system use. Later, Davis et al. [34] refined the model by explicitly introducing behavioral intention as a mediating variable between attitude and actual use, highlighting intention as a more immediate predictor of usage behavior. In the modified TAM model of 1989, it also consolidated the influence of external variables through perceived usefulness and perceived ease of use. Subsequently, in 1996, Venkatesh and Davis [32] further simplified the model by removing the attitude component and directly linking perceived usefulness and perceived ease of use to behavioral intention, which then leads to usage behavior. In academic research, this streamlined version has become the most commonly adopted TAM framework due to its parsimony and substantial empirical support. This study also adopts the 1996 TAM model as the foundational framework (see Figure 1).

2.2. Service Convenience

The concept of convenience can be traced back to the early 20th century when it was first associated with product categories. Copeland [35] introduced the term “convenience goods” to describe products that are widely available and require minimal time and effort for consumers to purchase. Later, Kelley [36] expanded this concept to services, emphasizing that reducing the time and effort needed to acquire goods and services is the key to attracting consumers. Subsequently, as the service economy grew, Kotler and Zaltman [37] applied the concept of convenience to the service sector, suggesting that simplifying the service acquisition process can significantly enhance customer experiences. Brown [38] further defined service convenience as any factor that improves consumer comfort during the purchase and use of services.

Based on these theoretical foundations, Berry et al. [17] proposed that service convenience refers to the amount of time and effort that customers believe it takes to purchase and use services. Berry et al. [17] also emphasized that service convenience encompasses multiple dimensions (i.e., access, transaction, decision-making, benefit, and post-benefit) and should be analyzed throughout the consumer’s service journey, which includes problem recognition, information search, evaluation, purchase, and post-purchase stages. Building on this perspective, Colwell et al. [19] further argued that time and effort are crucial non-monetary resources that affect service convenience. Moreover, Colwell et al. [19] also stressed that minimizing the time and effort required to obtain a service not only enhances service convenience but also improves the overall customer experience, thereby increasing the perceived value of the service. While most previous studies have emphasized time and effort as key elements of service convenience, Thuy [39] provided an alternative perspective, highlighting monetary cost as a crucial resource that influences consumers’ perception of convenience.

In the context of e-commerce, because consumers place considerable importance on the ease of access and use of online services, service convenience has been recognized as playing a significant role in shaping the overall user experience and has become a key factor influencing purchase decisions [40]. In line with this, service convenience has emerged as a critical determinant of success for digital platforms [41]. This convenience is often reflected in seamless and user-friendly shopping experiences, which many studies have found to significantly enhance user satisfaction [42,43,44]. Among the convenience-related factors contributing to user satisfaction, time savings have been identified as particularly important in the e-commerce environment [42]. Building on this, Rajamma et al. [43] further emphasized that reducing wait times can notably enhance consumers’ perceptions of online platforms, which in turn leads to greater satisfaction. In addition, Collier and Kimes [44] also noted that streamlining payment processes and making transactions effortless can significantly enhance the shopping experience, especially in e-commerce, where seamless payments are crucial for attracting and retaining users.

In particular, for OFD services, the importance of service convenience has become more pronounced as demand surged during the COVID-19 pandemic. Prasetyo et al. [45] argued that the COVID-19 pandemic led to increased consumer demand for time-saving and streamlined operations and that these changes in OFD services have positioned service convenience as a key factor in promoting customers’ purchase intentions. Similarly, Shah et al. [46] demonstrated that features such as contactless delivery and secure packaging through OFD services provide consumers with a more convenient and safe service experience, ultimately improving customer satisfaction. On the other hand, as consumer expectations for service convenience continue to expand, search convenience has emerged as another critical factor in shaping the digital experience [2]. In addition to avoiding the inconvenience of visiting restaurants, consumers increasingly expect to find the products or services they want on digital platforms more quickly and easily. In line with this, Hasan [47] noted that search convenience has become a key element affecting consumer behavior and purchase decisions. Therefore, to reflect this trend, this study incorporates the concept of search convenience proposed by Hasan [47], alongside the five service dimensions suggested by Berry et al. [17], to capture the characteristics of consumer behavior in the e-commerce environment. Collectively, these six dimensions provide a comprehensive framework for understanding service convenience in digital service environments. The concepts of these six dimensions are summarized as follows:

• Access convenience: Reflects how easily consumers initiate services through various channels, including remote communication (e.g., contacting restaurants via app), with minimal time and effort.
• Transaction convenience: Involves the ease of completing payments and transactions, with simplified processes and various payment options that enhance the transaction experience.
• Decision-making convenience: Refers to how easily consumers make decisions by simplifying information access and reducing complexity.
• Benefit convenience: Relates to the quality and efficiency of services, focusing on timely deliveries and fulfilling consumer expectations.
• Post-benefit convenience: Covers support for post-purchase needs, such as refunds and exchanges, which contribute to customer loyalty through accessible after-sales services.
• Search convenience: Involves the ability to quickly find products or services using efficient search tools, filters, and recommendation systems, thereby benefiting from time savings and enhancing the overall user experience.

3. Research Model and Hypotheses

3.1. Research Model

This study aims to explore strategic measures for promoting consumer use of OFD apps by investigating the mechanism through which service convenience is associated with usage intention and exploring ways to facilitate this process. To achieve this, the study modifies the original TAM [32] by introducing the independent constructs of convenience along with two moderating variables, as shown in Figure 2. Prior studies (e.g., [32,48,49]) have consistently demonstrated that perceived usefulness exerts a stronger and more enduring influence on behavioral intention than perceived ease of use. For example, Venkatesh and Davis [32] found that perceived usefulness had a consistently more substantial effect on behavioral intention, particularly as users gained more experience with the system. Similarly, Gefen and Straub [48] emphasized that perceived usefulness is the primary belief influencing intentions to use IT, especially in service contexts such as e-commerce. Venkatesh et al. [49] further reinforced this by identifying perceived usefulness—captured within the performance expectancy construct—as the most consistent and strongest predictor of behavioral intention across multiple user acceptance models. Moreover, Ahn et al. [50] also stated that compared with perceived ease of use, perceived usefulness has a more direct and significant influence on users’ behavioral intention. Beyond the above, perceived ease of use has been increasingly criticized for its limited conceptual scope. As Collier and Kimes [44] pointed out, while perceived ease of use primarily focuses on the interactivity and usability of a system interface, it fails to account for broader contextual and environmental factors such as the location of the technology, system accessibility, and the surrounding social environment. Moreover, perceived ease of use tends to overlook the time and effort exerted across the entire service process, focusing narrowly on the moment of interaction.

In contrast, the concept of convenience itself offers a more comprehensive framework by capturing the cognitive and physical effort required before, during, and after a transaction [44]. Following this perspective, this study adopts convenience as an independent construct to reflect user effort across all stages of service use. In addition, to better align the TAM framework [32] with the characteristics of OFD services, this study retains perceived usefulness, behavioral intention, and actual usage as core variables, while not including perceived ease of use. Meanwhile, this study also set coupon proneness and online review as moderator variables that can promote OFD app usage intention and behavior. This research model will help clarify the association of each dimension of service convenience on consumers’ perception of OFD services and verify key factors that promote app usage.

3.2. Research Hypotheses

3.2.1. From Convenience to Perceived Usefulness

In this study, the construct of convenience is operationalized through six dimensions, namely, access, search, transaction, decision-making, benefit, and post-benefit convenience, which appear to offer practical means to reduce users’ operational complexity and time costs [17,47], thereby enhancing their perception of the app’s usefulness in the context of the OFD industry. A detailed discussion of this perspective concerning each of the six dimensions of service convenience is as follows.

First, access convenience reflects the ease with which users can engage with online stores regardless of geographic constraints [51,52]. In the context of the OFD industry, this type of convenience helps reduce time costs for consumers, thereby increasing their perception of the service’s usefulness. Moreover, the 24/7 availability of access further allows consumers to shop at any time that suits their schedule, thus removing the constraints of physical store hours and enhancing overall convenience [53]. Thus, the following hypothesis is proposed:

H1-1. 

The access convenience of OFD apps is positively associated with users’ perception of app usefulness.

Next, search convenience refers to the ease with which consumers can quickly and efficiently find relevant information about food providers (e.g., restaurants) within the OFD apps. A smooth search process enhances user experience and strengthens their perception of the app’s usefulness. As Almarashdeh et al. [2] highlighted, effective and smooth search functions significantly reduce users’ time and cognitive effort, enhancing the overall shopping experience. Moreover, their study also showed that precise categorization and intuitive interface design play a crucial role in enhancing search efficiency, which is particularly important in digital platforms like OFD apps. Thus, the following hypothesis is proposed:

H1-2. 

The search convenience of OFD apps is positively associated with users’ perception of app usefulness.

Transaction convenience means the speed and security of the payment process in OFD apps. A reliable and seamless payment process generally can reduce consumer concerns, leading to greater customer satisfaction [14,54]. Moreover, offering multiple payment options and ensuring secure transactions can reduce users’ perceived uncertainty and enhance the overall convenience of the payment process, thereby increasing their perception of the OFD apps’ usefulness [55]. Thus, the following hypothesis is proposed:

H1-3. 

The transaction convenience of OFD apps is positively associated with users’ perception of app usefulness.

Decision-making convenience refers to the use of decision-support tools within web-based store environments (e.g., OFD apps) that enable consumers to efficiently form a consideration set and make informed purchase decisions, especially when facing time constraints [56]. This aspect of convenience can increase consumer satisfaction and reinforce their perception of the platform’s usefulness. Tools like comparison options and personalized recommendations save consumers’ time and effort in making choices, making the overall shopping experience more efficient [57]. Thus, the following hypothesis is proposed:

H1-4. 

The decision-making convenience of OFD apps is positively associated with users’ perception of app usefulness.

Benefit convenience refers to the extent to which a service effectively fulfills consumer needs, thereby enhancing perceived value and overall user experience [17,18]. Yum and Yoo [58] pointed out that while factors such as product usefulness, design, and safety contribute to customer satisfaction, benefit convenience—the value consumers gain from using a product or service—is the most critical factor influencing both perceived usefulness and customer satisfaction. These findings suggest that benefit convenience not only makes the service easy to use but also correlates with greater perceived value of the service. Thus, the following hypothesis is proposed:

H1-5. 

The benefit convenience of OFD apps is positively associated with users’ perception of app usefulness.

Finally, post-benefit convenience refers to the ease of accessing after-purchase services, such as refunds, exchanges, and other post-transaction support [59,60]. In general, well-designed post-purchase services benefit from building consumer trust in the platform, especially when issues arise. Lee and Nam [60] argued that post-purchase services can significantly enhance consumers’ perception of service usefulness by providing timely and efficient approaches to problem resolution. Thus, the following hypothesis is proposed:

H1-6. 

The post-benefit convenience of OFD apps is positively associated with users’ perception of app usefulness.

3.2.2. From Convenience to Intention to Use

Several previous studies have demonstrated that service convenience plays a significant role in influencing the intention to use e-commerce platforms [14,52,61,62]. For example, Yeo et al. [14] argued the significance of convenience in OFD service platforms, highlighting its substantial association with consumers’ attitudes and behavioral intentions. Similarly, Duarte et al. [52] also noted that the convenience of digital services not only enhances customer satisfaction in online shopping but also positively correlates with their behavioral intentions. The detailed dimensions of service convenience also appear to have a distinct relationship with the intention to use, as discussed below.

In terms of access convenience and intention to use, Lucas et al. [63] analyzed how different access platforms, such as mobile versus desktop, shape the overall user experience, highlighting that mobile commerce (m-commerce) significantly enhances accessibility and usability compared with traditional e-commerce platforms. Their findings suggest that improved access, primarily through mobile devices, reduces barriers to entry, allows for greater flexibility, and supports more frequent and spontaneous usage behavior. In the context of OFD apps, this type of access convenience—being able to place orders anytime, anywhere—plays a similarly critical role. Thus, the following hypothesis is proposed:

H2-1. 

The access convenience of OFD apps is positively associated with users’ intention to use the apps.

Regarding search convenience, Hasan [47] found that it has a significant relationship with consumer behavior and decision-making processes. Similarly, Jung et al. [64] noted that customers tend to prefer platforms with simple and effective search functions, particularly in the context of OFD apps. A well-designed search feature not only enhances the overall user experience but also reduces the time and effort required for decision-making, which in turn increases users’ intention to continue using the platform [64]. Thus, the following hypothesis is proposed:

H2-2. 

The search convenience of OFD apps is positively associated with users’ intention to use the apps.

Moving to transaction convenience, prior studies have shown that when a platform offers secure and seamless payment methods, users are more likely to develop trust in the platform [65,66]. A safe and smooth payment process reduces perceived uncertainty and risk [65], which thereby helps to increase users’ intention to use the service. Thus, the following hypothesis is proposed:

H2-3. 

The transaction convenience of OFD apps is positively associated with users’ intention to use the apps.

Regarding decision-making convenience, Jameson et al. [67] argued that a simplified menu display and personalized recommendations help users reduce information overload, while speeding up the decision-making process and enhancing their usage intention. When users can proceed efficiently through the purchasing decision process, their experience becomes more satisfying, which ultimately strengthens users’ willingness to use the service and increases the likelihood of repurchase [20]. Thus, the following hypothesis is proposed:

H2-4. 

The decision-making convenience of OFD apps is positively associated with users’ intention to use the apps.

In terms of benefit-related convenience, Kim and Lee [15] found that when users can easily obtain the benefits of a service, their perceived value and satisfaction increase, which in turn boosts their intention to reuse the platform. In addition, Kim and Lee [15] also noted that when users experience efficient and seamless services, they become more satisfied and dependent on the e-commerce platform, which in turn helps foster users’ long-term usage intentions, habits, and loyalty. Thus, the following hypothesis is proposed:

H2-5. 

The benefit convenience of OFD apps is positively associated with users’ intention to use the apps.

Regarding post-benefit convenience, when an e-commerce platform can quickly and effectively resolve post-purchase issues that users encounter, users’ satisfaction and trust can be significantly enhanced, which in turn increases their usage intention [60]. This suggests that by providing adequate post-purchase support and making users feel valued, platforms can increase long-term dependency and increase users’ intention to use the service repeatedly. Thus, the following hypothesis is proposed:

H2-6. 

The post-benefit convenience of OFD apps is positively associated with users’ intention to use the apps.

3.2.3. From Perceived Usefulness to Intention to Use

Perceived usefulness in TAM usually measures how much users perceive a service or technology as enhancing convenience and efficiency [34]. Perceived usefulness has been widely validated in numerous studies as a central variable driving intention to use, particularly in the fields of e-commerce and digital services. For example, Gefen et al. [68] discovered that consumers’ purchase intentions are significantly associated with their perceptions of a platform’s usefulness. In addition, Venkatesh and Davis [32] and Venkatesh et al. [69] found that perceived usefulness plays a crucial role in shaping users’ intentions to use and adopt new technologies. Specifically, Cheng et al. [70] demonstrated that perceived usefulness can significantly improve consumers’ inclination to use Internet banking. In addition, Lu and Su [61] emphasized that perceptions of a platform’s usefulness can greatly enhance consumers’ willingness to make purchases on mobile shopping websites. Beyond the above, Kim and Park [71] noted that perceived usefulness plays a crucial role in shaping consumers’ intentions to adopt health information technology. These studies consistently show that when users perceive a service or a product as useful, their intention to adopt it increases significantly. Thus, the following hypothesis is proposed:

H3. 

The usefulness of OFD apps is positively associated with users’ intention to use the apps.

3.2.4. From Intention to Use to Usage Behavior

Intention to use refers to the tendency of consumers to transform their willingness and beliefs (e.g., trust in the object) into potential actions after forming an attitude toward a particular object [72], while usage behavior generally refers to the frequency or intensity of a user’s actual engagement with a specific technology or service over a particular period [45,73]. In information systems and user behavior research, it reflects the extent to which behavioral intention is translated into action, specifically whether and how frequently users utilize the technology (e.g., OFD apps or mobile shopping applications). It is typically measured using Likert-scale items that assess how often users interact with the app within a defined time frame (e.g., weekly or monthly). Previous studies have shown that behavioral intention serves as an immediate determinant of actual behavior. For example, Venkatesh and Davis [74] emphasized that behavioral intention serves as a key mediator between system beliefs and actual system use, with empirical evidence across longitudinal field studies supporting the strong association between intention and subsequent behavior. Building on this general relationship, Gefen et al. [68] demonstrated how trust, a core belief in e-commerce contexts, shapes users’ behavioral intentions, which in turn translate into actual usage. In addition, by expanding the TAM framework into the Unified Theory of Acceptance and Use of Technology (UTAUT), Venkatesh et al. [49] identified behavioral intention as the most consistent and strongest determinant of actual usage behavior across diverse IT acceptance contexts among users. Subsequently, Venkatesh et al. [69] further emphasized that in various technology adoption contexts, users’ intention to use plays a crucial role in driving actual usage behavior, with stronger intentions more likely to exhibit higher levels of behavioral executions. Although all the above research aligns with the widely accepted theoretical assumption that behavioral intention precedes actual behavior, as posited in models such as the theory of reasoned action [75], the theory of planned behavior [76], protection motivation theory [77], and the theory of interpersonal behavior [78], it still cannot be ignored that the relationship may also be reciprocal, where past usage behavior influences current intention to use, primarily through the formation of habits or reinforced perceptions of usefulness [79,80]. Positive online shopping experiences have also been shown to be associated with consumers’ intentions to make future purchases [40,81,82]. Such experiences can be viewed as reflections of consumers’ past usage behavior, which in turn reinforces their intention to reuse the service. Specifically, consumers who have had positive experiences in the past are more likely to reuse the service [83]. However, since this study adopts the TAM [32] as the theoretical framework, which conceptualizes behavioral intention as a precursor to actual behavior, the potential reverse relationship is not addressed. Thus, the following hypothesis is proposed:

H4. 

Users’ intention to use OFD apps is positively associated with their app usage behavior.

3.2.5. The Moderating Effect of Coupon Proneness

Coupon proneness relates to a consumer’s tendency to seek out and use coupons when making purchasing decisions actively [84]. This behavioral tendency reflects consumers’ sensitivity to coupons and their habitual engagement with promotional offers provided by sellers [84]. Such engagement increases the perceived value of a product or service by effectively lowering its actual cost to consumers [25,84]. The enhanced perceived value, in turn, strengthens purchase intentions, especially among consumers with high coupon proneness [84]. Many existing studies have explored the broader implications of coupon proneness on consumer psychology and behavior. For example, Zheng et al. [85] emphasized that a strong propensity to use coupons is essential for fostering repeat business and customer loyalty, especially on e-commerce platforms where discounts significantly enhance consumers’ motivation to engage with the platform. Similarly, Ren et al. [86] also argued that the advancement of online marketing tools has made the effect of electronic coupons especially significant for consumers with strong coupon proneness. Thus, the following hypothesis is proposed:

H5-1. 

OFD app users’ coupon proneness moderates the association between their perception of app usefulness and their intention to use the app.

In the meantime, the coupon proneness also has a significant relationship with customers’ actual usage behavior. Chen et al. [87] found that coupons can be positively associated with consumers’ perceptions of price, thereby affecting their purchasing decisions. Consumers are more likely to complete a purchase, especially when promotional information is presented effectively through coupons [87,88]. Similarly, Zheng et al. [85] argued that consumers who are more sensitive to coupons tend to make quicker purchase decisions when coupons are available. Moreover, Zheng et al. [85] further stressed that if consumers are repeatedly exposed to such promotional incentives, they are more likely to engage in frequent transactions and develop a habit of repeat purchasing. Additionally, Sigala [89] also showed that coupons are a powerful tool not only for attracting new customers but also for enhancing the loyalty of existing users. These studies indicate that coupons play a critical role in influencing both consumers’ purchase intentions and their actual purchasing behavior. This association is even more substantial among consumers with high coupon proneness as they are more likely to convert their purchase intentions into actual actions by perceiving greater value and benefits offered through coupons in the promoted product or service. Thus, the following hypothesis is proposed:

H6-1. 

OFD app users’ coupon proneness moderates the association between their intention to use the app and their usage behavior.

3.2.6. The Moderating Effect of Online Reviews

As a form of user-generated content, online reviews have become an indispensable source of information for consumers when making purchasing decisions [27]. In the context of OFD services, such reviews typically convey users’ personal experiences, including feedback on food quality, delivery speed, and overall satisfaction [90]. These shared experiences not only reflect previous users’ service evaluations but also offer prospective users valuable information that helps shape their behavioral expectations. When potential customers are exposed to a high volume of positive reviews, particularly those emphasizing hygiene, food pricing, and reliability, they are more likely to develop favorable expectations about the OFD app and, consequently, have a stronger intention to use it [90,91]. Thus, the following hypothesis is proposed:

H5-2. 

Online reviews in OFD apps moderate the association between the perception of app usefulness and intention to use.

At the same time, previous studies have argued that online reviews can also be related to customers’ actual usage behavior. For example, Floh et al. [92] found that the emotional tone of online reviews significantly correlates with consumers’ buying behavior, with positive emotional reviews being particularly effective in driving purchase decisions. Yagci and Das [27] highlighted that well-written and positive reviews can increase consumers’ confidence and trust in the platform, which then encourages them to make actual purchases. Alalwan [93] argued that review-related features, such as online review contents, ratings, and tracking, shape users’ expectations and perceived value, which ultimately promote continuous use of mobile food ordering apps. All these studies indicate that when consumers are positively associated with online reviews, they are more likely to turn their intention into actual usage. Thus, the following hypothesis is proposed:

H6-2. 

Online reviews in OFD apps moderate the association between users’ intention to use the apps and their actual usage behavior.

4. Research Method

4.1. Survey and Samples

Data for verifying the research model was collected through an online survey. Data collection was conducted through a professional survey agency in China (“Wenjuanxing”). To ensure the clarity and appropriateness of the measurement items, the questionnaire first underwent expert review before distribution. Based on expert feedback, the wording was refined, and the language was localized to enhance semantic clarity and cultural relevance. Next, after the confirmation of the questionnaire design, participants were recruited from a nationwide consumer panel based on the following inclusion criteria: (1) aged 18 years or older and (2) had prior experience using OFD apps. After going through the process of selecting OFD app users, 513 responses were collected. Subsequently, to ensure data quality, responses were screened and excluded based on the following criteria: (1) completion time under 5 min (i.e., less than one-third of the median duration) and (2) uniform responses across multiple Likert-scale items (e.g., straight-lining). Finally, after applying these quality control procedures, 478 valid responses remained for analysis. During this survey period, the survey agency distributed a small monetary reward to participants who completed the questionnaire on behalf of the authors as a gesture of appreciation. The demographic characteristics of the respondents are as shown in

Table 1. Sample characteristics.

Characteristics	Options	No. (478)	(%)
Sex	Male	259	54.18%
	Female	219	45.82%
Age	18~20 Yrs	89	18.62%
	20~29 Yrs	223	46.65%
	30~39 Yrs	102	21.34%
	40~49 Yrs	37	7.74%
	More than 50 Yrs	27	5.65%
Monthly income	Under RMB 4000	128	26.78%
	RMB 4000~5999	111	23.22%
	RMB 6000~7999	96	20.08%
	RMB 8000~9999	61	12.76%
	Over RMB 10,000	82	17.15%
Number of OFD apps uses	1	133	27.82%
	2	107	22.38%
	3	122	25.52%
	Over 4	116	24.27%
Main time of using OFD apps	Weekdays	139	29.08%
	Weekends	175	36.61%
	No difference	164	34.31%

.

4.2. Measures

In this study, a research model was proposed that comprises six dimensions of convenience, core TAM variables, and moderating variables such as individual traits (i.e., coupon proneness) and environmental factors (i.e., online reviews). All measurement items of these research constructs were developed as reflective measurement constructs tailored to the context of this study, based on the existing literature with established validity and reliability (see

Table 2. Research constructs and measurement items.

Constructs	Codes	Operational Definitions and Measurement Items	Sources
Access convenience (AC)	Consumers’ perception of how easily they can start using the OFD app service and communicate with registered food providers (e.g., restaurants).
	AC1	I can order food at any time through the OFD app.	[15,52,94]
	AC2	I can order food from anywhere through the OFD app.
	AC3	I can easily contact registered restaurants through the OFD app.
	AC4	I can easily communicate with registered restaurants through the OFD app.
Search convenience (SC)	Consumers’ perception of how easily they can find the food they want to order in the OFD app.
	SC1	In the OFD app, I can easily find the food I want.	[2]
	SC2	In the OFD app, I can quickly find the food I want.
	SC3	In the OFD app, I can easily identify foods as they are categorized in an intuitive manner.
	SC4	In the OFD app, I can use search conditions to repeatedly find the same food.
Transaction convenience (TC)	Convenience level of the entire transaction process, including point accumulation, product selection, purchase procedures, payment process, and choosing a payment method that suits the consumer.
	TC1	In the OFD app, I find the product selection process convenient.	[16,52]
	TC2	In the OFD app, I find the purchase process convenient.
	TC3	In the OFD app, I find the payment process convenient.
	TC4	In the OFD app, I can choose a payment method that is convenient for me.
	TC5	In the OFD app, I find the point accumulation process convenient.
	TC6	In the OFD app, I can easily complete the purchase process.
Decision-making convenience (DC)	The amount of time and effort spent on deciding on a purchase and usage.
	DC1	In the OFD app, I find that the menu, provided with pictures, makes it easy to decide on the food I want.	[20]
	DC2	In the OFD app, I find that I can see detailed information about the food, making it easy to decide on my meal.
	DC3	In the OFD app, I find that sufficient information is provided, which helps me make an informed decision about the food.
	DC4	In the OFD app, I find that using it saves me a lot of time when deciding on food.
Benefit convenience (BC)	The extent to which consumers perceive the benefits they gain relative to the time and effort invested in experiencing the core advantages of the service.
	BC1	By using the OFD app, I can easily compare the prices of various restaurants.	[1,60]
	BC2	By using the OFD app, I can avoid the inconvenience of visiting the restaurant.
	BC3	By using the OFD app, I can avoid the waiting time at the restaurant.
	BC4	By using the OFD app, I can benefit from accumulating points.
	BC5	By using the OFD app, I can get discount coupon benefits.
Post-benefit convenience (PC)	The extent to which issues arising after a purchase are resolved quickly and accurately.
	PC1	In the OFD app, I find that it quickly provides solutions when issues arise.	[60]
	PC2	In the OFD app, I find that it quickly provides answers to my inquiries.
	PC3	In the OFD app, I find the cancellation process for orders to be convenient.
	PC4	In the OFD app, I find that the order modification and refund process is convenient.
Perceived usefulness (PU)	It refers to a consumer’s belief that using the OFD app will improve their work performance.
	PU1	I find the OFD app generally useful.	[95]
	PU2	I find the OFD app efficient.
	PU3	I find that the OFD app allows me to order and pay for food efficiently.
Intention to use (IU)	After forming consumer attitudes through the use of the OFD apps, it is defined as the consumer’s behavioral intention and belief toward possible future actions.
	IU1	I will continue to use the OFD app in the future.	[96]
	IU2	I will recommend the OFD app to my friends and acquaintances.
	IU3	I will use the OFD app again next time.
	IU4	I plan to use the OFD app frequently.
	IU5	I intend to prioritize using the OFD app over other methods, such as phone calls.
Usage behavior (UB)	Consumers’ actual actions involving ordering food and checking available options through OFD apps.
	UB1	Over the past 6 months, I have used the OFD app when ordering food.	[45,73]
	UB2	Over the past 6 months, I have used the OFD app to order various foods.
	UB3	Over the past 6 months, I have preferred using the OFD app rather than the restaurant’s own ordering system when ordering food.
	UB4	Over the past 6 months, I have checked the availability of food or restaurants I wanted on the OFD app.
Coupon proneness (CP)	Consumers’ tendency to actively seek, collect, and use coupons when shopping and to respond positively to coupon-related promotions.
	CP1	I feel good when I use coupons in the OFD app.	[85]
	CP2	I enjoy searching for coupons issued by restaurants in the OFD app.
	CP3	I feel that I am getting a good deal when I use coupons in an OFD app.
	CP4	I enjoy using coupons in the OFD app, regardless of how much I can save.
	CP5	I feel joy not only from saving money when I use coupons but also from the act of using the coupons themselves.
Online reviews (OR)	Consumers’ opinions on the reliability, usefulness, and satisfaction of online reviews posted by OFD in relation to food ordering.
	OR1	The online review information provided in the OFD app is highly relevant to the things I am interested in.	[93]
	OR2	The online review information provided in the OFD app is based on facts.
	OR3	The online review information provided in the OFD app is sufficiently detailed.
	OR4	The online review information provided in the OFD app is enough to satisfy my interests.
	OR5	The online review information provided in the OFD app helps me make decisions about food.

). All measurement items were evaluated using a 7-point Likert scale from 1 (strongly disagree) to 7 (strongly agree).

5. Data Analysis and Results

This study employed the PLS-SEM (partial least squares-structural equation modeling) technique to analyze the proposed research model. Given this study was designed from an exploratory approach perspective that applies the concept of fine-grained convenience, the utilization of the PLS-SEM approach is suitable for exploratory research focusing on the relationship between variables and is not constrained by the assumption of normal distribution of the sample [97]. The analysis was conducted using SmartPLS 4.0 (v. 4.1.0.9).

5.1. Measurement Model Evaluation

To evaluate the measurement model used in this study, first, reliability was assessed. The reliability of the measurement items is commonly assessed using composite reliability (CR) and Cronbach’s alpha. When both indices exceed 0.7, the reliability is ensured [97]. As shown in

Table 3. Analysis results of reliability and convergent validity.

Constructs	Items	Loadings (>0.7)	Cronbach’s Alpha (>0.7)	Composite Reliability (>0.7)	AVE (>0.5)
Access convenience (AC)	AC1	0.902	0.898	0.899	0.765
	AC2	0.865
	AC3	0.872
	AC4	0.860
Search convenience (SC)	SC1	0.872	0.858	0.862	0.701
	SC2	0.850
	SC3	0.819
	SC4	0.808
Transaction convenience (TC)	TC1	0.901	0.907	0.908	0.685
	TC2	0.814
	TC3	0.795
	TC4	0.830
	TC5	0.809
	TC6	0.810
Decision-making convenience (DC)	DC1	0.897	0.873	0.874	0.724
	DC2	0.838
	DC3	0.846
	DC4	0.820
Benefit convenience (BC)	BC1	0.879	0.894	0.897	0.702
	BC2	0.826
	BC3	0.824
	BC4	0.821
	BC5	0.838
Post-benefit convenience (PC)	PC1	0.902	0.878	0.879	0.732
	PC2	0.836
	PC3	0.844
	PC4	0.840
Perceived usefulness (PU)	PU1	0.894	0.843	0.850	0.761
	PU2	0.868
	PU3	0.854
Intention to use (IU)	IU1	0.937	0.903	0.906	0.723
	IU2	0.834
	IU3	0.835
	IU4	0.826
	IU5	0.813
Usage behavior (UB)	UB1	0.885	0.874	0.876	0.726
	UB2	0.845
	UB3	0.816
	UB4	0.860
Coupon proneness (CP)	CP1	0.929	0.947	0.992	0.823
	CP2	0.900
	CP3	0.918
	CP4	0.908
	CP5	0.879
Online reviews (OR)	OR1	0.942	0.950	0.973	0.832
	OR2	0.900
	OR3	0.898
	OR4	0.914
	OR5	0.907

, the minimum value for CR is 0.850, and the minimum value for Cronbach’s alpha is 0.843, indicating that both indices demonstrate sufficient reliability for the construct concept.

Next, the validity of the measurement model was evaluated from two perspectives: convergent validity and discriminant validity. To assess convergent validity, the indicator reliability of all items was first examined. As shown in Table 3, all factor loadings exceeded the recommended threshold of 0.7, indicating acceptable convergent validity [97]. Then, convergent validity was further assessed through the average variance extracted (AVE) for each construct. If the AVE value is greater than 0.5, it indicates that convergent validity is achieved [98]. As shown in Table 3, the minimum AVE value is 0.685, indicating that the convergent validity of each construct is satisfactory.

Meanwhile, discriminant validity analysis can be conducted using Fornell–Larcker and HTMT (Heterotrait–Monotrait) criteria [99,100]. The results of the Fornell–Larcker criterion are presented in

Table 4. Results of Fornell–Larcker criterion analysis for discriminant validity assessment.

	AC	SC	TC	DC	BC	PC	PU	IU	UB	CP	OR
Access convenience (AC)	0.875
Search convenience (SC)	0.452	0.837
Transaction convenience (TC)	0.429	0.510	0.827
Decision-making convenience (DC)	0.422	0.424	0.439	0.851
Benefit convenience (BC)	0.480	0.484	0.470	0.415	0.838
Post-benefit convenience (PC)	0.433	0.504	0.451	0.374	0.497	0.856
Perceived usefulness (PU)	0.433	0.445	0.456	0.421	0.451	0.397	0.872
Intention to use (IU)	0.598	0.603	0.623	0.574	0.637	0.551	0.580	0.850
Usage behavior (UB)	0.323	0.354	0.353	0.294	0.400	0.375	0.341	0.592	0.852
Coupon proneness (CP)	0.028	0.005	0.089	0.057	0.026	0.052	−0.029	0.092	0.099	0.907
Online reviews (OR)	0.052	0.015	0.037	0.069	0.071	0.063	−0.03	0.076	0.070	0.654	0.912

Note: Diagonal elements are the square root of AVEs.

. The square roots of the AVE values are higher than the correlation coefficients with other variables, indicating that there are no issues with discriminant validity [101]. The results based on the HTMT ratio are presented in

Table 5. Results of HTMT analysis for discriminant validity assessment.

	AC	SC	TC	DC	BC	PC	PU	IU	UB	CP	OR
Access convenience (AC)
Search convenience (SC)	0.507
Transaction convenience (TC)	0.471	0.567
Decision-making convenience (DC)	0.472	0.480	0.486
Benefit convenience (BC)	0.528	0.540	0.514	0.460
Post-benefit convenience (PC)	0.483	0.573	0.501	0.418	0.556
Perceived usefulness (PU)	0.488	0.508	0.513	0.479	0.507	0.455
Intention to use (IU)	0.663	0.677	0.687	0.642	0.703	0.616	0.658
Usage behavior (UB)	0.363	0.403	0.395	0.332	0.450	0.427	0.393	0.665
Coupon proneness (CP)	0.034	0.023	0.091	0.058	0.032	0.054	0.048	0.092	0.102
Online reviews (OR)	0.054	0.030	0.042	0.084	0.077	0.068	0.048	0.079	0.073	0.702

. If the values in the HTMT ratio table are less than 0.85, discriminant validity can be considered met [100]. As shown in Table 5, the highest HTMT ratio value in this study is 0.703, which indicates that there are no issues with the discriminant validity of each construct in our model.

Finally, the risk of common method bias (CMB) was further evaluated using Harman’s single-factor analysis [102]. All measurement items were subjected to an exploratory factor analysis without rotation, and the results showed that the first extracted factor explained 30.68% of the total variance. Since this value is lower than the 50% threshold, it implies that CMB is unlikely to be a significant concern in this study.

5.2. Structural Model Evaluation

To evaluate the direct relationships in the structural model, this study first assessed collinearity among the constructs by calculating the variance inflation factor (VIF). In this study, the results showed that the VIF values for all constructs fell well below the acceptable threshold of 5.0 [103], ranging from 1.031 to 1.785, indicating that multicollinearity was not a significant issue.

Next, this study conducted hypothesis testing through significance testing of direct path coefficients using bootstrapping (resampling 5000 times), and the results are as shown in Figure 3 and

Table 6. Significance testing results of direct path coefficients in the structural model.

Hypotheses	Paths	Path Coefficient	t-Statistics	p-Values	Supported?
H1-1	Access convenience (AC) -> →perceived usefulness (PU)	0.142	2.761	0.006	Supported
H1-2	Search convenience (SC) → perceived usefulness (PU)	0.128	2.405	0.016	Supported
H1-3	Transaction convenience (TC) → perceived usefulness (PU)	0.165	2.958	0.003	Supported
H1-4	Decision-making convenience (DC) → perceived usefulness (PU)	0.147	2.933	0.003	Supported
H1-5	Benefit convenience (BC) → perceived usefulness (PU)	0.149	2.851	0.004	Supported
H1-6	Post-benefit convenience (PC) → perceived usefulness (PU)	0.068	1.286	0.198	Not Supported
H2-1	Access convenience (AC) → intention to use (IU)	0.16	4.438	0.000	Supported
H2-2	Search convenience (SC) → intention to use (IU)	0.139	3.767	0.000	Supported
H2-3	Transaction convenience (TC) → intention to use (IU)	0.171	5.077	0.000	Supported
H2-4	Decision-making convenience (DC) → intention to use (IU)	0.165	4.958	0.000	Supported
H2-5	Benefit convenience (BC) → intention to use (IU)	0.198	5.743	0.000	Supported
H2-6	Post-benefit convenience (PC) → intention to use (IU)	0.077	2.213	0.027	Supported
H3	Perceived usefulness (PU) → intention to use (IU)	0.176	5.243	0.000	Supported
H4	Intention to use (IU) → usage behavior (UB)	0.572	22.403	0.000	Supported

. This study also assessed the practical significance of standardized path coefficients based on Cohen’s [104] widely accepted guidelines, where coefficients around 0.10 are considered small, around 0.30 medium, and 0.50 or higher large effects. Among the observed paths, the one from perceived usefulness to intention to use (β = 0.572) stands out as a large effect, indicating substantial practical implications for influencing users’ behavioral intentions. By contrast, although the path from post-benefit convenience to intention to use (β = 0.077) reached statistical significance, it falls within the small-effect range, so it may have limited practical relevance in guiding managerial strategies. Beyond the above, this study also verified the 95% bias-corrected bootstrap confidence intervals of all structural paths, confirming that significant effects did not include zero in their respective intervals, thereby mitigating the risk of inflated type I errors due to multiple testing. As such, these results collectively indicate that all hypotheses were supported except for H1-6, which examined the relationship between post-benefit convenience and perceived usefulness. Specifically, five dimensions of service convenience—namely, access, search, transaction, decision-making, and benefit convenience—significantly enhanced users’ perceived usefulness of OFD apps. Additionally, although post-benefit convenience did not show a significant correlation with perceived usefulness, all six dimensions still exhibited a positive association with users’ intention to use. In addition, perceived usefulness also had a significant positive association with the intention to use, and the intention to use, in turn, had a significant relationship with actual usage behavior.

Next, this study also conducted additional mediation analysis to gain a deeper understanding of how the various dimensions of service convenience are associated with users’ intention to use OFD apps. As a result, as shown in

Table 7. Results of mediation (indirect effect) analysis.

No.	Paths	Indirect Effect	t-Statistics	p-Values	Mediating Results
1	Access convenience (AC) → perceived usefulness (PU) → intention to use (IU)	0.025	2.368	0.018	Partial mediation
2	Search convenience (SC) → perceived usefulness (PU) → intention to use (IU)	0.023	2.220	0.028	Partial mediation
3	Transaction convenience (TC) → perceived usefulness (PU) → intention to use (IU)	0.029	2.581	0.010	Partial mediation
4	Decision-making convenience (DC) → perceived usefulness (PU) → intention to use (IU)	0.026	2.480	0.013	Partial mediation
5	Benefit convenience (BC) → perceived usefulness (PU) → intention to use (IU)	0.026	2.431	0.015	Partial mediation
6	Post-benefit convenience (PC) → perceived usefulness (PU) → intention to use (IU)	0.012	1.276	0.202	None

, perceived usefulness significantly mediated the association between the five conveniences of access, search, transaction, decision, and benefit convenience and the intention to use OFD apps. On the other hand, perceived usefulness did not mediate the relationship between post-benefit convenience and intention to use. This result implies that users may not be particularly interested in functions related to order cancellation or modification, possibly because such situations occur infrequently.

Then, to further explore the factors that strengthen the cognitive mechanisms underlying consumers’ acceptance of OFD apps, this study conducted a moderation analysis by examining coupon proneness and online reviews as moderating variables. In this analysis, all predictor and moderator variables were mean-centered before the interaction term construction to reduce multicollinearity [105]. As a result, it was found that coupon preference and online reviews had significant associations with the relationships between perceived usefulness and usage intention (H5-1 and H5-2) and between usage intention and usage behavior (H6-1 and H6-2) (see

Table 8. Results of moderating effects analysis.

Hypotheses	Paths	Path Coefficient	t-Statistics	p-Values	Supported?
H5-1	Coupon proneness (CP) x × perceived usefulness (PU) → intention to use (IU)	0.059	2.047	0.041	Supported
H6-1	Coupon proneness (CP) × intention to use (IU) → usage behavior (UB)	0.104	2.460	0.014	Supported
H5-2	Online reviews (OR) × perceived usefulness (PU) → intention to use (IU)	0.060	2.072	0.038	Supported
H6-2	Online reviews (OR) × intention to use (IU) → usage behavior (UB)	0.095	2.362	0.018	Supported

).

These moderating effects were further examined through slope analyses conducted at three different levels (−1 SD, mean, and +1 SD), as recommended by Dawson [106]. As shown in Figure 4a,b, the interaction slopes become noticeably steeper at higher levels of coupon proneness and online review quality compared with the lower levels. This pattern indicates that both moderators significantly moderate the relationship between perceived usefulness and intention to use. These findings support H5-1 and H5-2. This suggests that consumers who are more responsive to coupons and those who have a high level of satisfaction with online reviews tend to be more strongly influenced by their perception of the usefulness of OFD apps when forming their intention to use them. Similarly, the interaction slopes shown in Figure 4c,d also exhibit a steeper upward pattern at higher levels of coupon proneness and satisfaction with online reviews compared with other levels. This indicates that both moderators significantly moderate the relationship between intention to use and actual usage behavior. These results provide support for H6-1 and H6-2. These findings suggest that consumers who are more responsive to coupons and those with higher satisfaction toward online reviews are more likely to translate their intention to use OFD apps into actual usage behavior.

Subsequently, the R² values were further examined to evaluate the model’s explanatory power. As shown in Figure 2, the R² values were 0.350 for perceived usefulness, 0.692 for intention to use, and 0.386 for usage behavior, all of which exceed the threshold of 0.35, indicating that the model exhibits acceptable explanatory power [97].

Finally, to ensure the statistical validity of the model and to address potential concerns about type II errors, a post hoc power analysis was conducted using G*Power (v. 3.1.9.7). According to Cohen’s [104] guidelines, detecting a very small effect size (f² = 0.004) with 80% power at α = 0.05 would require a minimum of 395 participants. This f² value represents the smallest effect size among the structural paths examined in the model. Since this study included 478 valid respondents, the sample size was sufficient to detect even the weakest effects, thereby further supporting the robustness and statistical validity of the structural model results.

6. Discussion and Conclusions

The primary purpose of this study is to deeply investigate the cognitive mechanism by which consumers make ordering decisions on OFD apps. To this end, this study adopts the TAM [32] as its theoretical foundation and examines how various dimensions of convenience—namely, access, search, transaction, decision-making, benefit, and post-benefit—are associated with consumers’ perceived usefulness and intention to use. In addition, this study also examines the moderating effects of coupon proneness and online reviews on the relationships between perceived usefulness and intention to use, as well as between intention to use and actual usage behavior, in the context of OFD.

As a result of the analysis, all hypotheses presented in this study, except for hypothesis H1-6 (i.e., post-benefit convenience has a positive association with perceived usefulness), were supported. Based on these findings, the study provides answers to the two RQs presented in the introduction. In response to RQ1, “What is the cognitive mechanism by which service convenience is associated with consumers’ intentions to use OFD apps?”, this study demonstrated that five dimensions of convenience—namely, access, transaction, decision-making, benefit, and search convenience—are positively associated with consumers’ perceived usefulness of OFD apps. Furthermore, this study empirically confirmed that all six dimensions of convenience, including post-benefit convenience, have a significant association with users’ intention to use OFD apps. In addition, this study also found that perceived usefulness mediates the relationship between five of the convenience dimensions (excluding post-benefit convenience) and users’ intention to use. Finally, this study further confirmed that consumers’ intentions to use OFD apps has a significant association with their actual usage behavior. In response to RQ2, “What factors can further promote consumers’ intentions and behaviors to use OFD apps?”, this study investigated the moderating roles of coupon proneness and online reviews. The findings revealed that both factors exerted a significant positive moderating effect on the relationship between perceived usefulness and usage intention, as well as on the relationship between usage intention and actual usage behavior.

The findings of this study emphasize that service convenience plays a critical role in consumers’ acceptance of OFD services. These results align with previous studies that have highlighted the importance of convenience in enhancing consumers’ perceived usefulness and engagement with technology [15,17,18,56,107,108]. Ultimately, to derive valuable insights from the specific findings of this study, the following section discusses its practical implications and theoretical contributions.

6.1. Practical Implications

First, this study confirmed that the five dimensions of service convenience, namely, access (H1-1, H2-1), search (H1-2, H2-2), transaction (H1-3, H2-3), decision-making (H1-4, H2-4), and benefit convenience (H1-5, H2-5), have significant associations with both perceived usefulness and users’ intention to use OFD apps. These findings suggest that improvements in these dimensions can enhance users’ perceptions of the OFD apps’ usefulness and strengthen their intention to use them, which in turn is associated with an improved user experience and promotes continuous usage behavior of OFD apps. On the other hand, the hypothesis on the relationship between post-benefit convenience and perceived usefulness (H1-6) was not supported, which can be explained by the nature of post-benefit convenience as a reactive rather than a proactive factor. According to decision-making and technology acceptance theories, perceived usefulness is primarily shaped by pre-use and during-use experiences that enhance goal achievement and interaction efficiency [32,34]. This is because post-benefit services are typically engaged only after a transaction is complete or a problem arises, making them less influential in users’ initial utility evaluations [80]. Furthermore, from a behavioral economics perspective, such post-benefit services serve as loss-mitigation mechanisms rather than benefit-enhancing features and thus may not substantially elevate perceived usefulness [109]. However, given post-benefit convenience still has a positive association with users’ intention to use (H2-6) in this research, it should still be taken into account in strategic planning aimed at sustaining engagement. In other words, the role of post-benefit in fostering long-term customer relationships should not be overlooked. The following paragraphs discuss ways to enhance each service convenience dimension, aiming to increase consumers’ perception of the usefulness of OFD apps and their intention to use them.

• Access Convenience

To enhance access convenience, OFD apps should offer users multiple input methods, such as keyboards and voice commands, to facilitate seamless interaction. Additionally, incorporating accessibility features like screen readers can help support users with visual impairments or other disabilities, thereby aiding in the creation of a more inclusive environment. Additionally, it is also essential to ensure seamless functionality across various devices and network conditions. In particular, minimizing login barriers and reducing loading times can also enhance consumers’ first impressions and reduce their frustration, ultimately encouraging continued use of the app.

• Search Convenience

To enhance search convenience, OFD apps should adopt image recognition technology that allows users to upload pictures of food or menu items for quicker and more intuitive searches. Moreover, OFD apps can make it easier for users to find what they want by improving the organization of information, such as through the use of smart filters, personalized suggestions, and clear menu layouts. This helps users locate desired items with minimal effort, thereby improving their overall decision-making experience.

• Transaction Convenience

To enhance transaction convenience, OFD apps should offer a wide range of payment options, including credit cards, mobile wallets, QR codes, and cash-on-delivery, to cater to diverse user preferences while simplifying the overall ordering process. Additionally, integrating location-based promotions or automatically applying available discounts can streamline the checkout process, creating a more satisfying and efficient user experience.

• Decision-Making Convenience

To enhance decision-making convenience, OFD apps should help users make better choices by offering clear nutritional information; showcasing healthy meal options; and allowing filters for specific dietary needs, such as vegan, low-sodium, or gluten-free. These features can empower users to make informed decisions and build trust by catering to individual health concerns and lifestyle preferences.

• Benefit Convenience

To enhance benefit convenience, OFD apps should focus on maintaining food quality during delivery by adopting smart packaging and effective temperature control solutions. At the same time, offering customized incentives and ensuring consistent delivery performance can help users feel they are receiving greater value, thereby enhancing the actual benefits they gain from using the apps.

• Post-Purchase Convenience

To enhance post-purchase convenience, OFD apps should establish robust customer service systems that can respond quickly to complaints and provide effective solutions. OFD apps should also consider offering follow-up guidance, such as meal reheating tips, or utilizing customer data to deliver personalized after-service offers, which can increase satisfaction and encourage continued engagement.

Second, this study confirmed that perceived usefulness has a positive association with customers’ intention to use (H3), which, in turn, is related to their actual usage behavior (H4). These findings suggest that OFD companies should take active steps to enhance users’ perceived usefulness, thereby strengthening both their behavioral intention and platform engagement. The specific ways in which these efforts should be directed have been discussed just above. In addition to enhancing those conveniences, the following measures should also be considered. For example, OFD apps should demonstrate how their services help consumers save time and reduce daily decision-making burdens, thus reinforcing the perceived practicality of the OFD apps. In particular, OFD platforms can enhance perceived usefulness by sending personalized notifications at appropriate times. For instance, delivering targeted promotional messages during peak meal hours can effectively remind users of the app’s utility in streamlining daily meal planning. Moreover, by analyzing user data, OFD apps can provide customized recommendations tailored to individual preferences, which can further enhance consumers’ recognition of the app’s practical value. In addition, OFD apps are encouraged to provide short tutorial videos or onboarding guides to help new users understand key features more efficiently, thereby lowering the learning barrier and increasing immediate perceived usefulness. Beyond the above, by regularly collecting feedback through satisfaction surveys or in-app channels, OFD apps can continuously improve their services to align with user expectations, ultimately fostering long-term engagement and consistent usage behavior.

Third, with the confirmation of H5-1 and H6-1, this study revealed that coupon proneness plays a positive moderator between perceived usefulness and intention to use, as well as between intention to use and actual usage behavior. These results suggest that OFD platforms should actively adopt promotional strategies that utilize discount or reward coupons to increase users’ usage intentions and actual usage behaviors of OFD apps. To this end, OFD apps should consider incorporating games or quizzes into the coupon distribution process, enabling users to earn rewards in a fun and interactive way. By gamifying the otherwise routine act of ordering food, such features can inject a sense of enjoyment, anticipation, and engagement into the experience, ultimately enhancing the overall dining experience, even before the food is delivered. Additionally, OFD apps should consider hosting events centered on food experiences, voucher distribution, and coupon-sharing activities to enhance user engagement further. Incorporating donation programs—where users can contribute to charitable causes by using coupons—can also help foster a socially responsible brand image, thereby improving the overall reputation of the OFD app. In this regard, OFD apps can draw on successful practices from major e-commerce giants (e.g., Alibaba’s Ant Forest in China and Coupang’s Smile Donation in Korea), which combine coupon-based promotions with charitable contributions. Such initiatives not only appeal to socially conscious users but also enhance brand reputation and user engagement. Beyond targeting coupon-sensitive users, it is also crucial for OFD apps to encourage interest among users who were previously uninterested in coupons. After all, the more users become responsive to coupon-based promotions, the greater the potential impact of these strategies will be. In doing so, OFD apps can consider offering rewarding and straightforward tasks, such as daily check-ins or fundamental challenges, that allow users to earn coupons easily. Such efforts can foster greater user engagement while simultaneously encouraging users to recognize the value of coupons, which potentially results in more favorable attitudes and increased promotional effectiveness. Consequently, all these approaches help users recognize the additional value provided by the OFD app, thereby strengthening their intention to use, as well as their actual usage behavior.

Fourth, with the confirmation of H5-2 and H6-2, this study demonstrates that online reviews serve as a positive moderator between perceived usefulness and intention to use, as well as between intention to use and actual usage behavior. This suggests that OFD companies should prioritize managing online reviews while implementing effective promotional strategies to enhance their online presence. In today’s world, consumers often rely on online reviews written by previous customers before trying new restaurants when using OFD apps because these reviews typically offer valuable insights into food quality, service experience, and overall satisfaction [27]. Considering this growing reliance on online reviews for consumer decision-making, OFD companies should proactively manage customer reviews to protect and enhance their reputation. Positive reviews can help attract new customers, while ignoring negative feedback can harm a company’s credibility and deter potential customers from trying their services [92]. Moreover, as the volume of online reviews increases, OFD companies should also consider utilizing automatic review analysis tools to efficiently process customer feedback, which can quickly identify service issues and pinpoint areas for improvement in real time. Such proactive management of reviews is crucial for building a robust image of OFD apps and maintaining a competitive advantage in the market.

6.2. Theoretical Contributions

First, this study contributes by proposing a theoretical framework that explains the cognitive mechanisms underlying users’ acceptance of IT systems for online ordering services. In particular, the proposed framework incorporates clearly defined and independent dimensions of convenience as key factors, enabling a deeper understanding of consumers’ cognitive decision-making processes. Such understanding plays a crucial role as it provides valuable insights for developing effective strategies to enhance consumers’ purchase intentions. Although this framework was developed in the context of OFD service apps, many services within the broader category of online ordering share common attributes. Therefore, the framework proposed in this study is expected to provide a valuable theoretical basis for future research on various types of online services as well.

Second, this study provides empirical evidence that coupon proneness and satisfaction with online reviews positively moderate the relationships between perceived usefulness and usage intention, as well as between usage intention and actual usage behavior. These findings clarify the conditions under which users are more likely to translate their perceptions and intentions into concrete usage behavior. By applying such moderating variables, this study contributes to a more structured theoretical understanding of how consumer perceptions are translated into behavior in the context of OFD services.

6.3. Limitations and Future Research Directions

First, the six convenience dimensions in the research model were developed as reflective constructs based on the prior literature. However, the conceptual proximity among some of these dimensions might raise concerns about their structural independence. This suggests potential conceptual overlaps, which may limit the distinctiveness and explanatory power of each construct. To address this issue, future research may consider employing alternative modeling strategies, such as adopting a second-order factor structure or conducting an exploratory factor analysis (EFA), to validate discriminant validity further and enhance both the parsimony and theoretical robustness of the model.

Second, this study primarily focused on individual-level cognitive factors but did not sufficiently account for broader social, contextual, and personal characteristics that may act as potential confounders. Specifically, it overlooked influences such as users’ prior OFD experience, digital literacy, cultural attitudes toward convenience, and external social environments like peer recommendations or situational usage contexts (e.g., family meals, group orders, or workplace lunches). Future research should consider incorporating these contextual factors to gain a better understanding of how various contexts influence the use and interaction with OFD apps.

Third, although this study identifies a significant relationship between intention and behavior, it does not fully establish their temporal ordering, leaving the possibility of reverse causality open. Usage behavior is measured retrospectively (e.g., “over the past 6 months”), while intention is assessed prospectively, which introduces a potential temporal mismatch. Future longitudinal or experimental designs are recommended to clarify the directionality of this relationship.

Fourth, this study measured usage behavior through self-reported Likert-type agreement scales. While this approach has been widely adopted in prior studies, it may not fully capture the objective frequency of user behavior. As highlighted by Podsakoff et al. [110], using agreement scales to assess factual behavior may introduce measurement bias and compromise construct validity. Future studies should consider adopting more objective behavioral metrics (e.g., usage logs and frequency-based scales) and disaggregating complex actions into distinct items to improve construct clarity and external validity.