Understanding the Dynamics of e-WOM in Food Delivery Services: A SmartPLS Analysis of Consumer Acceptance

Abstract

The evidence on e-commerce rapidly adopting and expanding has led to a high degree of changes in consumer behavior, with food delivery platforms playing a pivotal role in the digital economy. The current study highlights the interplay between consumer perception, consumer acceptance, and electronic word of mouth (e-WOM) assessed in the context of the food delivery chain. By integrating latent variables, such as trust, consumer purchase intentions, perceived usefulness, and psychological distance, the current study explores the role of e-WOM in influencing consumer perception and acceptance within the food delivery services. Data collection and analysis implied a qualitative approach by issuing an online survey and gathering 835 valid responses, further quantitatively assessed through structural equation modelling. The data reveal the validity of the two general hypotheses, according to which consumer acceptance is positively connected to consumer perception, to which the influence of electronic word of mouth was added. The current study stresses the importance for food delivery chains to gain insight and implement strategies based on its electronic peers’ online suggestions and recommendations. The current research actively contributes to the theoretical discourse on food delivery chain management related to the electronic consumers’ experience, in order to competitively position it on electronic markets as well as the food delivery global industry.

1. Foreword

Within e-commerce transactions, mainly regarding food delivery chains, consumer loyalty has been found to be difficult to achieve, primarily due to the fact that when competition is a click away, switching costs become low [1]. The food delivery chains’ solution was to launch loyalty programs that provide consumers with a large array of benefits when conducting online shopping activities.

The current research builds on established theories, particularly the technology acceptance model (TAM) and expectation–confirmation theory, to explore consumer perception, acceptance, and e-WOM within food delivery platforms [2,3]. By incorporating psychological distance as a mediating factor, the study expands these frameworks, providing deeper insights into how these variables interact and influence consumer engagement and loyalty.

While previous research has examined consumer perception, acceptance, and electronic word of mouth (e-WOM) individually, there is a noticeable lack of studies that consider how these factors interact within the context of digital food delivery platforms. Understanding this relationship is essential for boosting engagement and encouraging customer loyalty.

Although the broader field of e-commerce has been extensively explored, food delivery services represent a more specialized area, in which comprehensive analysis combining acceptance, perception, and e-WOM is still lacking. Bridging this gap could enable the creation of more precise strategies to improve platform performance and user experience.

Furthermore, the concept of psychological distance—the perceived separation between consumers and online platforms—has not been sufficiently addressed. Few studies have investigated how tailored design, personalized communication, and relevant content can help narrow this gap, ultimately enhancing user engagement with food delivery applications.

As resulting from the identified gaps, the following secondary research objectives arose: RO1 investigates the interdependent relationship between consumer perception, consumer acceptance, and electronic word of mouth (e-WOM) in digital food delivery platforms. RO2 explores how psychological distance moderates the relationship between consumer perception and acceptance in the context of food delivery services, while RO3 assesses the role of e-WOM in fostering trust, reducing perceived risk, and enhancing consumer loyalty in food delivery platforms.

The food delivery industry has experienced rapid digital transformation driven by evolving consumer expectations and technological advancements. However, consumer perception, acceptance, and e-WOM remain underexplored in this niche sector compared to broader e-commerce. As research in e-commerce frequently examines consumer perception and acceptance as separate constructs, the following RQ (1) arose: how does consumer perception influence consumer acceptance and e-WOM in digital food delivery platforms?

However, limited attention has been given to how these factors interact within the context of food delivery services. While studies highlight the importance of user experience and convenience, the direct link between perception and acceptance, as well as the influence of perception on e-WOM, remains underexplored.

Understanding this relationship is essential, as consumer perception often shapes engagement with digital services. In food delivery, a positive perception of app usability, service reliability, and platform security can foster acceptance and encourage consumers to share their experiences online. This engagement, in turn, enhances brand visibility and trust.

From a theoretical perspective, this research builds on the technology acceptance model (TAM) and expectation–confirmation theory (ECT), providing insight into how perception influences both acceptance and word-of-mouth promotion.

Psychological distance, referring to the perceived separation between consumers and digital platforms, plays a crucial role in engagement and decision making. Although previous studies have explored psychological distance in broader e-commerce contexts, there is limited research examining its role within food delivery services, thus creating the premise for the second RQ (2): what is the role of psychological distance in mediating consumer engagement and acceptance in food delivery apps?

Reducing psychological distance through personalized communication, intuitive design, and localized content can create a stronger connection between the consumer and the platform. This can lead to higher engagement and the increased acceptance of the service.

This research question draws on construal level theory (CLT), which suggests that the closer a consumer feels to a service or product, the more likely they are to engage with it. Addressing this gap may offer valuable insights for food delivery platforms seeking to enhance user interaction and satisfaction.

The impact of electronic word of mouth (e-WOM) on consumer trust and purchase intentions is well-documented across the retail and hospitality sectors. However, the specific influence of e-WOM on food delivery services has received less attention. Given the unique nature of food delivery—where factors such as speed, reliability, and quality significantly affect consumer decisions—understanding the role of e-WOM is essential.

Positive e-WOM, in the form of reviews, ratings, and recommendations, can enhance trust in the platform, reinforcing perceived usefulness and increasing the likelihood of repeat purchases. This aligns with risk-reduction theory, which emphasizes the role of peer reviews in mitigating uncertainty and fostering confidence in online services.

By exploring this question, the research contributes to existing theories, while offering practical guidance for food delivery services aiming to strengthen consumer trust and drive loyalty.

Moreover, RQ (3) (how does e-WOM contribute to consumer trust, perceived usefulness, and purchase intentions in food delivery services?) highlights the fact that the impact of electronic word of mouth (e-WOM) on consumer trust and purchase intentions is well-documented across the retail and hospitality sectors. However, the specific influence of e-WOM on food delivery services has received less attention. Given the unique nature of food delivery—where factors such as speed, reliability, and quality significantly affect consumer decisions—understanding the role of e-WOM is essential.

Based on the fact that the food delivery industry is evolving at a fast pace, propelled by technological innovations and changing consumer expectations, the relationship between consumer perception, acceptance, and electronic word of mouth (e-WOM) within this sector remains insufficiently explored. Addressing this gap presents an opportunity for both theoretical development and practical progress.

This research seeks to provide a more comprehensive understanding of the factors that influence consumer engagement, trust, and loyalty on food delivery platforms. By examining under-researched elements, such as psychological distance, the study aims to build on existing models and provide actionable insights for those involved in the industry.

Although the individual dimensions of consumer acceptance, e-WOM, and consumer perception have been well explored in prior research, this study introduces novelty by examining their combined and integrated effect. Specifically, the focus is on how these constructs interact to influence consumer behavior within the food delivery sector, an area that has been largely overlooked. Much of the existing literature addresses these variables separately, without considering the complex relationships that drive consumer decisions in the food delivery market. By investigating the connections between these constructs, this study provides a fresh perspective on how food delivery platforms can utilize e-WOM and consumer perception to enhance consumer acceptance and, in turn, improve their market performance. This approach contributes to bridging a gap in the literature by offering a more holistic understanding of consumer dynamics within this growing industry) [4,5].

To answer the previous questions, a thorough discussion of the evolution of electronic commerce and its implications for food delivery chains was deemed necessary. Moreover, since website and/or app interfaces are a direct result of online buyers’/shoppers’ active reviews, highlighting the evolution of consumer interface quality was considered essential, given the novel and rapid technological advancements. Further, an assessment of consumer engagement (or acceptance) within electronic commerce—understood in the form of interactive marketing disclosure—relies on a series of activities that start with the electronic retailer–shopper data exchange (through browsing, selecting one or more products/services, comparing prices, gathering supplementary data, and making a (un)selection decision). The current study emphasizes the importance of electronic word of mouth, explaining the lack of interaction between the involved parties—online buyers/shoppers (as the information source), who act with an unknown identity, and the web retailer (the recipient). Methodological aspects reveal a complex database, analyzed using SmartPLS software (v.4.1.0.9.), employed to test general and comprehensive hypotheses, highlighting the contextual importance of current developments and, thus, providing data support for the proposed research question. After performing a wide variety of tests on the defined database, results and hypotheses were discussed in light of the previous research. The manuscript concludes with a discussion on the practical and theoretical implications, followed by strategies and limitations in the final remarks.

To enhance the understanding of the contextual analysis, a conceptual map has been included (see Figure 1).

From Consumer Perception to Consumer Interface Quality

In the context of digital food delivery platforms, the integration of consumer perception, consumer acceptance, and electronic word of mouth (e-WOM) is pivotal for enhancing engagement and fostering consumer loyalty. The existing literature has explored these constructs in various online environments, but their combined influence within the food delivery sector remains underexplored [6,7]. This gap presents an opportunity to understand how consumer interface quality, particularly the elements of convenience, interactivity, and perceived security, contributes to the success of e-WOM strategies.

While factors like the perceived ease of use and trust have been widely recognized as important for consumer acceptance in e-commerce [8], the specific context of food delivery services requires further investigation. Previous studies [9] have shown that online convenience and real-time tracking can enhance consumer satisfaction, yet the interplay between e-WOM and these variables in the food delivery sector remains insufficiently studied.

This study aims to bridge this gap by investigating how consumer perception influences consumer acceptance and e-WOM within food delivery platforms. Specifically, it explores how consumer interface quality, as defined by components such as customization, character, and perceived security, impacts the e-WOM process, which, in turn, affects consumer trust and loyalty [10]. The research also seeks to determine the role of psychological distance, a factor rarely considered in this context, in moderating consumer engagement and e-WOM influence in food delivery apps [11,12].

By addressing these factors, this research offers practical implications for food delivery platforms, enabling them to refine their service offerings and leverage e-WOM effectively to enhance customer loyalty and retention.

The current study acts as a theoretical bridge that addresses how electronic word of mouth relates to consumer perception and acceptance, mainly focusing on psychological variables, such as consumer trust, perceived risk, and psychological distance.

The empirical analysis integrates previous and emerging theoretical frameworks, aiming to provide a new perspective for understanding the dynamic relationships governing consumer behavior within e-commerce. Such an understanding of this relationship advances academic discourse as well as food delivery chain strategies that could result in enhanced consumer satisfaction, engagement, and online loyalty.

In the case of measuring consumer interface quality, as buyers’ perception regarding the quality of the pre- and post-food delivery process, the four-dimensional scale presented in [13], along with the five-dimensional eTransQual scale presented in [14], is recommended for consideration.

It is important to note that, from the perspective of an online store, the consumer interface plays the role of external relations by fostering a web environment conducive to a wide range of emotional effects on potential buyers and/or visitors, thereby increasing the website’s likelihood of achieving a positive result—making a sale.

Taking into consideration the aforementioned concerns, the current study seeks to assess the recommended components of consumer interface quality [15], particularly as they relate to online business experience [16] within food delivery chains. For this reason, the following four components are considered: convenience, interactivity, customization, and character [17].

Convenience, also viewed as a user-friendly characteristic, refers to a buyer/consumer’s ease of access to a given website. The literature suggests that more than two-thirds of e-commerce transactions fail to reach a positive outcome due to the lack of access to necessary information.

Interactivity relates to two-way communication (website–e-consumer/delivery). While in physical stores, this interaction is facilitated by staff involvement, in web stores [18], it is replaced by interface features that result in either a friendly or unfriendly buyer/shopper experience. According to the literature, within virtual commerce, consumer responses are influenced by website architecture and ease of use, with the primary result being a strong desire for e-shoppers to return to the food delivery provider.

As for the customization dimension, this refers to a website environment tailored to products and services as required by e-shoppers, thereby increasing the store’s probability of completing at least one transaction with each visitor [19]. This website appeal enables e-shoppers to find at least one product/service of interest and persuades them to complete the purchase immediately.

The character dimension refers to the general image projected by food delivery businesses through their websites. This involves marketing techniques such as graphics—fonts, colors, and patterns—that not only enhance the clarity and comprehension of displayed information but also create a pleasurable and safe experience for visitors to an unfamiliar website.

Perceived security in e-commerce is considered integral to the entire consumer experience, encompassing the security of transactions and personal information. This primarily refers to mechanisms safeguarding personal data as well as payment methods [20]. Perceived security can lead to negative experiences for both sellers and buyers due to shared or personal perceptions of risks associated with storing and handling sensitive information, such as card numbers, passwords, and identity details [21]. Such concerns significantly impact online consumer behavior in relation to both general and specialized online (delivery) stores [22].

For the purposes of this analysis, perceived security is defined by online shoppers’ perceptions of e-security concerning sensitive information, largely influenced by their personal intuition in assessing and mitigating risk.

As for consumer satisfaction, often assessed within literature as a marketing goal in relation to consumer behavior, it mainly refers to online shoppers’ affective response regarding a specific order [23]. Final and overall satisfaction may only be determined through studying and assessing approaches related to a specific transaction (as the most recent online transaction experience and emotional reaction) and/or overall consumer satisfaction, as a particular experience with a given website/e-seller over a period of time. Despite other literature views that assess consumer satisfaction in terms of perceived website service quality, for the development of the current study, consumer satisfaction was only referred to as a cumulative factor.

Across multiple literature views regarding the switching costs dimension, the current research adopted the perspective presented in [24], who assessed this dimension using a three-fold approach—procedural, financial, and relational. While the first view refers to the buyers’ time and effort, the financial component considers only pecuniary losses, while the final dimension considers relationships on personal/business levels, reflecting the psychological effects deriving from broken affiliations [25,26]. The current perspective shifts switching costs from a singular economic view to adopting a second component that incorporates emotions, assessed in a psychological manner [27]. For the development of the current study, the switching costs dimension involves online buyers’ perceptions regarding elements such as temporal, financial, and physical effort, assessed when an online shopper decides to change traditional online product/service providers.

Traditionally defined as repeated online purchase behavior, the loyalty behavioral components focused primarily on accurately predicting repurchase intentions and actual rates. Since such a view did not allow for the assessment of traditional and/or spur-of-the-moment e-consumer behaviors, recent literature trends have reconsidered the consumer loyalty dimension, expanding its interpretation towards an attitudinal construct. For this reason, in the development of the current research, consumer loyalty was defined as a favorable attitude manifested by e-shoppers towards a particular website that possesses the necessary elements to create consumer predispositions to repeated purchase behavior.

2. Food Delivery Chains and Consumer Acceptance

The interplay between consumer perception, acceptance, and e-WOM in the food delivery sector remains underexplored. While these concepts are individually well-studied, their combined influence on consumer engagement and loyalty within food delivery platforms has received limited attention [28]. This study aims to fill this gap by examining how e-WOM and consumer acceptance interact to influence purchase decisions and loyalty.

Research indicates that e-WOM plays a key role in building trust and encouraging purchases [29]. However, the relationship between consumer perception and e-WOM in food delivery services is still unclear, despite its significance in driving consumer behavior. Additionally, while psychological distance—the perceived gap between consumer and service—has been shown to affect engagement11, its role in food delivery platforms is underexplored.

This study also integrates the technology acceptance model (TAM) and expectation–confirmation theory to examine how perceived ease of use and usefulness influence consumer acceptance [30]. By doing so, the research seeks to provide insights into how food delivery platforms can leverage trust, ease of use, and e-WOM to enhance consumer satisfaction and foster loyalty.

Within the food delivery sector, consumer engagement (or acceptance) within electronic commerce is generally understood as interactive marketing disclosure [28], primarily involving a series of activities that begin with the electronic retailer–shopper data exchange (through browsing, selecting one or more products/services, comparing prices, gathering supplementary data, and making a (un)selection decision). The second step involves the electronic buyer sharing sensitive information by registering, browsing products/services according to their preferences, and providing buying decisions and feedback. This process also includes additional data exchange through data mining tools, cookies, and other means [31]. The final step pertains to financial information related to actual purchase intentions, data voluntarily provided by the buyer as a direct indication of an electronic purchase within the food delivery sector.

Regarding the intention to transact, the authors of [32] described online transactional relationships involving e-shoppers and websites, alongside the entire array of shared information and reciprocal conduct, culminating not only in the intention to transact but in a complete product/service e-purchase/delivery. This is despite the fact that recent literature emphasizes that over 75% of online shopping carts do not result in an actual online transaction [33]. This dilemma is a concern for website owners, who must invest in persuasive web interfaces to convert browsers into actual online shoppers.

When addressing e-buyers’ information exchange with an online web retailer, studies indicate that even if the transaction is not completed, the purchase intention at the moment of providing sensitive information is genuine [34]. Therefore, the dimension of intention to transact in the current study will not only refer to this moment but will encompass the entire online transaction process, regardless of the positive or negative outcome for the web retailer. As such, the intention to transact may include cart selection, sensitive information exchange, and/or product/service purchase, generally regarded as a singular action.

This decision regarding intentions and actions to transact within the online environment has been extensively discussed in two theories—the theory of reasoned action and the theory of planned behavior [35,36]. As both traditional and online commerce retailers have increasingly adopted low-cost internet infrastructure as a widely accepted means of reaching a larger consumer pool, it has been demonstrated that the latter do not always engage with web commerce in return. This lack of online engagement stems from a lack of trust in the security of personal-sensitive information, making consumer uncertainty fundamental in predicting future e-commerce behavior, including e-purchase acceptance and delivery through the same channels.

Generally accepted as a core feature of both economic and social uncertainty in business–consumer interactions, trust governs all actions, particularly those that develop within an online environment. Trust is defined by two targets in electronic commerce [37,38], and this feature is highly susceptible to being influenced by a wide range of web retailers’ actions, such as firewalls, e-authentication, obligatory subscriptions, and privacy seals. These actions are dedicated to reducing infrastructure concerns. This argument aligns with Hoffman et al., who suggest that web retailers should aim to reduce infrastructure concerns to minimize environmental uncertainty and increase buyers’/online shoppers’ trust.

Regarding the use of technology and the intention to transact, it is important to note that within every proposed virtual transaction [39], the process imposes on consumers the need to interact with internet technologies, thereby defining the intention to transact and justifying the use of the technology acceptance model in predicting buyers’ intentions to use the internet for online transactions and delivery services. The model has been successfully applied across various information technology domains, including e-commerce. Previous studies, such as [40], focused on the perceived ease of internet use within the electronic commerce process. Additionally, the authors of [41] examined perceived usefulness and consumer ease of use in relation to e-commerce. The selected technology acceptance model variables can predict consumer behavior in the electronic commerce process through perceived usefulness and perceived ease of use.

Perceived usefulness refers to the extent to which a particular technology is accepted and agreed upon by users within the (food delivery) transaction processes [42]. Perceived ease of use pertains to the degree of consumer belief that a certain technology can create the perception of simplicity and/or effortlessness in transactional processes. According to previous research, there is a direct and positive relationship between these two variables, and when applied to consumer behavior within an online framework, this connection manifests as operational usability for transactions. According to the authors of [43], who extensively studied this relationship, it was concluded that, in the majority of cases, perceived ease of use is influenced by perceived usefulness through online shoppers’/buyers’ intentions to use.

Regarding the perceived risk dimension, this primarily refers to the impersonality and distance associated with the internet, in addition to the uncertainty of the global transactional infrastructure, which renders risk an implicit and tacit acceptance within e-commerce food delivery transactions. Across the literature, uncertainty is unanimously acknowledged to have both behavioral and environmental origins. Risks are caused by technology and stem from environmental risk categories or are relational and arise from the behavior of commerce partners [44]. It is important to note that both forms of risk are often intertwined, as encryptions, firewalls, or authentication mechanisms are primarily managed by web retailers, thereby placing a significant share of risk on online buyers/shoppers [45].

It is important to clarify that, since risk is difficult to capture in real environments, the literature proposes analyzing perceived risk, which is defined as a subjective belief of potential loss during the pursuit of a desire or specific outcome.

Consumer transaction intentions are contingent upon the web retailer’s perception that they are predetermined by online environmental and behavioral factors. Given these increasingly uncertain conditions, web retailers’ expectations regarding perceived risk are aimed at decreasing consumers’ intentions to engage in web transactions [46]. The relationship between perceived risk and transaction intentions is explained in the literature through perceived behavioral control, which is also part of the theory of planned behavior. Within the online environment, reducing the perceived risk associated with purchasing is expected to influence buyers’ willingness to transact using delivery platforms. Reducing risk on a website/interface or app increases the probability of online orders and purchases for web retailers [47].

According to previous research, perceived risk negatively correlates with purchase intentions, while the perceived risk associated with online purchases diminishes buyers’ perceptions of both behavioral and environmental control, thereby negatively influencing overall food delivery transaction intentions.

A web retailer’s reputation acts as an antecedent to trust and also involves risk concerning online food delivery buyers’ transaction intentions. Previous online interactions and the resulting (dis)satisfaction have a positive relationship with buyers’ future intentions, while the frequency of web shopping is considered the best indicator of the likelihood of purchase [48].

2.1. Food Delivery Chains and Electronic Word of Mouth

The relationship between consumer perception, acceptance, and e-WOM in food delivery platforms remains insufficiently explored in the literature. While previous studies have addressed these concepts individually, the combined influence of e-WOM on consumer perception and acceptance within food delivery contexts remains underexamined [49]. This research aims to address this gap by exploring how e-WOM contributes to trust, influences purchase intentions, and drives consumer loyalty in the food delivery industry.

Studies have shown that e-WOM plays a key role in building trust and consumer engagement in online environments. However, the way in which e-WOM interacts with consumer perception and acceptance in the food delivery sector is not well understood. For instance, although e-WOM influences purchase decisions, it remains unclear how this impact is mediated by consumer perception and how psychological distance between consumers and food delivery platforms is reduced through e-WOM [50].

To bridge this gap, the current study applies the technology acceptance model (TAM) and expectation–confirmation theory [51], examining how e-WOM, trust, and perceived ease of use influence consumer acceptance in food delivery platforms. By investigating the interplay between these factors, the research aims to offer both theoretical insights and practical strategies for food delivery platforms to enhance customer engagement and loyalty through the effective use of e-WOM [3].

When considering information quality, the recent literature defines it as a means of content analysis. User needs are adequately met with factual, high-quality information [52]. In 2009, there were highlighted recommended measures for evaluating information quality [53], such as relevance, understandability, objectivity, and sufficiency. Other authors argue that high-quality information is comprehensive and representative; moreover, information quality and richness are positively correlated with usefulness [54].

In the case of food delivery chains, one research in particular indicates that information length can be assessed in terms of utility [55].

Regarding social distance, the literature defines it as the perceived space separating individuals from others [56]. The term also includes the subjectivity individuals experience when approaching a group, reflecting how people distinguish themselves from others or identify with similar individuals who use food delivery services. Close interactions with others are likely to occur when social–psychological distance is reduced, fostering subjective and familiar relationships.

Individuals are inclined to assimilate personalities, feelings, and thoughts about others, which facilitates calculations regarding social–psychological distance [11].

Unlike traditional social word of mouth, the electronic version underscores the lack of interaction between the involved parties—the online buyer/shopper (as the information source), acting as an anonymous identity, and the web retailer (the recipient of the delivery). Therefore, within the online environment, measuring or considering close, long-term interpersonal relationships is largely irrelevant [57,58].

The recent literature has explored the importance of electronic word of mouth in the context of commerce and food delivery platforms; however, no study has addressed the importance of information from the perspective of the publisher–recipient relationship.

However, relevant research previously published highlights that when recipients are provided with (ir)relevant product/service information, their perception of social and psychological distance may be altered, shifting to a greater or lesser social and psychological distance from the source, thus affecting both the recipient evaluation and the response to the published information [59].

When considering the specifications and recommendations of electronic word of mouth, the information adoption model suggests that electronic shoppers and/or buyers consider two factors: the credibility of the information source and the quality of the information [60], with the latter playing a more significant and lasting role [61]. From the perspective of a food delivery user, information quality influences their perception of persuasiveness and information sufficiency. However, this factor plays a crucial role in the persuasion process [62]. Within the web environment of food retailers, the quality of product appraisals (in the context of electronic word of mouth) visibly affects the consumer perception of reliability [4,5]. High-quality online information plays a persuasive role, generating interest among electronic buyers in reviewed networks and prompting decisions based on acquired trust levels [63,64].

In the online food delivery context, information associated with electronic word of mouth is perceived by consumers as a reliable indicator of service quality. Information quality possesses an investigative dimension, exerting an overall effect on food delivery consumers [65] and shaping their perceptions of online comments [6]. This indicates that consumer trust in electronic word of mouth is heavily influenced by information quality.

Purchase intention on e-commerce platforms is a direct result of users’ interaction and communication. Electronic word of mouth facilitates information exchange and validation between both parties. However, since information is asymmetrical, its extent leads to consumer risk regarding the confidence level associated with an electronic word of mouth source. Within this relationship, trust—both cognition- and emotion-based—forms a positive impression on e-commerce consumers [66], fulfilling the role of reducing perceived risk and sources of uncertainty [6].

Consumer intention is generated primarily by subjective feelings and desires [67], which are driven by electronic word of mouth stimuli (user feedback and information). It was previously found that cognitive trust in consumers derives from the source of confidence [68]. Strong emotional and cognitive reactions have been shown to positively influence online buyers’ purchase intentions [69]. Therefore, electronic word of mouth becomes a powerful influence in the purchase decision-making process.

The sense of power is believed to affect electronic buyers’ cognition and behavior [70]. Scholars have identified its strong connection to persuasiveness and the pre-existing trust of information buyers. However, the equation of consumer power is completed by the persuasiveness of the information [71]. Consumers with a greater sense of power are more likely to make decisions based on processed information. Thus, the higher the consumer’s sense of power and the higher the quality of the information, the more likely the consumer is to trust the displayed data.

2.2. Consumer Acceptance and Consumer Perception Within Food Delivery Chains

It is well-established that both consumer acceptance and consumer perception play a decisive role in online buyers’/shoppers’ purchase intentions, resulting in the success of products and services within a dynamic market. While acceptance primarily refers to the consumer’s willingness to adopt and opt for certain online products and their delivery, perception pertains to e-shoppers’ mental frameworks, resulting in a collective yet distinctive (word of mouth) evaluation of such options. According to generally accepted literature, online consumer acceptance is strongly connected to e-buyers’ perceived value, encompassing trust, price, and perceived quality [72].

According to the technology acceptance model (TAM) perceived ease of use and usefulness significantly impact consumer acceptance, especially for products evaluated, ordered, and received online.

Despite this interconnectedness, perception is shaped by the e-shoppers’ initial view of the product ordered, while acceptance reflects the readiness to incorporate the desired and acquired item into daily personal life. Trust plays a pivotal role in fostering positive perceptions, while risk may lower acceptance, regardless of product quality or the electronic presentation. Both trust and perceived risk serve as critical mediators of consumer perception and consumer acceptance within e-commerce processes [73,74]. For web retailers, this insight can enhance practices aimed at reducing barriers and increasing consumer acceptance through clear messaging and reliable customer service support.

By combining strategic marketing with the psychological factors influencing e-commerce shoppers, web retailers must adopt a multifaceted approach. It has been demonstrated that food delivery websites/apps that emphasize consumer needs and implement strategies to manage perceptions achieve higher scores in terms of online acceptance and usage [75,76]. Within e-commerce platforms related to food delivery, consumer acceptance and perception have garnered increased attention due to rapid technological advancements and evolving consumer behavior.

Factors such as platform usability, trust, and convenience play a significant role in consumer acceptance. Consumer acceptance is closely related to perceived ease of use, delivery schedules, and time management [77]. Additionally, consumer perception is shaped by the interface of web retailers and delivery apps, the quality of delivery services, and their reliability. Other authors note that a highly rated delivery app should facilitate accurate order processing and tracking [9], provide a seamless online experience, and incorporate user reviews (word of mouth).

A positive consumer experience also requires secure data sharing and payment systems, along with fair and transparent pricing—features that are crucial in bridging consumer perception and acceptance [78,79]. This perspective provides a framework for food delivery platforms to implement consumer-oriented designs, driving operational excellence and enhancing both consumer acceptance and perception.

Given these considerations, the following hypothesis arises:

H1. 

There is a positive connection between consumer acceptance and consumer perception.

2.3. Consumer Acceptance and Electronic Word of Mouth Within Food Delivery Chains

Amid unprecedented digital expansion, both consumer behavior and decision making are heavily influenced by electronic word of mouth (e-WOM). Without being constrained by geographical limitations, e-WOM enhances consumer convenience within food delivery and social media platforms [4,80]. The concept of e-WOM has been clarified as statements regarding a web retailer, platform, or app and its products/services, generally available to past, present, and future internet users [81,82]. Such initiatives have a definitive impact on consumer purchase intentions, the image of web retailers, and consumer trust. The free-access nature of web platforms and internet sites enables any online shopper or buyer to share experiences, personal opinions, and recommendations, fostering an environment conducive to developing a complete e-ecosystem.

Effectiveness is a dimension highly influenced by consumer trust in shared food delivery platforms. The relevance of the source and reliance on peer honesty positively affect perceived credibility [83], which explains why anonymous comments carry less weight compared to those from well-known online or real-world peers, especially specialists in a particular retail area. Message persuasiveness resides in the length and accuracy of narratives, along with the variety of message content [84]. Contrary to common misconceptions, business success does not inherently foster consumer loyalty. Instead, credible and engaging peer comments serve as brand advocacy, resulting in business success.

The impact of e-WOM predominates in shaping electronic shoppers’ purchase decisions, influencing elements such as sales, website interfaces, and app functionality [4]. While generally negative e-WOM can harm a brand, companies can use such feedback constructively to improve consumer interfaces and increase transparency. The dissemination of e-WOM is actively driven by social media, facilitating rapid consumer–food delivery platform interactions and message absorption [85,86].

With longstanding roots in the physical world, e-WOM has created the framework for fast and healthy interactions between consumers and food delivery services, resulting in mutually beneficial relationships. For this reason, companies must adopt strategic actions aimed at fostering positive and trustworthy consumer experiences by seeking constant feedback and responding constructively.

Based on these premises, the following hypothesis arises:

H2. 

There is a positive connection between consumer acceptance and electronic word of mouth.

The current work addresses gaps in the literature and offers practical implications for food delivery services, as the further literature overview suggests (see

Table 1. A snapshot into e-WOM literature review framework.

Author(s) & Year	Focus/Context	Key Findings	Gap Identified	Future Research Calls
[6]	e-WOM in e-commerce	e-WOM influences consumer trust and purchase intentions	Limited focus on niche sectors (e.g., food delivery)	Explore e-WOM’s role in niche platforms like food delivery
[9]	Food delivery platforms	Convenience and tracking features enhance consumer perception	e-WOM not deeply explored as a driver of acceptance	Examine how e-WOM influences acceptance in food delivery
[10]	Platform engagement and interactive design	Interactive designs boost consumer satisfaction	Overlooks e-WOM’s contribution to engagement	Investigate e-WOM’s effect on interactive platform designs in food delivery
[11]	Psychological distance theory	Psychological distance reduces engagement	Lacks application to food delivery platforms	Study how psychological distance affects food delivery platform use
[29]	Consumer perception of online reviews	Review quality drives consumer confidence and reduces uncertainty	Lacks integration with consumer acceptance variables	Investigate the link between review perception and acceptance in online food delivery
[42]	Usability of food delivery apps	Positive user experience fosters loyalty	Lacks e-WOM’s mediation role	Examine e-WOM as a mediator between usability and loyalty
[87]	Consumer empowerment in digital spaces	Empowered consumers drive brand loyalty	Limited exploration in food delivery contexts	Investigate empowerment and e-WOM in food delivery apps
[88]	Social media and e-WOM	Social media boosts e-WOM, increasing consumer loyalty	No niche application to food delivery	Investigate e-WOM’s effect on food delivery platforms
[89]	e-WOM and consumer purchase intention	e-WOM directly influences purchase intentions	Does not consider acceptance–perception interplay	Explore the interplay of perception, acceptance, and e-WOM in food delivery
[90]	Online platform acceptance	Perceived ease of use and trust drive acceptance	Overlooks the influence of e-WOM	Study how e-WOM affects acceptance in digital contexts
[91]	Psychological distance in e-commerce	Reducing psychological distance improves consumer engagement	No connection to e-WOM	Analyse how psychological distance shapes e-WOM in food delivery

).

3. Materials and Methods

The methodology used in this study is specifically designed to explore the interactions between the three main variables: consumer perception, consumer acceptance, and e-WOM. Despite the reviewer’s concerns, structural equation modeling (SEM), particularly with SmartPLS, is an appropriate method for examining the complex relationships among these variables. SEM enables the analysis of direct, indirect, and mediating effects between the constructs and their sub-dimensions, allowing for a deeper understanding of their interdependencies [92]. Furthermore, the inclusion of quadratic analysis is crucial for investigating non-linear relationships, such as the observed U-shaped connection between electronic commerce and consumer perception, which carries significant theoretical implications. The multi-group analysis offers valuable insights into how various consumer segments, such as different age or income groups, engage with and perceive food delivery platforms, providing important guidance for segmentation strategies in the food delivery sector [93].

3.1. Participants and Procedure

The current study involved 835 respondents, aged 18 to 35+ years, who freely agreed to complete a survey in full. Eligibility to participate in the research required respondents to have engaged in at least one online food delivery transaction. Of the respondents, 66.44% identified as female, with an average income of less than 2000 RON per month. Most participants were undergraduates and/or students, and the typical frequency of food delivery app usage ranged from one to five times per month.

3.2. Methods, Setting, and Sample

Previous research practices informed the design of this study, which employed convenience sampling based on voluntary response [94,95,96,97]. Accessibility was considered of utmost importance, and the methodological features and scope of the research were carefully addressed. Google Forms was selected as the data collection platform, and the questionnaire was distributed through various electronic communication channels and groups.

The research complied with the General Data Protection Regulation (GDPR). Respondents accessing the online form were assured that no personal or retention data would be requested, and the completed forms were treated confidentially and used exclusively for academic purposes.

Relying on the practicality and pragmatism of previous literature, the current research employed a convenience sampling method based on voluntary response [98]. Data gathering through the Google Forms platform generated a hyperlink, which was subsequently distributed across various online formal and informal communication channels, providing researchers with ease of access and greater reach for potential respondents.

By adopting the convenience sampling methodology, the form layout was designed to be easily identifiable, comprehensible, and interactive for participants engaging with the research items and variables. To mitigate sampling bias, the research methodology emphasizes that fully completed online forms were expected only from respondents who had actively engaged in at least one online food delivery transaction, a criterion clearly outlined and explained in the introductory section of the form.

Between 4 April and 18 April 2024, data were collected for a pilot study, which was subsequently validated and used to generate hypotheses.

The research instrument was designed to provide a comprehensive understanding of the relationship between consumer perception, acceptance, and word of mouth within the online shopping environment of business and retail markets. Respondents were informed that no form of compensation or benefit was offered for participating in the study by completing the questionnaire, provided they met the previously stated conditions. Since the questionnaire was distributed online, the results reflect diverse demographic categories, including age, gender, income, and occupational status. The registered response rate was relatively low, with a seven-month period yielding 835 fully completed forms.

Due to the increasing popularity of food delivery apps, this market is experiencing rapid growth in Romania. The online food delivery market consists of two main categories—meal and grocery delivery. Orders are typically placed directly through retailers’ websites or apps. Delivery is generally conducted under the just-in-time (JIT) concept, with scheduled retail deliveries averaging 180 min and meal/restaurant deliveries averaging 85 min [99].

The data indicate that the vast majority of respondents have used food delivery apps/websites for over two years and engage with food delivery services between one and five times per month.

3.3. Measures

Through the distribution of Google Forms, the questionnaire was scaled down to maximize respondent engagement and minimize bias resulting from incomplete forms and missing answers.

The research design consisted of four parts. Initially, basic demographic data were collected, including age, gender, income thresholds, educational background, frequency of food delivery app/website usage, expenditure on food delivery, and general experience or habits related to food delivery.

The three theoretical components examined are as follows: consumer perception—measured across the following eight dimensions: convenience (CPC), interactivity (CPI), customization (CPC), character (CPCH), perceived security (CPPS), switching costs (CPSC), consumer satisfaction (CPCS), and consumer loyalty (CPCL); electronic commerce–consumer acceptance—measured across the following seven dimensions: transaction behavior (CAAT), intention to transact (CAIT), trust (CAT), perceived risk (CAPR), perceived usefulness (CAPU), perceived ease of use (CAPEU), and web retailer reputation (CARR); and electronic word of mouth—measured across the following five dimensions: information quality (EWMIQ), e-commerce trust (EWMT), social–psychological distance (EWMSD), purchase intentions (EWMPI), and sense of power (EWMSP).

To better understand how each sub-construct functions within the broader framework, the authors explain that the three primary dimensions—consumer acceptance, electronic word of mouth (e-WOM), and consumer perception—are closely interconnected.

More specifically, consumer perception shapes consumer acceptance by influencing the user’s experience and expectations when interacting with the food delivery platform. This relationship highlights the importance of factors such as perceived ease of use, perceived usefulness, and trust in fostering consumer acceptance [84,100].

Consumer acceptance affects e-WOM by prompting consumers to share positive word of mouth when they feel satisfied and trust the platform. This sharing of experiences helps to build consumer confidence and attract new users [101].

e-WOM creates a feedback loop with consumer perception, as the quality of the information shared through e-WOM can influence other users’ perceptions of the platform, thus shaping their decision-making and purchase intentions [6,102].

A 7-point Likert scale was adopted for the entire research instrument, replacing the traditional 5-point scale. The rationale behind this choice lies in the increased accuracy provided by a broader range of options, particularly for instruments distributed exclusively within an online environment. To ensure suitability, higher response accuracy, and enhanced statistical interpretation, the 7-point Likert scale was selected, surpassing the traditional reliance on smaller scales. According to the literature, the broader scope of such a scale, along with the benefit of offering respondents a greater variety of response options, enhances individual expression and yields results closer to reality by engaging higher reasoning faculties.

Comprising a 7-point Likert scale (ranging from 1—totally agree to 7—totally disagree) with eight dimensions and 27 items, the consumer perception variable was constructed using insights from Srinivasan et al. on consumer interface quality, Salisbury et al. on perceived security, Ping et al. on switching costs, and Anderson et al. on consumer satisfaction and loyalty, with slight revisions applied [5,103,104,105]. The Cronbach’s alpha value for this construct is 0.928.

The consumer acceptance variable, also assessed using a 7-point Likert scale (ranging from 1—totally agree to 7—totally disagree), encompasses seven dimensions, adapted or developed from existing measures and scales, and contains 22 items. The perceived usefulness and perceived ease of use variables were initially analyzed by [43,106,107]. Trust, perceived risk, and web retailer reputation were derived from the studies of Järvenpää [108]. Intention to transact was based on the work of Venkatesh [2], while transaction behavior [109], reflecting the likelihood of purchase, was inspired by Brown et al. [110]. The Cronbach’s alpha value is 0.953.

The electronic word of mouth (e-WOM) dimension, also evaluated through a 7-point Likert scale (ranging from 1—totally agree to 7—totally disagree), includes five dimensions and 22 items, adapted for the purposes of this study. The information quality variable was first utilized by Park et al. [111]. Word-of-mouth trust was described by McAllister [112] and Halepete et al. [113], while social–psychological distance originates from the work of Hernandez-Ortega et.al. [114,115,116]. The purchase intentions dimension was adapted from Pavlou et al. [117], and the current format of the sense of power dimension was based on Anderson et al. [107]. The Cronbach’s alpha value for this construct is 0.927.

3.4. Analysis Strategy

As part of the strategic development of this research, the general assessment of the three primary constructs and 20 subconstructs was conducted using SmartPLS (v. 4.1.0.9) software. The internal consistency and reliability of the proposed research instrument were evaluated, yielding valid results. The analysis included structural equation modelling (SEM), followed by FIMIX segmentation analysis. This was further complemented by process emulator with quadratic non-linear effects, specific indirect effects, and the examination of the quadratic curvilinear relationship between electronic commerce and consumer perception. Finally, the structural hypotheses were tested.

The selection of SmartPLS was driven by the goal of providing researchers and reviewers with deeper insights into consumer perception, choices, and reasoning within the context of electronic (web/app) food delivery [107,117]. The software is widely recommended in the literature for cases involving at least one formative construct, necessitating the use of partial least squares (PLS) algorithms. This process generates two outcomes—the outer model, which provides data on observable variables related to latent variables, and the inner model, which offers structural data on the relationships between observed variables and other latent variables [118].

In terms of validity and reliability, the current model analyzed the outer model [119]. The inner model is employed to examine path coefficients in relation to other variables.

4. Results

According to recent literature [120], when building on previously accepted results, a new research instrument should compare construct collinearity severity values with generally accepted thresholds, such as those specified by SmartPLS software [121]. The current database’s VIF values fall within the acceptable threshold of less than 5.0, as suggested within previous studies [122,123], indicating no collinearity issues.

The AVE values, which measure convergent validity, must exceed thresholds of 0.5 (>0.5). Indicators failing to meet this criterion should be excluded [120,124]. However, it was proposed that AVE thresholds ranging from 0.4 to 0.7 may be retained if composite reliability (CR) remains unaffected [125]. The current results indicated that five items needed to be removed: CPCS2 = 0.26, EWMSP6 = 0.18, EWMSP7 = 0.14, CARR2 = −0.024, and EWMSP5 = 0.496.

For item convergence, the composite reliability (CR) Rho_c values must exceed 0.7 to ensure the retention of variables for further analysis. Construct reliability and validity assessments require Cronbach’s alpha and composite reliability values to meet satisfactory criteria, with thresholds exceeding 0.6. The current model meets these criteria, with Cronbach’s alpha ranging from 0.51 to 0.92 and CR values between 0.74 and 0.96. The construct was confirmed as valid and reliable according to these standards.

When conducting the reliability test, literature offers complementary perspectives to traditional approaches, agreeing on the feasibility of reporting Rho_a values as an alternative to Cronbach’s alpha [126,127,128]. Based on this criterion, Rho_a values ranging from 0.5 to 0.9 confirm the reliability of the database and model, providing sufficient convergent validity (see

Table 2. Construct reliability and validity (final values).

	Cronbach’s Alpha	Composite Reliability (rho_a)	Composite Reliability (rho_c)	Average Variance Extracted (AVE)
CONSUMER PERCEPTION	0.928	0.936	0.936	0.378
Character	0.808	0.809	0.887	0.723
Convenience	0.864	0.869	0.917	0.787
Consumer Loyalty	0.868	0.869	0.91	0.717
Consumer Satisfaction	0.825	0.828	0.896	0.742
Customization	0.755	0.756	0.86	0.672
EC Trust	0.886	0.888	0.921	0.746
ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE	0.953	0.955	0.957	0.497
ELECTRONIC WORD OF MOUTH	0.927	0.931	0.936	0.438
Information Quality	0.833	0.833	0.889	0.666
Intention to Transact	0.85	0.852	0.909	0.769
Interactivity	0.514	0.638	0.745	0.513
Perceived Ease of Use	0.869	0.873	0.911	0.719
Perceived Risk	0.789	0.789	0.877	0.703
Perceived Security	0.861	0.864	0.915	0.782
Perceived Usefulness	0.849	0.849	0.899	0.689
Purchase Intentions	0.775	0.779	0.869	0.69
Sense of Power	0.83	0.844	0.882	0.601
Social–Psychological Distance	0.822	0.822	0.894	0.738
Switching Costs	0.817	0.82	0.891	0.733
Trust	0.859	0.86	0.914	0.78
Web Retailer Reputation	0.825	0.848	0.88	0.603

).

The SEM analysis revealed that the values of the standardized coefficients ranged between 0.1 and 0.6, falling within the absolute accepted threshold of <|1|, indicating no multicollinearity issues.

In assessing a new model, due diligence requires the analysis of R² values [129,130]. For this procedure, the values 0.1, 0.3, and 0.6 indicate weak-to-substantial explanatory power [126,131]. Alternatively, authors describe R² peak values as follows: <0.2 (weak), <0.4 (moderate), <0.6 (substantial), <0.8 (strong), and 0.9–1 (very strong) [95,132]. In behavioral sciences, the effect size r is classified as weak (0.1), moderate (0.3), and strong (>0.5). Despite these variations, the literature consistently suggests that an R² explained variance of ≥1 for a specific construct is considered adequate (see

Table 3. Construct R² values.

CONSUMER PERCEPTION	0.75
ELECTRONIC WORD OF MOUTH	0.569

).

According to recent studies [133,134,135,136], intervals peaking at multiples of 0.25 represent weak-to-substantial R² values for explained variance. In the current construct, R² values range from 0.56 to 0.75, indicating satisfactory results.

The construct reliability and validity ensure the model’s measures are consistent and accurate, while the R² values show how well the model explains the variation in the outcomes, confirming its predictive strength.

For both the estimated and saturated models, the current SRMR values fall within the accepted thresholds of <0.1, as previously proposed [137], and align with the more conservative threshold of <0.08 [133] (see

Table 4. Model fit summary.

	SRMR	d_ULS
	Original sample (O)	Original sample (O)
Saturated model	0.081	59.778
Estimated model	0.089	72.164

).

By performing a confirmatory tetrad analysis (CTA-PLS), the current model indicators (variables with at least four indicators) were tested and confirmed to be best classified as reflective or formative. A threshold of 80% of p-values >0.05 indicates a best reflective dimension, while thresholds of p-values <0.05 below 80% are considered to be best classified as formative. The results show that sense of power variables are best measured formatively.

Furthermore, to assess the possibility of hidden heterogeneity and latent segments, a FIMIX analysis was conducted [138,139]. Such latent class analysis is recommended when a priori prevention methods are not feasible for external sources [126,140,141]. In the current analysis, the assumption consists of 80% power and a 15% effect size. The number of extracted groups for the current database should approximate 83 [141].

After performing the analysis for 10 segments, the results indicate a declining trend, which researchers assume will continue. Initial results show that 0.21% of the data are retained by the first two segments, with segment sizes continuing to decrease (see

Table 5. FIMIX segments sizes.

Segment 1	Segment 2	Segment 3	Segment 4	Segment 5	Segment 6	Segment 7	Segment 8
0.215	0.213	0.164	0.146	0.13	0.093	0.022	0.018

).

To determine the number of segments to be retained and their degree of separation, Akaike’s information criterion (AIC) modified by factor 3, along with normed entropy (EN) and consistent AIC (CAIC) criteria, are considered [142,143]. EN thresholds fall within the 0–1 interval and reflect segment reliability; the higher the value, the greater the partition quality [144].

According to an alternative view [145], if the EN value exceeds 0.5, any additional predetermined segments may be disregarded (see

Table 6. FIMIX results.

	Segment 1	Segment 2	Segment 3	Segment 4	Segment 5	Segment 6	Segment 7	Segment 8
AIC3 (modified AIC with Factor 3)	35,307.15	33,399.576	32,278.065	31,898.64	31,473.91	31,304.497	31,239.47	31,079.779
CAIC (consistent AIC)	35,516.66	33,823.356	32,916.116	32,750.96	32,540.5	32,585.36	32,734.61	32,789.184
EN (normed entropy statistic)	0	0.824	0.862	0.848	0.796	0.818	0.802	0.808
NFI (non-fuzzy index)	0	0.849	0.862	0.831	0.76	0.77	0.754	0.754
SUMMED FIT	70,823.81	67,222.932	65,194.181	64,649.6	64,014.41	63,889.857	63,974.08	63,868.963

).

As the initial results suggest, the data are not confined to a single segment. The lower the AIC3 value, the better the model fit; for this reason, Segment 8 can be considered the best option. Additionally, model complexity adjustments show that AIC3 reaches its lowest value at Segment 8, supporting the eight-segment solution.

However, EN plays a crucial role in highlighting segment membership clarity. According to the displayed values, the third segment peaks at 0.862, indicating the best solution for segment clarity. It is generally accepted that EN thresholds of 0.8 or above signify well-defined segments and are sufficient as clarity indicators [146,147].

In contexts where NFI values also indicate segment membership distinctions, the 0.862 NFI value for Segment 3 highlights it as an optimal choice for segment differentiation. The final aggregate indicator—discrete segment assignment—favors lower values, with thresholds not exceeding 0.2 [125,148]. As the overall fit measure (summed fit) shows, the best fit is observed in Segment 8, as indicated by both AIC3 and CAIC, reinforcing Segment 8 as the optimal solution for the general model fit.

To obtain deeper insights and refined understanding, importance–performance map analysis (IPMA) was deemed necessary. This analysis offers a nuanced understanding of construct performance across the eight predetermined segments. IPMA visual maps evaluate the importance (total effects) and performance of each construct for both specific latent constructs and the overall model. As the results demonstrate (see Figure 2), the FIMIX segmentation results are overlaid, allowing each construct to undergo dual evaluation, in the context of its own importance/performance and its impact across other segments.

The first IPM concerns electronic word of mouth (red marker) and sense of power as key constructs. The x-axis plots total effects (importance), while performance is displayed on the y-axis (see Figure 1). The current construct achieves high levels of importance, with a total effect of 0.95, positioned towards the right side of the scale.

Similar results are observed regarding construct performance, with values rounding to approximately 75–80. This position indicates that electronic word of mouth is highly impactful and performs well. As it already delivers significant results, it should be maintained and further leveraged through specific practices and policies on web platforms.

The sense of power indicator (blue marker) is positioned on the left side in terms of importance, with low scores of 0.92. Regarding performance levels, the scores range between 65 and 80, which are relatively high. Overall, sense of power appears to perform well but does not significantly impact the overall model. While it does not require immediate attention, close monitoring could enhance its alignment with other high-importance constructs within the model.

Regarding the electronic commerce–consumer acceptance (red marker) construct, it is noteworthy that it displays the highest importance on the map (close to 0.93), indicating a strong influence on the outcome. In terms of performance, it also scores highly (0.75), supporting the model’s objectives. As this dimension demonstrates both high impact and performance, continued attention is recommended to sustain its positive contribution.

The electronic word of mouth (blue marker) scores slightly lower in importance compared to consumer acceptance (0.91) but demonstrates high performance levels, close to 0.80. This construct also exhibits significant impact and performs well, warranting continued attention (see Figure 3).

Sense of power (purple marker) registers a moderate level of importance (0.87) and relatively high performance (75). Although it is less critical than the previous two constructs, sense of power contributes positively to the model. Monitoring this construct could be beneficial, as it may play a more significant role in response to foreseeable shifts.

The FIMIX analysis identified eight distinct segments, highlighting latent heterogeneity. Each segment corresponds to a unique subgroup, reflecting varying responses associated with the form items and underscoring the importance of tailored strategies for each emerging group. The FIMIX analysis, which facilitated the segmentation of the sample, offers deeper insights into how constructs such as electronic word of mouth and sense of power influence different groups to varying degrees.

By conducting the IPMA analysis, the overlayed segmentation reveals segment-specific variations, reinforcing the universal importance of some constructs (e.g., electronic word of mouth). Simultaneously, the results suggest that constructs such as sense of power may hold greater relevance for certain segments but not for the overall dataset. This segmentation enables web retailers to implement customized strategies [149,150]. A layered approach can maximize model outcomes by directing resources and strategies towards the most influential constructs within each segment.

PLS predict analysis reveals that Q² values compare the errors encountered in the path model with the simple mean [151]. The PLS-SEM evaluation incorporates the model’s predictive power. In the current model, Q² > 0, indicating a correctly specified model with strong inner model predictive power [152].

Additionally, a process emulator with quadratic non-linear effects, controls, and moderated mediation was conducted. The path model diagram highlights constructs, such as consumer acceptance (CA), consumer perception (CP), trust (TR), perceived risk (PR), and ease of use (EU), which were tested for significance. Overall, the path weights (e.g., CA → CP or CP → CA → TR) reveal the strongest relationships, with the most substantial influence illustrated by the thickness of the arrows (see Figure 4).

R² values exceeding 0.67 indicate a substantial model fit, values between 0.33 and 0.67 indicate a moderate fit, and values between 0.19 and 0.33 indicate a weak fit [137]. In the given model, the R² value for consumer acceptance is 0.673, suggesting a substantial model fit. Similarly, the R² value for web retailer reputation (WRR) is 0.765, indicating strong explanatory power and high variance.

The quadratic analysis was employed to capture non-linear effects, particularly the U-shaped relationship between electronic commerce and consumer perception, indicating that increased engagement has a stronger influence on perception [153]. The use of control variables ensured that external factors did not skew the primary relationships, allowing for a clearer understanding of the core effects. Moreover, the moderated mediation analysis explored the role of ECT (electronic commerce trust) in shaping the relationship between consumer acceptance and e-WOM, providing valuable insights into how trust dynamics differ across consumer segments [154] This combination of approaches strengthens the model’s capacity to reflect the complexities of consumer behavior in the food delivery context.

The direct effects (path coefficients) are considered meaningful in the literature [131] if they meet the 0.1 threshold. However, it was suggested that path coefficients should be interpreted alongside t-values and p-values to ensure precise statistical significance [95]. T-values are expected to meet a threshold of at least 1.96 for a 95% confidence level (p < 0.05), as authors recommend [155].

In the current analysis, path coefficients, such as CA → EU (t = 23.041, p < 0.001) and CA → PU (t = 26.849, p < 0.001), demonstrate high significance, indicating robust direct effects.

The total effects, interpreted as the sum of direct and indirect effects, provide insights into the overall impact of one variable on another within the model. This analysis aids in understanding comprehensive influence, particularly when mediated paths are present [156]. For the current model, examples of substantial effects include CA → EU with a total effect of 1.273, highlighting the significant cumulative influence of CA on EU at both direct and indirect levels.

The specific indirect effects (mediation) are examined by identifying the indirect paths between constructs. If both direct and indirect effects are significant, complementary mediation is present [124]. If only the indirect effect is significant, indirect-only mediation is achieved. In the current model, examples of significant indirect effects confirm the presence of mediation, such as QE (CP) → CA → TR (p = 0.037), indicating that CA mediates the effect of CP on TR.

Moderated mediation (conditional direct and indirect effects) analysis highlights variations in construct relationships depending on the level of a moderator (in this case, ECT). For moderated mediation to occur, the conditional direct effects must vary significantly across different moderator levels [157]. In the current model, the effect of CA → EU is moderated by ECT. For example, at +1SD of ECT, the CA → EU path shows a significant effect (t = 33.005, p < 0.001), indicating that, at high ECT levels, CA strongly influences EU. This suggests that ECT intensifies the relationship between CA and EU.

Multi-group analysis (MGA), recommended by Henseler et al. [139], involves comparing path coefficients across different groups [158], with significant differences observed at p < 0.005. This analysis provides specific insights for each segment, offering a deeper understanding of how relationships vary between groups.

In the current study, multi-group testing based on age group comparison reveals that the relationship CP → CA is stronger for the 18–30 years age group (β = 0.65, t = 4.32, p < 0.005) compared to the 31–50 years age group (β = 0.40, t = 2.01, p < 0.05), suggesting a greater influence of CP on CA among younger respondents. For the CA → EU relationship, both age groups exhibit a significant positive relationship (β = 0.55/0.48, t = 3.82/2.75, p < 0.05).

These results indicate that a gender-neutral approach to perception-based marketing in online stores may suffice across different demographic groups.

According to the gender group analysis, the CP → CA path does not show a statistically significant difference, suggesting that both gender groups are equally influenced by perception factors when analyzed in terms of acceptance. However, for the CA → TR path, a significant effect was observed for females rather than males (β = 0.68/0.42, t = 4.90/2.35, p < 0.05). This result suggests that female online shoppers build trust by relying more on acceptance factors. Therefore, websites should focus on acceptance-related trust-building efforts, such as enhancing reliability and ease of use, to better engage female consumers.

In the income group comparison, paths such as CP → CA exhibit a stronger path coefficient for the lower-income group compared to the middle-income group (β = 0.60/0.35, t = 4.25/2.10, p < 0.05). This finding indicates that lower-income consumers place greater emphasis on perception factors (such as value for money and ease of access) when forming acceptance judgements. Conversely, middle-income consumers appear less influenced by perception when evaluating products for online purchases.

Regarding the CA → EU path, middle-income consumers display a stronger relationship compared to lower-income consumers (β = 0.58/0.45, t = 3.50/3.02, p < 0.05), suggesting that the former group is more likely to translate acceptance into actual usage. This result implies that acceptance strategies targeting middle-income segments have a higher impact on engagement levels.

4.1. Control Variables

Control variables should be included when they explain substantial variance in the dependent variable [159]. In the current model, control variables affecting IQ show varying levels of significance. The IQ → CO path demonstrates a significant effect (p = 0.014), indicating an independent contribution to IQ variance beyond the primary predictors. This suggests that the variable should be retained in the model to provide sufficient explanatory power.

Conversely, the IQ ← CUSA control variable is not significant (p = 0.651) and does not have a meaningful effect on IQ. As recommend, non-significant variables can be considered for removal in future model developments to enhance parsimony [138].

4.2. Process Emulator and Model Fit

The process emulator with quadratic non-linear effects, controls, and moderated mediation analysis concludes with an assessment of model fit (GOF), [160,161,162]. For the current model, GOF values greater than 0.36 are considered large, while values between 0.25 and 0.36 are deemed medium, and those below 0.1 are classified as small.

For the selected endogenous constructs, the following GOF values were obtained: consumer satisfaction: 0.747, purchase intentions: 0.662, trust: 0.762, web retailer reputation: 0.708.

The overall GOF value of 0.721 places the model within the large range (>0.36), suggesting a strong model fit that adequately captures variance within the endogenous constructs and demonstrates substantial explanatory power.

4.3. Specific Indirect Effect Sizes

By measuring specific indirect effect sizes and applying v² thresholds (0.01 for small, 0.04 for medium, and 0.09 for large), as suggested by Aiken et al. [163], thresholds with p > 0.05 are considered to display excessive error, warranting rejection. Conversely, values with p < 0.05 are deemed significant and support the existence of the observed effect.

For v² < 0.01, researchers suggest that results should only be reported in cases involving small sample sizes or underdeveloped research fields [162]. In sociological or psychological fields, lower thresholds are generally acceptable.

The results indicate a considerable number of large effects, including electronic word of mouth → electronic commerce–consumer acceptance → consumer perception, electronic word of mouth → electronic commerce–consumer acceptance → consumer perception → switching costs, electronic word of mouth → electronic commerce–consumer acceptance → intention to transact (see

Table 7. Specific indirect effects.

	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	p Values	The Effect Size Column
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION	0.407	0.408	0.064	6.402	0	LARGE EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → INTENTION TO TRANSACT	0.437	0.439	0.068	6.47	0	LARGE EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → PERCEIVED EASE OF USE	0.512	0.513	0.067	7.679	0	LARGE EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → PERCEIVED RISK	0.477	0.478	0.067	7.127	0	LARGE EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → PERCEIVED USEFULNESS	0.486	0.485	0.066	7.401	0	LARGE EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CHARACTER	0.362	0.36	0.058	6.217	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → TRANSACTIONS BEHAVIOR	0.07	0.082	0.03	2.317	0.021	NO EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMER LOYALTY	0.28	0.284	0.058	4.803	0	MEDIUM EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → INTERACTIVITY	0.128	0.148	0.044	2.896	0.004	SMALL EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → WEB RETAILER REPUTATION	0.174	0.203	0.059	2.956	0.003	MEDIUM EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CONVENIENCE	0.34	0.338	0.057	5.975	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → TRUST	0.143	0.167	0.051	2.806	0.005	SMALL EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMIZATION	0.244	0.244	0.055	4.402	0	MEDIUM EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → TRANSACTIONS BEHAVIOR	0.21	0.21	0.051	4.099	0	MEDIUM EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMER SATISFACTION	0.369	0.369	0.06	6.124	0	LARGE EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → TRUST	0.428	0.429	0.065	6.586	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → PERCEIVED USEFULNESS	0.162	0.189	0.055	2.931	0.003	SMALL EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → WEB RETAILER REPUTATION	0.521	0.522	0.069	7.566	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CHARACTER	0.121	0.14	0.043	2.832	0.005	SMALL EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CONVENIENCE	0.114	0.132	0.042	2.719	0.007	SMALL EFFECT
ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → INTERACTIVITY	0.382	0.382	0.059	6.446	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMER LOYALTY	0.093	0.11	0.037	2.535	0.011	SMALL EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMER SATISFACTION	0.123	0.144	0.045	2.759	0.006	SMALL EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMIZATION	0.081	0.095	0.033	2.432	0.015	NO EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION	0.136	0.159	0.048	2.812	0.005	SMALL EFFECT
ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CHARACTER	0.672	0.666	0.056	11.964	0	LARGE EFFECT
ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CONVENIENCE	0.631	0.626	0.064	9.896	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → PERCEIVED EASE OF USE	0.171	0.199	0.058	2.962	0.003	SMALL EFFECT
ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMER LOYALTY	0.519	0.524	0.069	7.566	0	LARGE EFFECT
ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMER SATISFACTION	0.685	0.683	0.057	12.077	0	LARGE EFFECT
ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → CUSTOMIZATION	0.452	0.45	0.074	6.098	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → PERCEIVED RISK	0.159	0.186	0.056	2.852	0.004	NO EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → EC TRUST	0.317	0.365	0.084	3.794	0	NO EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → INTENTION TO TRANSACT	0.146	0.171	0.052	2.798	0.005	NO EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE	0.18	0.21	0.061	2.975	0.003	NO EFFECT
ELECTRONIC COMMERCE–CONSUMER ACCEPTANCE → CONSUMER PERCEPTION → INTERACTIVITY	0.708	0.711	0.078	9.072	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → INFORMATION QUALITY	0.319	0.368	0.085	3.777	0	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → PURCHASE INTENTIONS	0.265	0.305	0.077	3.463	0.001	LARGE EFFECT
SENSE OF POWER → ELECTRONIC WORD OF MOUTH → SOCIAL–PSYCHOLOGICAL DISTANCE	0.217	0.249	0.068	3.18	0.001	MEDIUM EFFECT

).

Since the outer model is valid, the next step involves testing the structural hypotheses, considering the values of R² and f². The model assumes no direct effect between consumer perception and electronic word of mouth.

Hypothesis 1 (H1) posits that electronic commerce has a positive effect on consumer perception. The results reveal a path coefficient of 0.866, which is highly significant with a t-statistic of 58.277 and a p-value of 0. The R² value for consumer perception indicates that electronic commerce explains 75% of the variance in consumer perception, reflecting a substantial effect [163]. Based on the effect size (f²), the effect is considered moderate-to-high.

These results suggest that, as the perception of electronic commerce improves, there is a significant increase in consumer perception.

Hypothesis 2 (H2) assumes that electronic commerce positively affects electronic word of mouth. The analysis shows a path coefficient of 0.754 (t-statistic of 27.727 and a p-value of 0), with an R² value of 0.569 for electronic word of mouth. This indicates that electronic commerce explains 56.9% of the variance in electronic word of mouth, representing a substantial level of explained variance. The effect size (f²) is 1.32, considered small-to-moderate [93,164].

Based on the model’s path coefficients and the relationships among constructs [153], a list of additional hypotheses has been classified according to their importance. All secondary hypotheses have been accepted, following the rationale previously presented (see

Table 8. Hypotheses test results.

Hypothesis	Path Coefficient	t-Statistic	p-Value	Decision	Importance
H1: Electronic Commerce has a Positive Effect on Consumer Perception	0.866	58.277	0	Accepted	High
H2: Electronic Commerce has a Positive Effect on Electronic Word of Mouth	0.754	27.727	0	Accepted	Medium
H3: Electronic Commerce has a Positive Effect on Intention to Transact	0.894	66.004	0	Accepted	High
H4: Electronic Commerce has a Positive Effect on Perceived Usefulness	0.992	108.054	0	Accepted	High
H5: Electronic Commerce has a Positive Effect on Perceived Ease of Use	0.886	63.659	0	Accepted	Medium
H6: Electronic Commerce has a Positive Effect on Trust	0.931	76.108	0	Accepted	High
H7: Electronic Commerce has a Positive Effect on Web Retailer Reputation	0.992	118.621	0	Accepted	High
H8: Electronic Commerce has a Positive Effect on Transactions Behavior	0.622	26.266	0	Accepted	Low
H9: Electronic Commerce has a Positive Effect on Perceived Risk	0.945	62.755	0	Accepted	Medium
H10: Electronic Word of Mouth has a Positive Effect on EC Trust	0.962	94.735	0	Accepted	High
H11: Electronic Word of Mouth has a Positive Effect on Information Quality	0.955	76.035	0	Accepted	Medium
H12: Electronic Word of Mouth has a Positive Effect on Purchase Intentions	0.906	49.726	0	Accepted	Medium
H13: Electronic Word of Mouth has a Positive Effect on Sense of Power	0.819	31.298	0	Accepted	Low
H14: Electronic Word of Mouth has a Positive Effect on Social–Psychological Distance	0.892	49.053	0	Accepted	Low

).

By calculating the quadratic curvilinear relationship between electronic commerce (X) and consumer perception (Y), the results reveal a U-shaped relationship with a positive quadratic term of 0.021X², indicating an upward-opening curve. The linear term of 0.825X suggests a linear relationship between the two variables.

The results imply that low or even moderate levels of buyer engagement with electronic commerce do not significantly influence consumer perception. However, in cases where higher engagement levels are recorded (potentially due to familiarity, improved website features, ease of access, or reliability), consumer perception shifts positively (see Figure 5).

The U-shaped relationship could inform strategies for electronic food markets, focusing on enhancing electronic commerce engagement and fostering positive consumer perception.

The following graph illustrates the U-shaped (curvilinear) relationship between electronic commerce (X) and electronic word of mouth (Y). Similar to the previous case, the relationship forms an upward-opening parabola, suggesting that the relationship between the two variables initially decreases to a certain point before beginning to rise.

The U-shaped curve implies that moderate engagement in electronic commerce may initially have near-negative effects on electronic word of mouth. However, as activities related to electronic commerce increase, there is a shift towards a positive effect on electronic word of mouth (see Figure 6).

This result implies that, to achieve a high level of electronic commerce, a positive impact on word-of-mouth promotion must be fostered. Until that level is reached, sellers cannot effectively motivate consumers to engage in electronic word-of-mouth promotion. The importance of these findings lies in the strategies that food delivery websites/apps must develop, focusing on consumer engagement and promotion.

The findings of this study confirm that consumer perception, e-WOM, and consumer acceptance are all significantly interconnected within the food delivery context. Although all hypotheses were supported, this does not imply that each variable has an identical level of impact. Each factor contributes uniquely to the consumer decision-making process, with certain variables exerting a stronger influence than others. For instance, perceived usefulness and ease of use have a more substantial effect on consumer acceptance compared to other factors, such as trust [165]. These results emphasize the importance of prioritizing ease of use and perceived value to enhance consumer engagement.

The acceptance of all hypotheses suggests the model’s robustness and its capacity to capture key dynamics within the food delivery industry. However, future research should explore additional factors or refine the existing model to investigate potential moderating or mediating variables that could account for variations in the relationships between constructs. Moreover, the results highlight the significant role of e-WOM, not only as a predictor of consumer acceptance but also as a feedback mechanism that shapes consumer perception. This insight enriches the understanding of consumer loyalty and engagement within online environments.

The analysis methods employed in this study provide valuable contributions to both theory and practice. Construct reliability and validity assessments ensure the robustness of the model and offer reliable tools for future research [95]. The R² values demonstrate the explanatory power of the model, highlighting the key drivers of consumer behavior, which can help food delivery platforms focus on critical areas, like ease of use and trust [166]. SEM analysis enables a comprehensive exploration of direct and indirect effects, offering insights into consumer decision making and improving marketing strategies [165]. The FIMIX latent class analysis and IPMA further refine segmentation strategies, allowing platforms to tailor their services to different consumer needs [138]. Quadratic non-linear and moderated mediation analyses reveal the complexities of consumer engagement, showing the importance of trust and the U-shaped relationship between electronic commerce and consumer perception [144]. These findings offer both theoretical insights into consumer behavior and practical recommendations for enhancing user experience and engagement in the food delivery market.

4.4. Hypothesis Testing

The first hypothesis posited that electronic commerce (consumer acceptance) has a positive effect on consumer perception across all components; the path coefficient was found to be 0.866 (t-statistic = 58.277, p-value < 0.001). This indicates a strong and significant effect, confirming the validity of the hypothesis. After testing the model and analyzing the path coefficients, the hypothesis was validated. According to the factor loadings, consumer acceptance through electronic commerce in relation to consumer perception recorded the highest factor loadings for perceived usefulness (0.788), perceived ease of use (0.858), and web retailer reputation (0.854). The lowest value was observed for the transaction behavior dimension (0.376), but the overall model data support the hypothesis.

Subsequently, the relationship between electronic commerce and electronic word of mouth was tested, with a path coefficient of 0.754 (t-statistic = 27.727, p-value < 0.001), which also showed significant support for the hypothesis. The factor loadings indicate a positive effect, thereby supporting the second general hypothesis (H2).

Additionally, the interconnections among the three variables and their components were examined. For electronic commerce, the strongest scores were recorded for CARR, CAPEU, and CAPU. For electronic word of mouth, the strongest relationships were found for EWMT, EWMIQ, EWMPI, and EWMSD. Consumer perception was primarily defined by CPC, CPI, CPCS, and CPC, confirming all secondary hypotheses (H1–H14).

Following the methodology of Karakhan [167], the general results and discussions indicated that all seven and eight respective dimensions for consumer acceptance and consumer perception were confirmed after running tests for CR and variance. This demonstrates that the research tool is highly significant for the electronic food commerce and delivery industry.

Additionally, a quadratic curvilinear relationship was observed between electronic commerce and consumer perception. The U-shaped relationship was captured using a quadratic term (0.021X²) alongside the linear term (0.825X), showing that higher engagement with electronic commerce positively influences consumer perception, particularly when the engagement reaches higher levels [147].

These statistical results are consistent with the findings of previous studies in the field [161,167], confirming the robustness of the proposed relationships in the food delivery context. The next sections further elaborate on the mediation and moderation analysis, with a focus on how external factors, such as electronic commerce trust (ECT), influence the key relationships.

These findings provide a fundamental baseline for further investments in digital development and platform/app improvements. The study underscores the importance of consumer perceptions and word-of-mouth promotion within the online environment—a dimension often overlooked by many food delivery market players, who assume it to be relevant only to the physical world.

The proposed study is, therefore, of significant interest to the food industry, as home/work delivery has become a habitual practice for large population segments, differentiated largely by age and income levels. By evaluating the environment across multiple food delivery settings, the research has the potential to attract interest from individual and conglomerate food delivery platforms, restaurants, and food processing companies at both institutional and industry-wide levels.

5. Discussion

The current research presents a twofold aim. It examines food delivery chains operating through websites and app interfaces, focusing on online consumer perception, consumer acceptance, and electronic word of mouth (e-WOM). In exploring and explaining the intricacies of these relationships, consumer perception—analyzed across eight dimensions—was employed to provide a novel perspective on the roles of CPC, CPI, CPCS, and CPPS.

Additionally, electronic commerce–consumer acceptance was deemed essential to capture behaviors associated with CAAT, CAIT, CAT, CAPR, and CAPEU, while also highlighting the significance of web retailer reputation (CARR). Ultimately, e-WOM was incorporated to assess its influence, particularly in a context where individuals in the early stages of their life span exhibit evolving behaviors regarding the acquisition of raw and processed food, as well as other goods.

According to the PLS predict values, trust (as part of e-WOM) serves as a moderator for both consumer acceptance and consumer perception variables. This finding underscores the necessity of conducting this study—not only to contribute to the literature by reflecting the latest online consumer behavior trends but also to assist food delivery chains in implementing strategies that emphasize factors such as psychological distance, information quality, consumer e-WOM trust, perceived ease of use, and retailer reputation (both online and offline).

To provide a deeper understanding of these findings, the following sections will discuss the results in detail.

5.1. Consumer Acceptance and Consumer Perception

Driven by convenience, personalization, and increasingly efficient offerings, food delivery companies operating online have enhanced consumer perception, transforming traditional consumer interactions with delivery services. Authors highlight the role of tracking systems in improving the food delivery experience [164], with an additional positive impact derived from tailored web and app-based ordering systems, further enhancing consumer satisfaction [168].

Secure payment gateways play a crucial role in boosting consumer experience, standing as a cornerstone for consumer perception and satisfaction [80].

Consumer behavior in food delivery channels has been shaped by consumer acceptance, resulting in a significant impact on e-WOM. The personalized interfaces of online platforms and apps, combined with reliable delivery services, contribute to positive consumer satisfaction. This satisfaction is reflected on electronic platforms, enhancing credibility and reinforcing the importance of customary e-WOM. Efficiency and convenience create favorable conditions for e-WOM [169]. Ray et al. [78] synthesize consumer intention to transact with efficient website and delivery app navigation, as well as order efficiency.

Consumer trust is primarily enhanced by reviews, secure payment features, and firewalls [170]. The research indicates that perceived ease of use positively correlates with user adoption rates. Studies highlight the relationship between long-term consumer loyalty and positive past experiences [78,171]. e-WOM is strongly linked to users’ social experiences, influencing consumer purchase decisions either positively or negatively. A positive experience increases the likelihood of future interactions, fostering long-term engagement.

When authors explored the usability of food delivery apps, it was found that consumer experience directly correlates with consumer satisfaction [170], increasing the probability of repeat interactions and driving consumer loyalty [172,173]. These findings reinforce the notion that consumer acceptance shapes positive behaviors within the online food delivery community.

5.2. Consumer Acceptance and Electronic Word of Mouth

Trust, a core element of e-WOM, is frequently generated by online buyers who provide detailed feedback regarding previous purchases through food delivery channels [168]. The accuracy and relevance of online feedback actively influence buyer behavior [174]. High-quality peer reviews instill confidence in consumers regarding their purchasing decisions, reducing uncertainty and aligning with consumer needs [175]. Furthermore, consumer acceptance emerges as a direct effect of empowered consumers, driven by e-WOM [168].

Psychological distance, defined as the extent to which a consumer relates to a brand or delivery service, plays a pivotal role. Research suggests that e-WOM acts as a personalized online infrastructure, enhancing product accessibility and relevance for consumers [176]. In this way, e-WOM bridges the gap between the lack of physical brand presence and the consumer’s ability to evaluate services, driving businesses to enhance consumer loyalty through deeper engagement.

The positive connection between consumer acceptance and e-WOM across the food delivery industry highlights three key elements—trust, user engagement, and purchase intentions [87]. In 2023, studies revealed that food delivery platforms benefit from positive e-WOM, with such behaviors fostering consumer loyalty and encouraging repeat purchases [11,88,89]. It was emphasized that consumers are directly influenced by the quantity, detail, and multimedia content of previous reviews [177], particularly regarding the quality and reliability of food delivery services.

For example, Wang et al.89] conducted research demonstrating that social media platforms amplify the effects of e-WOM on consumer engagement and satisfaction. This further underscores the critical role of e-WOM in contemporary food delivery services, particularly in driving consumer acceptance.

5.3. Theoretical Implications

In terms of theory building, the current research seeks to introduce new variables to the existing framework of consumer acceptance and behavior towards e-WOM, applying them to the context of electronic food delivery.

One of the primary shortcomings of the previous literature is the lack of comprehensive examination of consumer behavior and acceptance within the context of electronic word of mouth. The rapid evolution of social media, electronic platforms, and apps has amplified the challenge of managing the quality of online information shared by consumers, particularly for web retailers and food delivery chains. While earlier studies have highlighted the influence of previous online food delivery users through comments on ordering decisions, they have not sufficiently addressed consumer perception and the persuasive power of e-WOM in driving (or deterring) purchase decisions.

The current research builds upon established theoretical frameworks, primarily the technology acceptance model (TAM) [44] and Bhattacherjee’s expectation–confirmation theory (2001), focusing on food delivery chains [3].

The findings position consumer perception as a mediating factor in shaping online consumer satisfaction and loyalty. This extends previous research [178,179], which underscored the importance of platform engagement and interactive design. By incorporating variables such as character and switching costs, the results offer deeper insights into factors influencing online consumer loyalty and retention in food delivery services [3].

In this study, e-WOM is shown to positively contribute to the quality of electronic information, influencing variables such as sense of power and psychological distance. Although the previous literature has largely regarded e-WOM as an informal influence on consumer perception [180], the present study identifies e-WOM as playing a pivotal role in empowering consumers, fostering brand trust, and reducing perceived social barriers in electronic environments. These findings align with prior research [181] and contribute to risk-reduction theories [182], suggesting the significance of online trust-building frameworks in mitigating consumer perceived risk and enhancing trust in food delivery apps and websites.

In line with the existing research, which highlights the mediating role of e-WOM in online purchase intentions [89], this study offers new insights into the perceived usefulness of online interactions, with a focus on transactional behavior. The findings provide practical support for food delivery chains, demonstrating that consumer acceptance—driven by perceived ease of use and web retailer reputation—underpins e-commerce adoption, consistent with the previous literature [183].

By integrating the concept of psychological distance, the research introduces a new perspective on overcoming psychological barriers to decision making, thereby contributing to earlier studies [184] that examine the role of relatability in virtual environments [90]. These insights not only reshape the understanding of consumer acceptance in the digitalized food delivery sector but also offer pathways for optimizing user experiences and developing tailored sales strategies for food delivery apps and service providers.

The findings lay the groundwork for future research exploring the dynamic interplay between e-WOM, consumer acceptance, and behavior in rapidly expanding sectors, such as food delivery.

5.4. Practical Implications

The current study offers several actionable strategic insights for food delivery platforms. Firstly, it is essential to highlight the importance of providing intuitive and secure web/app interfaces, thereby enhancing the perceived ease of use and convenience for online consumers. Such factors contribute to building trust and motivating buyers to repeatedly use services. In this process, real-time order placement, secure online transactions, and order tracking can serve as simple yet necessary means for food delivery platforms to achieve improved outcomes [11].

Additionally, consumer trust and purchase intentions are directly influenced by e-WOM, derived from positive experiences and ratings [91]. Based on these findings, loyalty programs and exclusive discounts could enhance the popularity of food delivery businesses, thereby increasing returns.

Reducing psychological distance can foster the perception of high reliability among consumer groups and individuals, making services more relatable and accessible. Personalized communication and content on websites can bridge the psychological gap, resulting in higher levels of consumer engagement [10].

Moreover, food delivery platforms should prioritize addressing perceived risk at an organizational level. Effective strategies include transparent pricing and positive, tailored customer service policies [185]. Trust can also be nurtured through participation in initiatives such as verified business programs and unbiased consumer testimonials.

Finally, consumer empowerment strategies within food delivery chains should focus on real-time feedback and order decision-making processes. Engaging with empowered consumers fosters brand loyalty [172].

Adopting and implementing these strategies could yield positive effects not only on existing consumer perceptions but also on e-WOM, thereby establishing a solid foundation for repeat transactions and long-term loyalty.

5.5. Limitations and Future Research

Despite the theoretical and practical contributions to the food delivery sector, a number of limitations must be acknowledged.

Firstly, the uneven sample distribution—with the majority of respondents identifying as female—presents a potential bias. The literature suggests that perceptions of information credibility may vary across genders [186]. Hence, future research should further investigate the influence of gender on the relationship between online peer review quality and trust.

Additionally, as food delivery users predominantly consist of individuals in the first half of their life span—who are also frequent users of social media and web apps—the impact of e-WOM may differ across age groups. Experienced internet users are more influenced by online information [30], while less experienced users may encounter uncertainty and interpret online content with greater skepticism, potentially resulting in mistrust.

To enhance familiarity with food delivery app content, future research could consider consumer experience as a moderating variable, examining its influence on purchase intentions.

In the present study, electronic commerce trust was hypothesized as a moderating variable. However, the results indicate a limited effect. Trust is recognized as a subjective element of consumer perception [187] Nonetheless, with the rapid growth of social networks and peer review systems, consumer trust is becoming increasingly important and versatile.

External factors, such as the content provided on food delivery websites and apps, may gain significance for opinion leaders [188], reinforcing positive consumer interactions and fostering stronger relationships.

Consumer loyalty has become a prevailing objective for marketers, while consumer satisfaction in relation to website/app interfaces has generally been overlooked. While in the physical world, salespeople interact directly with consumers, having indirect control over their satisfaction, in online commerce, such actions are replaced by web interfaces designed to bring the possibility of purchase closer [189]. As a direct result, the quality of the consumer interface plays a major role in a business’s financial performance. Additionally, online comments, shared opinions, and multimedia—namely, electronic word of mouth (e-WOM)—ultimately translate into significant effects on online purchases [16].

For future research, it is essential to focus on key constructs specific to the food delivery sector, rather than including general elements from related studies. While the current study examined a broad range of constructs, future research should identify the unique factors that influence consumer behavior in food delivery, such as delivery speed, product quality, and service reliability. By narrowing the scope to these key constructs, the research can offer more relevant insights for managing consumer engagement and loyalty in food delivery chains.

Additionally, future studies should consider exploring moderating and mediating variables specific to the food delivery context, such as logistics and market characteristics, to enhance understanding of the industry’s dynamics. This will provide more targeted strategies and help food delivery platforms tailor their services more effectively, differentiating them from other sectors.

The current results enhance the research stream by clarifying the dynamics of consumer perception, acceptance, and e-WOM in the food delivery sector. The study extends existing models, showing how trust, perceived ease of use, and usefulness shape consumer engagement. It also demonstrates that information quality (IQ) positively influences consumer acceptance (CA) through electronic commerce trust (ECT), and that ECT moderates the impact of IQ on purchase intentions (PI). These findings contribute valuable insights for both research and the management of food delivery chains by providing a deeper understanding of the interactions between information quality, consumer acceptance, and electronic commerce trust [190]. These insights can inform strategies aimed at enhancing consumer engagement and fostering greater trust in digital platforms. Furthermore, this study lays the groundwork for future research, encouraging further exploration of these relationships and supporting the evolution of management strategies in the fast-changing food delivery industry [191].

6. Final Remarks

The current research advances theories concerning the interdependent relationships between consumer acceptance, consumer perception, and electronic word of mouth (e-WOM) within the complex online interfaces of food delivery chains. The key findings highlight consumer perception factors, such as convenience, interactivity, and perceived security, as mediators influencing consumer acceptance.

Furthermore, e-WOM was found to have a significant impact on elements such as trust, consumer purchase intentions, and perceived usefulness, thereby not only reducing perceived risks but also diminishing psychological distance. This research underscores the importance of information quality and consumer empowerment, contributing to the evolving landscape of digital food delivery platforms and their relationship with consumer engagement. In this dynamic, web retailers must navigate interactions with empowered consumers who actively shape digital platform experiences.

These findings contribute to existing theories by incorporating underexplored constructs and variables, such as sense of power and sociological distance. Given the accelerating pace of technological advancement, e-WOM has emerged as a key driver of consumer loyalty and adoption. The current study provides valuable theoretical and practical insights, supporting the optimization of future consumer experiences within the global e-commerce environment.