What motivates consumers’ continued usage intentions of food delivery applications in post-COVID-19 outbreak? Comparing Generations X, Y and Z

Abstract

Purpose

Food delivery applications (FDAs) are becoming more and more well-liked across Generations X, Y and Z, with Asia experiencing the biggest growth. These three generations of consumers have distinct views toward using FDAs because they were influenced by various social environments, cultures and experiences. Therefore, marketers ought to present customers with various values. The purpose of this study is to investigate how the theory of consumption values (TCV) affects the intention of various generational cohorts to continued usage intentions.

Design/methodology/approach

Online surveys were distributed to 745 Thai customers who had previously ordered meals from FDAs to collect information. The data were analyzed using partial least squares structural equation modeling and multigroup analysis.

Findings

Results reveal that there are significant differences among Generations X, Y and Z, while emotional value is insignificant. Generation X is concerned about functional and conditional value, whereas Generation Y expresses the highest concern on epistemic and social value. Meanwhile, Generation Z pays attention to epistemic, functional and social value.

Originality/value

The novelty of this study contributes to TCV affecting the different FDAs consumption value among Generations X, Y and Z, while the previous research only focused on gender and age difference, and this study firstly combines FDAs with TCV to predict the consumers’ intention to continuously use FDAs in the post-COVID-19 outbreak.

Citation

Rungruangjit, W. and Charoenpornpanichkul, K. (2024), "What motivates consumers’ continued usage intentions of food delivery applications in post-COVID-19 outbreak? Comparing Generations X, Y and Z", Journal of Asia Business Studies, Vol. 18 No. 1, pp. 224-251. https://doi.org/10.1108/JABS-06-2023-0234

1. Introduction

The COVID-19 pandemic has increased the company’s demand for digital transformation. Even traditional businesses are now somewhat compelled to conduct business online (Kumar and Shah, 2021). Particularly, in the restaurant business, it changed consumer habits and contributed to the rapid transformation into online services for survival from the outbreak (Silva et al., 2022). To keep afloat or take advantage of the digital platform, new companies have initiated the food delivery applications (FDAs) business (Kumar and Shah, 2021). FDA implementation has met consumer needs for convenient food services and personal cleanliness in addition to commercial requirements for businesses (Silva et al., 2022). These applications make an easy way to order food and drinks from various restaurants without having to go to the restaurant, help the consumers conveniently see listed restaurants, menus, track the status of orders, find out interesting promotion, find the nearest restaurant, ratings, make online payment and check order statuses without physical contact with restaurants (Burlea-Schiopoiu et al., 2022). Moreover, FDAs have also offered more opportunities for different restaurant business operators, including street food and local food vendors, to earn more income and attract new groups of customers (Chotigo and Kadono, 2021). FDAs have emerged as a global trend. A segment of mobile applications with high growth rates includes this one (Silva et al., 2022). According to Statista (2022a), the global market for meal delivery was around US$130.2bn in 2022 and was projected to reach US$223.7bn by 2027. Asia accounted for 44% of all finalized deals worldwide for online food and meal delivery in the third quarter of 2022. With about 31% of the deals, the USA came in second, while 17% were in Europe (Statista, 2023a, 2023b). One of the Southeast Asian industries with the highest growth rates is online meal delivery in Thailand (Sirikeratikul, 2020). When things return to normal, a huge worry is that consumers will continue to hang out and eat out. Consumer behavior of using FDAs during outbreaks will probably change either in a positive or negative way and that might affect the restaurants or food service providers significantly. Thus, the restaurants and FDA providers should be well-prepared. Nevertheless, prior studies explore the factors affecting the adoption intentions to FDAs (Ali et al., 2021), intention to use FDAs (Chotigo and Kadono, 2021), customer satisfaction with FDAs (Bunarunraksa and Nuangjamnong, 2022), during the COVID-19 crisis period. There are a few research on the intention to continue using FDAs after the COVID-19 pandemic. This study tries to bridge that gap in primary research. In addition, there have been significant conceptual gaps about FDAs in the past.

The existing studies primarily focus on the application of technology acceptance, such as the unified theory of acceptance and use of technology (UTAUT) (Okumus et al., 2018), the technology acceptance model (Prabowo and Nugroho, 2019), and extended technology acceptance model (Lee et al., 2017). Few research have used any important theories of consumer behavior to investigate how customers react to FDAs. For instance, Ray et al. (2019) used the uses and gratifications theory to examine how customers behaved toward FDAs. Similar to this, Kaur et al. (2020) investigated customer opposition to using FDAs. The lack of comprehensive consumer behavior research restricts the knowledge offered to practitioners and researchers because the majority of results provided so far only touch on the technological aspects of FDAs. However, a number of academics have used the theory of consumption values (TCV) to explain consumer choice behavior. Nevertheless, few findings have demonstrated the values influencing the use of the FDAs (Kaur et al., 2021). TCV encourages comprehension of how consumers select a particular product from a range of goods and brands and decide whether to buy it or not (Sheth et al., 1991). This theory might have been applied to the online setting to learn more about how much value consumers place on digital products and services, especially how social media can improve brand experiences (Mäntymäki and Salo, 2015). TCV is a well-established and widely used theoretical framework used to study the nuances of purchase intention among consumers (Choe and Kim, 2019; Talwar et al., 2020) and continuous intentions (Dhir et al., 2020). TCV includes both cognitive and affective aspects of consumption and creates a multidimensional and holistic understanding of associated values (Tandon et al., 2021).

Second, not all facets of the hypothesis have been covered in prior research on TCV in the context of FDAs. The scale is made up of five parts: functional, social, emotional, conditional and epistemic value (Sheth et al., 1991). In the study of Wahyuningsih et al. (2022), TCV is used to examine customer satisfaction and online word of mouth regarding FDAs while combining all five components into one measure, namely, consumer value. This study’s findings only offer broad conclusions; they do not, however, adequately explain how functional, social, emotional, conditional and epistemic values relate to FDAs. In addition, Tandon et al. (2021) predicted customers’ purchase intent for FDAs by examining functional, social and conditional values. However, they have not looked at the aspect of emotional and epistemic values. To assess customers’ intention to continually buy food via mobile applications in the post-COVID-19 outbreak, the current study applies the TCV to cover research gaps. The following are the three research questions (RQs) for this study:

RQ1.

Which five consumption values influence consumers’ intention to continuously use food delivery applications in the post-COVID-19 outbreak?

Third, consumers are worried and afraid of the spread of disease during the COVID-19 situation, so food delivery services are an option to alleviate such concerns because there is a wide variety of food menus, as well as providing convenience, helping alleviate the hassle of going out to find food and drink, saving travel time and ensuring hygienic safety. During the outbreak of the virus, this kind of consumer behavior is clearly more prevalent. Nevertheless, there is an important question whether consumer behavior in using the FDAs will change once the situation returns to normal. It is likely that travel and in-store dining restrictions will continue to ease within the next five years. This should result in substantial growth in the food and travel related industries, so it is a doubt whether consumers will continue to use the FDAs. Although many consumers are currently using FDA services, it cannot be confirmed that they will continue to use them in the long term (Yap and Lee, 2023). Moreover, there are less consumers who want to use the same mobile applications repeatedly than application developers expected (Lee and Kim, 2019). Another important thing is the long-term sustainability of the industry’s significant digital transformation. It is necessary that businesses ensure that FDAs, which help mediate between restaurants and consumers, will continue to be used after the end of the COVID-19 pandemic (Kumar and Shah, 2021).

Moreover, few past studies have focused on factors influencing the continued use of FDAs across generations (Zhao and Bacao, 2020), especially in the post-COVID period. Age segmentation is a very useful tool for marketers (Thompson et al., 2018). Considering the diversity of consumer demographics, it is difficult to satisfy all customers with just one service or product (Dolnicar et al., 2018). Because each generation has different lifestyle preferences, targeting consumers therefore requires understanding the values and motivations of different generations (Rahulan et al., 2015). Each generation of consumers is shaped by different social environments, cultures and experiences. Therefore, they have different attitudes toward purchasing products or services. Marketers are required to provide them with different offers, customer experience and business models. The mass market segmentation is primarily classified into generation because each generation born and grown up in the same time period will go through the same experiences and milestones. As a result, people with similar social and cultural experiences are more likely to have similar sets of values, attitudes and behaviors (Kotler et al., 2021). Most past studies have suggested a link between the overall consumption value structure, without making comparisons between consumers in each generation. The market segmentation theory asserts that specific strategies are needed for each type of consumer (Zhou et al., 2020). In addition, researchers have previously used cohorts to examine generational differences in consumer behaviors, such as purchasing (Schade et al., 2016), online shopping (Brand et al., 2022), branding (Gentina et al., 2016) and online food purchasing behavior during the COVID-19 pandemic for Generations X, Y and Z (Wahyuningsih et al., 2022); however, those studies did not explore the intentions of Generations X, Y and Z to continue using FDA services after the COVID-19 pandemic. This research was therefore conducted with the aim of filling this gap.

Finally, the study’s findings are summarized in Kaur et al. (2021), who investigated the impact of gender and age on the link between TCV and purchase intention in relation to FDAs, which has divided consumers into two groups, namely, young users and old users, but did not study consumers in each generation. In addition, Wahyuningsih et al. (2022) conducted a comparative study of Generations X, Y and Z consumers who use online applications to purchase food from restaurants; that study proposed a model based on TCV. The differences between three generations were tested using analysis of variance (ANOVA) statistics, and the mean was compared to test how consumers across the three generations had different opinions toward various aspects of TCV. In this regard, ANOVA and the mean value calculation do not estimate causal–predictive relations. Due to the aforementioned research gap, the differences were compared using a multigroup analysis (MGA) technique with more than two groups (Henseler et al., 2009). This test combines the permutation and bootstrapping approaches to produce a criterion that is similar to the overall F-test. For assessing more than two groups, this method can be compared to ANOVA F-test while maintaining the familywise error rate and offering the necessary statistical power without relying on distribution assumptions (Cheah et al., 2023). The current study compares the moderating impacts of different generations in an effort to close research gaps. The following is a description of the RQs.

RQ2.

What are the differences in relationship between TCV and consumers’ intention to continuously use food delivery applications in the post-COVID-19 outbreak among Generations X, Y and Z?

2. Literature review

2.1 Theory of consumption values

Sheth et al.’s (1991) multifaceted theoretical framework revealed the generic consumption values theory, which explains why consumers pick one product or brand over another and why they buy or use a particular product. There are five different types of consumption values that affect consumer behavior: functional, social, emotional, conditional and epistemic values.

2.1.1 Functional value.

Functional value is defined as the advantage felt or attained from a situation’s functional, practical and physical performance (Sheth et al., 1991). Functional value takes into account a product’s performance, reliability, soundness and price benefits. Consumers’ rational buying behaviors and preferred products are primarily influenced by price, quality and value (Candan et al., 2013). According to utility theory, when consumers make a purchase decision, they always try to maximize their purchase utility, which is derived from the comparison between the benefit of the product and the cost incurred (Rizkalla and Setiadi, 2020). In the context of FDAs, functional value defined as the benefits and value for money, which FDA consumes perceived that they can ordering food at a reasonable price.

2.1.2 Social value.

The benefit that is perceived and acquired by one or more social groupings is referred to as social value. Through affiliation with stereotypical demographic, socioeconomic and cultural-ethnic groups, an alternative might gain social significance. In other words, goods that encourage social interaction among users or improve their social standing have higher social values. On the basis of a profile of chosen imagery, social value is assessed (Sheth et al., 1991). According to Wang et al. (2020), consumers typically gauge social value based on a product or service’s association with and relevance to their social group. This pertains to the perceived social status increase that consumers experience when using FDAs in the context of the current study (Talwar et al., 2020), as well as the improvement in self-image that consumers experience in their social contexts as a result of using FDAs to order food (Kaur et al., 2021).

2.1.3 Emotional value.

The emotional value is defined as the perceived utility acquired from an alternative’s capacity to arouse feelings or affective states (Sheth et al., 1991). Emotions are characterized as feelings or emotional reactions to elements such as situations, things, commercials and brands in the context of consumer behaviors. The advantage derived from the emotional and sensual state is known as emotional value. This value is connected to how consumers react to a product (Candan et al., 2013). Products and services are frequently linked to strong emotions (such as the romance sparked by a candlelit supper or the dread sparked by a scary movie) (Sheth et al., 1991). This study defined emotional value as a product’s ability to elicit emotions or affective states in the ordering via FDAs.

2.1.4 Conditional value.

The perceived utility gained by an option as a result of the particular circumstance or conditions facing the decision-maker is referred to as the conditional value of an option. When physical or social conditions exist that increase an option’s functional or social worth, it gains conditional value. On a profile of desirable contingencies, conditional value is evaluated. The usefulness of a choice frequently depends on the circumstances (Sheth et al., 1991). Conditional value is concerned with the temporally limited benefits that a good or service offers, depending on a particular situation (Wang et al., 2020). This study defined conditional value as the “features, advantages and merits that drive value derived by consumers when the users purchase food via FDAs, such as the listing of favorite restaurants, delivery fee discounts” (Kaur et al., 2021) and better promotional schemes (Talwar et al., 2020).

2.1.5 Epistemic value.

The ability of an option to pique interest, offer novelty and/or satisfy a need for information is what is referred to as the epistemic value of that option. A product would be seen as having epistemic value if it could provide customers with novel experiences that would satisfy their requirements for novelty, curiosity and knowledge. Epistemic value is undoubtedly provided by the brand-new experiences (Sheth et al., 1991). According to this study, epistemic value is the novelty and sense of interest sparked by seeing advertising and other people using FDAs, all of which can increase the value associated with using FDAs (Kaur et al., 2021).

2.2 Continued usage intentions

As it is claimed that the cost to acquire a new client can be up to five times more than the cost to keep an existing customer, ongoing usage is more important than initial usage (Bhattacherijee, 2001).

Studies on the intention to continue using information systems (IS) have multiplied recently and now encompass a variety of scenarios, including the intention to keep using mobile applications, m-services and m-commerce, among others (Franque et al., 2020). Because it is thought to be crucial for the long-term viability of an IS, it is desired to explore the direct effects on the continuity intention of using mobile applications, even if the majority of earlier research on these systems strongly stressed the initial acceptance (Silva et al., 2022). It is challenging to comprehend the continued usage intention of consumers from the perspective of IS alone, and in FDAs systems, it is necessary to understand the factors that influence the willingness of consumers to continue usage (Wang et al., 2020). For the FDAs system, unlike ordinary mobile applications, it integrates different online and offline entities, such as restaurants or vendors, delivery staff and customers. Therefore, it is important for relevant stakeholders to comprehend the drivers behind consumers’ continued use of FDAs in the context of the COVID-19 outbreak (Zhao and Bacao, 2020).

2.3 Market segmentation

Smith (1956), who underlined that market segmentation offers a reasonable and more exact adjustment of product and marketing effort to consumer or user requirement, was the first to recognize and establish market segmentation in academic literature. Due to the intense market competition of today, many businesses are now required to understand the competitive factors that influence the particular market to which the segmentation strategy will ultimately apply and then choose the most advantageous segmentation and the market and customer demands can be better understood with the use of effective segmentation (Zhou et al., 2020). It is almost hard to please every consumer with a single service or product due to the diversity of consumers. Therefore, the segmentation of the market into smaller, more focused and generally homogeneous consumer segments provides the business with a number of competitive advantages, and segmentation greatly improves comprehension of the traits of the relevant market, consumer behavior prediction, the discovery and exploration of new market prospects and identification of potential groupings (Dolnicar et al., 2018). Recently, many researchers have planned to partition the market based on the generational divide. This notion is based on the observation that the environmental background of the development process, including music, sports, regulations and many historical events at the time, has had a significant impact on each generation. These teams are referred to as “cohorts” by demographers. Cohort members share similar fundamental experiences (Rodríguez et al., 2015).

2.4 Generation cohort theory

Generation cohort theory considers that people who were born in the same time period, grow up in similar historical, social, cultural, political and economic environments share similar preferences for certain things in life, reside in the same community and encounter similar external experiences in their late adolescence and early adulthood. They share comparable values, beliefs, attitudes, preferences and traits because of acquired knowledge and common experiences as each generation cohort goes through key events together (Inglehart, 1997; Rahulan et al., 2015). Because each generation has its own beliefs about the kind of desired lifestyle, it is crucial to understand the values and motivations of each generation to target particular consumers (Rahulan et al., 2015). By concentrating on various generations, it is easier to segment the market and develop more effective strategies and product positioning (Brand et al., 2022). Because members of a cohort are thought to have comparable values, segmentation based on generation cohorts has become a very helpful tool for marketers (Thompson et al., 2018). To compare the differences between consumer generations in terms of shopping (Schade et al., 2016), consumer decision-making (Solka et al., 2011), online purchase (Brand et al., 2022) and online food ordering behavior (Wahyuningsih et al., 2022), researchers use generation cohorts. Generations X, Y and Z, which make up the segmentation used in this study, have been divided into three categories.

2.4.1 Characteristics of Generation X.

According to Koksal (2019), Generation X refers to those who were born between 1965 and 1979. They are one of the most educated generations and are known for their skepticism, pragmatism and awareness of risk avoidance (Gurau, 2012). Generation X people are those who were not born in the digital age but eventually accepted many or most features of modern technologies (Calvo-Porral and Pesqueira-Sanchez, 2020). Due to their higher level of discretionary income, they have a higher purchasing power than preceding generations. They want to shop with detailed knowledge about the products. When making new purchases, they are more cautious about product qualities, particularly those of internet information (Brand et al., 2022). With greater experience and family responsibilities, they are likely to seek high-quality products and services, make a comparison and try to receive adequate information about prospective purchases (Dabija et al., 2018).

2.4.2 Characteristics of Generation Y.

Individuals who were born between 1980 and 1994 are referred to as members of generation Y (Koksal, 2019). They have grown up around technology and have been in regular contact with digital media and the internet. They are more optimistic about information and communication technologies than earlier generations, which has affected their behavior, way of thinking and learning process. They heavily use mobile devices, the internet, connection, interactive media and social networks, and technologies have been incorporated into their daily lives (Calvo-Porral and Pesqueira-Sanchez, 2020). They are more knowledgeable about technology than their parents, with the tendency to be lazier and the behavior of consumerism (Mavilinda et al., 2022). They are not loyal customers while choosing products that match their personality and lifestyle; on the contrary, they are more likely to remain loyal to firms that share their social and civic ideals (Yan et al., 2019). The community has adopted the usage of FDAs as a new way of life, especially among Generation Y, who are potential FDA users (Mavilinda et al., 2022).

2.4.3 Characteristics of Generation Z.

Generation Z is defined as the individuals who were born between 1995 and 2000 (Koksal, 2019). Technology and internet are essential for this generation that is growing with the sudden advancement of computers and technology. The “mobile generation” is another term for this generation (Kapi and Roy, 2014).

Marketers are highly interested in this segment because Generation Z purchasing power is five or six times higher than previous generations (Djafarova and Foots, 2022). They anticipate ongoing innovation from the carefully chosen products because they have a wide range of options for goods and services. The preferred product is expected to promote convenience in their lives. They conduct research on the product before making a purchase (Özkan and Solmaz, 2017). They are likely to use electronic devices to suggest and spread their experiences to others [electronic word-of-mouth (eWOM)]. They do not keep their opinions to themselves when they have a good or bad experience with a business; instead, they post their opinions or criticism online, on social media (Wahyuningsih et al., 2022). Ordering food through a delivery service is viewed as sensible and practical by Generation Z. They prefer to place weekly as well as daily internet food orders. Worldwide, Generation Z consumers are increasingly using food delivery services (Banerjee et al., 2019).

2.5 Research framework and formulation of hypotheses

2.5.1 Research between functional value and continued usage intention.

Talwar et al. (2020) discovered a link between functional value and purchase intent for online travel agents. According to earlier research, functional value and usage intention are positively correlated (Kim et al., 2018). In addition, Yeo et al. (2017) hypothesized that customers’ perceptions of FDAs’ utility and their intention to continue using them are influenced by their ability to save money. Functional value is positively associated with FDAs’ buying intention, according to a study by Tandon et al. (2021). The following hypothesis was put forth to test this assumption based on the facts mentioned above:

H1.

Functional value has a significant positive impact on continued usage intention toward FDAs in the post-COVID-19 outbreak.

2.5.2 Research between social value and continued usage intention.

Kaur et al. (2021) discovered that among Indian consumers, prestige or social value could influence purchase intention. According to Dhir et al. (2020), social value was positively correlated with the intention to consistently use mobile messaging apps and online travel services. Similar to this, Talwar et al.’s (2020) earlier research discovered a relationship between social value and the propensity to use internet travel agents. Furthermore, Kaur et al. (2018) discovered that social value had a limited impact on people’s continued use of social media brand communities. In the context of FDAs, according to the study of Zhao and Bacao (2020), which explored the factor determining customers’ continuous use of food delivery apps during the COVID-19 pandemic, it was found that social value has direct positive impacts on users’ intention to continuously use FDAs. Therefore, the following hypothesis was proposed:

H2.

Social value has a significant positive impact on continued usage intention toward FDAs in the post-COVID-19 outbreak.

2.5.3 Research between emotional value and continued usage intention.

Accordingly, earlier research discovered a significant positive association between emotional value and purchase intention for online travel companies (Talwar et al., 2020). In a similar vein, Kaur et al.’s (2018) study discovered that emotional value had a role in the ongoing use of social media brand communities. Mäntymäki and Salo (2015) found that the primary variable impacting teens’ online spending was emotional value. In addition, Yeo et al. (2017) developed their model to identifying the key predictors of customers’ desire to consistently use online food purchase systems. Customers were observed to think that using such online food ordering platforms was fun. As a result, the hypothesis was created as follows:

H3.

Emotional value has a significant positive impact on continued usage intention toward FDAs in the post-COVID-19 outbreak.

2.5.4 Research between conditional value and continued usage intention.

Prior research found a significant positive association between conditional value and purchase intention for online travel firms (Talwar et al., 2020). In a similar vein, Hu and Yang (2019) discovered that promotional discounts have an impact on consumers’ purchasing decisions throughout the booking phase. Affordances were described as conditional value in the context of FDAs by Kaur et al. (2021), and it was discovered that affordances had a considerable impact on Indian consumers’ intent to purchase via FDAs. Finally, Tandon et al. (2021) found a favorable correlation between conditional value and FDAs purchase intention. Based on the aforementioned data, the following hypothesis was proposed to test this premise.

H4.

Conditional value has a significant positive impact on continued usage intention toward FDAs in the post-COVID-19 outbreak.

2.5.5 Research between epistemic value and continued usage intention.

Previous research suggests a favorable association between epistemic value and consumption behavior in relation to teenagers’ online shopping (Mäntymäki and Salo, 2015) Accordingly, research has shown a link between epistemic value and the consumption of healthy foods (Thomé et al., 2019). The relationship between customer value, satisfaction and behavioral intention was also examined by Williams et al. (2017) across two culturally distinct groups of Japanese and Western adventure travelers. According to the report, Japanese tourists value novelty and curiosity more than Western tourists. The same study (Talwar et al., 2020) discovered that epistemic value significantly has a positive link with purchase intention toward online travel companies. Thus, the hypothesis was proposed to test this assumption:

H5.

Epistemic value has a significant positive impact on continued usage intention toward FDAs in the post-COVID-19 outbreak

2.5.6 Generation cohort as a moderating role in the multigroup analysis.

The importance of moderating variables in elucidating individual differences in consumer behavior has been studied by academics. For instance, Tan and Ooi (2018) found that gender and age only slightly affect consumers’ choices when making purchases through mobile applications for travel-related goods. The moderating effect of age on the relationship between premium price payment, social network and intention toward online travel agents was also discussed by Ye et al. (2017). However, it is possible that each customer group will perceive distinct values (Bhatnagar and Ghose, 2004). The difference in intentions to use FDAs between Generations X, Y and Z found that Generation Z is particularly interested in emotional value and functional value, which includes enjoyment, ease of ordering payment, not having to prepare the food, not having to get dressed up and transaction security more than Generations X and Y (Wahyuningsih et al., 2022). Moreover, due to their sensitivity to societal issues on social media, Generation Z prioritizes values while making purchases. They are more likely to support essential ideas with brands or products and avoid those with ideas or behaviors at odds with their own values (Smaliukiene et al., 2020). The rationale behind the cohorts online socializing is the necessity to be involved and informed with other people (Yadave and Rai, 2017). From the above literature, it is probable that social value is quite important for this generation. Moreover, in comparison to Generations X, Y and Z, Generation Y particularly exhibits concern for epistemic value more than generation X and Z (Wahyuningsih et al., 2022). For the functional value, the Generations Y and Z perceive service quality and value differently (Yarimoglu, 2017). Furthermore, conditional value in terms of safety value is different across three generations X, Y and Z. Particularly, Generation X shows the highest concern over conditional value more than Generations Y and Z. From the discussion as mentioned above, we anticipate that the association between consumption value and intention toward FDAs will be moderated by the generation cohort of the consumers. As a result, the following hypotheses were put out to test this assumption (Wahyuningsih et al., 2022):

H6.

There are significant differences in relationship between functional value and continued usage intention toward FDAs in the post-COVID-19 outbreak among Generations X, Y and Z.

H7.

There are significant differences in relationship between social value and continued usage intention toward FDAs in the post-COVID-19 outbreak among Generations X, Y and Z.

H8.

There are significant differences in relationship between emotional value and continued usage intention toward FDAs in the post-COVID-19 outbreak among Generations X, Y and Z.

H9.

There are significant differences in relationship between conditional value and continued usage intention toward FDAs in the post-COVID-19 outbreak among Generations X, Y and Z.

H10.

There are significant differences in relationship between epistemic value and continued usage intention toward FDAs in the post-COVID-19 outbreak among Generations X, Y and Z.

According to the literature review and theoretical background, the conceptual framework of this study was established as shown in Figure 1.

3. Methodology

3.1 Sample characteristics

Three categories of Thai smartphone users make up the study’s target demographic. The first group consists of Generation X, who were born between 1965 and 1979 and range in age from 44 to 58. Generation Y, those born between 1980 and 1994 (aged 29–43), makes up the second group, followed by Generation Z, those born between 1995 and 2000 (aged 23–28) (Koksal, 2019). Three generation cohorts are considered as prospective consumers of FDAs in Thailand market with strong purchasing power (Statista, 2022c), and these three target groups have experienced in purchasing food from the popular food delivery applications, including GrabFood, Food Panda or LINE MAN (Statista, 2022b), at least once a week for three months (Kaur et al., 2021). Using these sample representatives, we expect to be able to confidently extrapolate the findings to the complete group of FDA users.

3.2 Sample size

Prior to doing the MGA, researchers must ensure that the total number of observations in each group meets the minimum sample size requirement. Large sample sizes are the best method to ensure statistical power for both groups. To determine if the lack of any moderating impact is due to low statistical power or the absence of a real moderating effect, a statistical power should be established at 80% (Hair et al., 2022). With the following settings: the effect size f² = 0.15 (medium), α = 0.05, number of predictors = 5 and power set at 95%, the minimum acceptable power was calculated using the G*power 3.1.9.7 program, whereas the lowest acceptable power is 80% (Gefen et al., 2011). For each group, the test recommended a minimum sample size of 138 cases. In this study, the sample size of each generation meets the requirement as follows: Generation X is 242, Generation Y is 215 and Generation Z is 288.

3.3 Data collection procedure

The survey was distributed via Facebook and Instagram using Google’s online survey by disseminating the tool through Instagram and Facebook, where that the best creator performance on social media in the food and dining category, namely, Mawinfinferr, TidReview, Stavingtime, BanKii and KP.Eat.Travel (Thailand Social Awards, 2023), to target respondents all over Thailand during the first week of September to November 2022. To make sure that every response met the requirements for inclusion, the following three screening questions were used:

Q1.

Are you in the Generation X range of 44–58 years old, the Generation Y range of 29–43 years old or the Generation Z range of 23–28 years old?

Q2.

Have you ever purchased food from food delivery applications, namely, GrabFood, Food Panda or LINE MAN?

Q3.

Did you order food via food delivery applications at least once a week for three months?

An online survey may be distributed to participants extremely easily and at a very low administrative cost, which makes data entry and analysis much simpler (Evans and Mathur, 2005). The research goals and the survey’s inquiry’s strictly academic aspects were initially explained to the participants. Clarification of the voluntary nature of participation allowed participants to withdraw at any moment during the process, and the same internet protocol address was only allowed to submit data once to prevent repetitive responses. In total, 978 people completed the survey, but only 745 of the 878 observations (76.17%) passed the requirements for inclusion.

4. Data analysis and results

4.1 Descriptive analysis

There were 454 participants in this study who completed the questionnaire. The majority of respondents (n = 356, 47.79%) were females, followed by men (n = 245, 32.89%) and people of other genders (n = 144, 19.33%). The majority of them were Generation Z, at the age of 23–28 years old (n = 288, 38.66%), followed by Generation X, at the age of 44–58 years old (n = 242, 32.48%) and Generation Y, aged between 29 and 43 years old (n = 215, 28.86%). Regarding the education level, 336 respondents (45.10%) held a master’s degree. A total of 425 respondents were private company employees (57.05%) and 324 respondents earned 30,001–50,000 baht per month (43.49%). Table 1 provides a summary of the respondents’ demographic information.

4.2 Data analysis

The measurement and structural model were assessed in the current investigation using the partial least squares structural equation modeling (PLS-SEM) method, version 4.0. MGA inside a PLS path modeling framework is one of the most efficient ways to assess moderation in many interactions, claimed by Hair et al. (2022). Standard moderation investigates a single structural relationship at the point of interaction between exogenous and endogenous variables and significantly improves researchers’ capacity to recognize significant variations in a variety of relationships across particular groups (Schlägel and Sarstedt, 2016). The aforementioned information explains why PLS-SEM was used in this study’s analysis.

4.3 Common method variance and nonresponse bias test

To examine common method variance, the single-factor test by Harman (Podsakoff and Organ, 1986) was used in this study. Principal component analysis, as recommended by Tehseen et al. (2017), is made to carry out the test. According to Kock (2021), all indicators passed the test based on the single-factor findings of the unrotated principal axis factoring analysis, which showed a variance of 37.770%, which is less than 50%. Therefore, there is no indication of common technique bias in this study. In other words, there are no significant worries that might influence how variables relate to one another. In addition, an ANOVA test was performed on a set of randomly chosen measurement items that were answered by both sets of respondents to make sure that there was no issue with nonresponse bias in this study (Armstrong and Overton, 1977). The findings indicate that there was no response bias in the data collection because there was no significant difference between early and late responders at p > 0.05.

4.4 Measurement model analysis

To verify the instruments’ reliability and content validity, a pretest of 30 copies was conducted. Following the completion of the surveys by 745 respondents; the construct reliability was examined, the constructs’ internal consistency is measured by Cronbach’s alpha and composite reliability and scales currently in use are regarded satisfactory if these values have a minimum value of 0.7 (Nunnally, 1978), which every item exceeded the recommended threshold; and regarding the convergent validity, all values exceeded the criteria for the average variance extracted, which is generally advised to be at least 0.50 (Fornell and Larcher, 1981), and the constructs’ items have outer loadings over 0.6, as suggested by Hair et al. (2010). Tables 2 and 3 include the results.

The heterotrait–monotrait (HTMT) ratio of correlations technique was used to assess the discriminant validity in this case. As shown in Table 4, each construct’s HTMT correlation ratio is less than 0.9, which is in accordance with the HTMT threshold value (Henseler et al., 2015). As a result, the measurement model’s discriminant validity was guaranteed.

4.5 Structural model analysis and hypothesis testing

According to Diamantopoulos and Siguaw (2006), if the variance inflation factor (VIF) value is less than or equal to 3.3, multicollinearity has no impact on the validity of the results. It is confirmed that multicollinearity was not a significant problem in the investigation because all of the VIF values for each hypothesis were less than or equal to 3.3. The results of the VIF are shown in Table 5.

The study hypothesis test was carried out using a bootstrapping technique using a 5,000-resampling process. A statistical metric in a model that indicates the effectiveness of the independent variables in expressing variance in the dependent variable is the coefficient of determination, abbreviated R². For the overall model, Generation X, Generation Y and Generation Z model, the R² value of continued usage intentions is 0.454, 0.580, 0.452 and 0.489, respectively. According to Cohen (1988), the R² value met the threshold requirement of 0.20. Figures 2–5 display the R² value’s findings. The results of hypothesis testing for overall model are shown in Figure 2 and Table 4, and Table 6 presents the results of MGA, using the Henseler’s MGA method.

4.6 Measurement invariance of composites

Prior to doing a MGA comparing two or more groups, Henseler et al. (2016) recommended the measurement invariance of composites (MICOM) test when conducting PLS-SEM studies. Because the MICOM test also helps ensure that the measurement models specify measures of the same attribute under diverse situations and safeguards the validity of results. This study illustrates the permutation-based invariance measurement testing for each generation. Comparing and interpreting the group-specific differences of PLS-SEM results requires invariance measurement testing.

The MICOM protocols have three steps, which are as follows: the invariance evaluation described in Step 1 ensures the same indications, data processing and algorithm setting/criteria; Step 2 involves evaluating compositional invariance to confirm that the initial correlation was over 5% and the permutation p-value was over 0.05, then this invariance is established; Step 3 involves determining whether the mean and variance of the groups were identical. Full invariance is indicated by both mean and variance falling between the 2.5% and 97.5% borders, whereas partial invariance is indicated by none or just one mean or variance lying between these bounds. The hypothesis testing is feasible if the study successfully completes the three necessary steps (Henseler et al., 2016; Dewi et al., 2019). In Table 6, the MICOM results are shown.

4.7 Multigroup analysis

The Generations X, Y and Z data files were tested for differences using PLS MGA. Henseler’s MGA approach served as the foundation for the results. This method states that a 5% level of significant differences between particular path coefficients across the three groups is indicated by a p-value of differences between path coefficients below 0.05 or over 0.95 (Henseler et al., 2009). Table 7 shows the MGA, which found that four hypotheses were supported while one was not.

5. Discussion and conclusions

5.1 Discussions

This study provides evidence that functional value has a positive influence on continued usage intentions toward FDAs in the post-COVID-19 outbreak (H1). This result further supports prior studies of Rizkalla and Setiadi (2020) and Tandon et al. (2021). This shows the significance of factors, such as offered benefits, convenience or various menus, reasonable price and consumers’ evaluation of whether the purchase is valuable for money in respect of cost–quality.

The second finding revealed that social value has a positive influence on continued usage intentions (H2). This result contradicts Tandon et al.’s (2021) conclusion that there is a negative correlation between social value and FDA purchase intention, but it is in line with Talwar et al. (2020) and Kaur et al. (2021). Customers who purchase food using FDAs see their personal status and improved self-image, and they associate their dining location with their personal status (Kim and Lee, 2017). By using cutting-edge technology to order meals, FDAs are seen as a source of enhancing social perception and building favorable impressions, increasing their consumption value thought to be a measure of personal status. Such value will consequently boost their desire to use FDAs.

As for the third finding, emotional value does not influence continued usage intentions (H3). This result is in line with earlier research by Kaur et al. (2020) and Tandon et al. (2021). FDA users are likely stimulated by the design and aesthetics, but the excitement may come from starvation rather than from the app’s design. Given the lack of significance in the association between emotional value and continued usage intentions, it is possible that customers are unaware of the FDAs’ original arousing feeling (Kumar and Shah, 2021). Moreover, food ordering via the FDAs focuses more on the benefits of convenience and travel time saving rather than amusement from using the application for shopping, watching movies or listening to music.

Regarding the fourth finding, conditional value was found to have a positive impact on continued usage intentions (H4). The findings of Talwar et al. (2020) are supported by this finding. The consumers want to take use of FDAs in more ways. Our research suggests that consumers’ perceptions of favorable outcomes, such as free delivery, availability to favorite restaurants, reasonable pricing, promotional offers and discounts, can be used to increase FDA’s continued usage intention. The fifth finding (H5) proposed that epistemic value was a powerful motivating factor toward continued usage intention. This finding contrasts with the prior study by Tandon et al. (2021), which conducted the qualitative study to explicate FDAs’ specific values. They eliminated epistemic value of the TCV in the context of FDAs. However, Thomé et al. (2019) confirmed an association between epistemic value and consumption behavior. Advertisements and a sense of curiosity brought on by seeing other people use FDAs can influence consumers’ continued usage intentions on FDAs favorably. If the consumers see their friends and other familiar people buying food from FDAs, they are more inclined to follow suit (Johnson et al., 2018). In addition, buying food through FDAs offers them novelty and an experience that is distinct from dining in a restaurant.

The findings of MGA compare the difference in consumption values theory among Generations X, Y and Z, which extends the current body of knowledge because the previous studies have not examined the consumption value theory among generation cohorts. The findings indicate significant difference in relationships among functional, social, conditional and epistemic values and continued usage intention toward FDAs between Generations X, Y and Z; thus, H6, H7, H9 and H10 were supported. On the contrary, H8 was not supported; emotional value has no significant.

Generation Y consumers did not give importance to the factors of functional value, while it has an impact on continued usage intentions among Generations X and Z consumers, with a higher level for Generation X than Generation Z consumers. Generation X have family responsibilities in providing care for their children and elderly family members. Thus, they were a group of people who spent their money cautiously while looking for products or services that offer multiple benefits, with reasonable price and value for money. This is consistent with Generation Z, who prefer to lower spending because they have been through economic downturns more frequently than other generations (Özkan and Solmaz, 2017). They saw their parents struggling to survive the financial crisis, so they were aware of spending money cautiously more than Generation Y. The values sought by Generation Z consumers are convenience in placing order, payment, not necessary to cook, not necessary to dress and preferred products or services for convenience in their lives.

For social value, Generation X consumers do not pay attention to this factor, while social value has an influence on continued usage intentions among both Generations Y and Z consumers, with a higher level for Generation Y than Generation Z consumers. Generation Y habitually like to compare themselves to their friends. They would like to gain recognition from their peers or society (Kotler et al., 2021). They participate in an online community to share their values. They seek to portray their own image, impress their friends with their product or service knowledge and are particularly interested in how other group members will react (Rungruangjit et al., 2023). They do not have a strong brand loyalty and just shop for things that fit with their personalities and way of life. They are more likely to remain loyal to firms that share their social and civic ideals.

On the contrary, emotional value has not affected significant difference in continued usage intentions toward FDAs among three generations. The excitement of Generations X and Y on the online platform or from using applications mostly arises from shopping because it makes them feel relaxed from hard work (Dabija et al., 2018). In addition, Generation Y consumers like using social networks, blogging and other forms of social media to create and share information (Noble et al., 2009). Meanwhile, Generation Z enjoy participating in brand activities on social media and expect brands to offer something exciting and new interactive experiences (Kotler et al., 2021). The use of FDAs is focused more on benefits, convenience, time-saving and value for money, rather than fulfilling the feeling of fun and excitement.

Furthermore, the result provides interesting new finding about the conditional value, that the Generation X consumers focus on this factor, while it has no influence among consumers in both Generations Y and Z. Conditional value, comprising fast delivery, promotional incentives, free delivery and listings of excellent restaurants are expected to boost Generation X consumers’ consumption value and their intention to continue using FDAs for food ordering. They are more inclined to purchase meals from FDAs that provide better bargains in terms of delivery costs, restaurant quality and the types of restaurants advertised.

Finally, Generation X consumers do not pay attention to the factor of epistemic value. This is probably because Generation X consumers exposure to online advertising does not exert a major effect on their buying decisions, because they have not much interest in such advertising and often ignore them. Family, friends or school are their predominant values, leading to more pragmatic and flexible behavior, as well as more cautious and skeptical of new things (Dabija et al., 2018). Meanwhile, epistemic value has influence on continued usage intention among consumers in Generations Y and Z, with a higher level for Generation Y than Generation Z consumers. Generation Y consumers often like to compare themselves to friends. What their friends speak or buy has a big influence on them. They trust their friends more than brands that have been popular in the market (Kotler et al., 2021). Their decision to buy products or services has been influenced by online comments and suggestions, which is considered more reliable than the data publicized by brands (Dabija et al., 2017). They will also take into account whether the eateries use FDAs to present fresh, innovative and substantial information on their menu items (Wahyuningsih et al., 2022).

5.2 Theoretical contributions

The study also expands the TCV theory so that it can be used in the context of FDAs. The proposed inclusive approach substantially advances the growing body of research on FDAs. Therefore, by investigating digital platforms such as FDAs, the TCV’s application is expanded to include the study of the food retail industry, and the empirical study revealed that the key elements of TCV influencing continued usage intentions toward FDAs consisted of only four dimensions as follows: functional, social, conditional and epistemic values, whereas emotional value has no influence in the context of FDAs. Thus, we recommend focusing on four dimensions of TCV and eliminating the emotional value in the context of FDAs.

Moreover, the finding contributes to the FDAs literature and TCV involving customer segmentation, while there are significant differences in relationships among four dimensions of TCV and intention to continue using FDAs in the post-COVID-19 outbreak between Generations X, Y and Z. The segmentation of the market is the key contribution of this study. Marketing literature about the various values felt by Generations X, Y and Z is improved by online purchasing behavior and diverse value perceptions among three generation groups in FDAs. Researchers might use our findings to examine how technology-oriented services, such as FDAs, are seen differently by each generational cohort and how that affects customers’ intentions to use such goods and services over time. The structural model’s many generational perspectives were scientifically evaluated and useful for marketing information, particularly in terms of market segmentation literature and consumer patterns. Marketers are urged to segment the market not just based on demographic information, such as age or income, but also based on generational cohorts. As a result, this study adds a number of new chapters to the literature on generational cohorts that are pertinent to FDA’s operations.

Finally, this study contributes to explore the data regarding continued usage intention after the COVID-19 outbreak, which is hardly examined in the FDAs’ context. Most studies emphasized user’s likelihood to use, intention to adopt or satisfaction of consumers with FDAs, during the COVID-19 outbreak. This study establishes the validity of exploring continuous usage of FDAs instead of adoption during a crisis situation. Thailand’s online food delivery sector is among those with the quickest growth rates in Southeast Asia. The study’s findings add to our understanding of customer attitudes on the usage of FDAs from the standpoint of a developing nation.

5.3 Managerial implications

To compete for greater market share, managers of FDA-regulated enterprises can benefit from this study’s practical advice. It also urges marketers to gain a full understanding of the particular aspects that affect how much Generations X, Y and Z customers perceive derived value. First, as for Generation X consumers, emphasis should be placed on delivering functional and conditional values so as to make these consumers perceive the value for money, offer good value for money, offer various benefits, offer discounts or coupons on each visit, reduce the delivery charges or offer free delivery, provide better promotional incentives, along with limited-time free subscriptions, free meals after a specific number of deliveries and freebies in collaboration with other businesses. Moreover, they are concerned about conditional value, which comes in the form of free and quick delivery. FDA service providers should look on ways to reduce delivery-related expenses and speed up food delivery. For instance, FDAs can avoid traffic jams by using drones to transport food, which has been tested in Korea (Hwang et al., 2019). Even though it is costly to acquire drones, the long-term benefits are expected to offset the cost, including the decrease in number of delivery personnel.

Second, Generation Y consumers should emphasize epistemic and social value. Through smart marketing that targets current users, such as eWOM, significant referrals (family, friends or coworkers), online reviews and advocacy, FDA companies can increase their Generation Y client base, which is considered more reliable than the data provided by companies. For the social value, they habitually require acceptance from friends or society. They would like to impress their friends with their knowledge on products or services, describe their own image, as well as being quite curious about reactions from other members. They are likely to connect prestige value to the use of these apps. Companies can formulate strategies, such as recognizing the most frequent user in a specific area within a certain period on their Facebook page or TikTok, and should create campaign to provide a space to express opinions, knowledge or talent, such as providing a space to review food delivery services, review menus, recommend great restaurants, hold a contest for the best local dishes or organize a user-generated content contest about creative food.

Finally, factors influencing Generation Z consumers are epistemic, functional and social values. They habitually like to do fun activities to engage with brands. They expect brands to offer excitement, new offerings all the time and new interactive experiences. Thus, companies should create interactive activities with experience, which is totally different from eating at a restaurant, especially via social media, such as TikTok challenging campaign. Moreover, the marketers should place emphasis on advertisements showing the use of the FDAs to order food quickly and conveniently, with contactless deliveries and other security measures to raise confidence amid the pandemic or PM2.5 pollution. In addition, they are aware of spending money reasonably. Therefore, marketers should communicate to this group of consumers the value for money of the products or services, time value and more benefits to receive, compared to going out for a meal. In addition, it is necessary to ensure the apps be up-to-date with high-resolution pictures of food and provide information of newcoming restaurants or food catering services.

5.4 Limitations and future research

The findings may not apply to FDAs in other nations or to other mobile apps used for making reservations for or purchasing products and services because the focus of this study is FDAs in the Thai market alone. This limited generalizability can be addressed, and researchers working to validate the findings of this study in diverse cultural contexts, particularly in countries where FDAs are often used, can help shed further light on this problem. We advise academics to expand this research area and provide a more complete comprehension of FDA consuming behavior by using additional fundamental theories from a variety of domains. For instance, one explanation for why customers reject the usage of FDAs is the innovation resistance argument. Customer experience frameworks can also be used to understand the sensory aspects of FDA purchases, as well as the influence of user reviews and app design elements (such as background color) on purchasing choices.

Figures

Figure 1

Conceptual framework

Figure 2

Results of structural model (overall model)

Figure 3

Results of structural model (Generation X)

Figure 4

Results of structural model (Generation Y)

Figure 5

Results of structural model (Generation Z)

Table 1

Characteristic of the respondents (N = 745)

Characteristic	Categories	Frequency	%
Gender	Female	356	47.79
	Male	245	32.89
	Other genders	144	19.33
Age	44–58 years old (Generation X)	242	32.48
	29–43 years old (Generation Y)	215	28.86
	23–28 years old (Generation Z)	288	38.66
Education	Below bachelor’s degree	104	13.96
	Bachelor’s degree	263	35.30
	Master’s degree	336	45.10
	Doctoral degree	42	5.64
Occupation	Government officers/employees	165	22.15
	Students	38	5.10
	Private company employees	425	57.05
	Business owners	117	15.70
Monthly income	Under 10,000 baht	33	4.43
	10,000–30,000 baht	246	33.02
	30,001–50,000 baht	324	43.49
	Above 50,000 baht	142	19.06

Source: Created by authors

Table 2

Results of measurement model (overall model)

Constructs	Items	Loading	CA	CR	AVE
Functional value (FNV) (Talwar et al., 2020)			0.854	0.856	0.774
	FNV1: FDA charges an affordable price	0.883
	FNV2: FDA is valuable for my money	0.896
	FNV3: FDA consistently offers various benefits, such as convenience or variety of menus	0.860
Social value (SCV) (Choe and Kim, 2019; Talwar et al., 2020)			0.930	0.931	0.744
	SCV1: Using FDA enhances my self-concept to others	0.824
	SCV2: Using FDA helps me gain social acceptance	0.907
	SCV3: Using FDA helps create a positive impression on other people	0.907
	SCV4: Using FDA helps you stand out of your peers	0.905
	SCV5: Using FDA enables me to demonstrate my online food ordering experiences to others	0.756
	SCV6: I have a feeling of having social status when consuming food purchased from well-known FDAs	0.867
Emotional value (EMV) (Huang et al., 2017; Xu et al., 2020)			0.894	0.898	0.614
	EMV1: I feel happy while using the FDA	0.816
	EMV2: I feel pleased with using the FDA	0.817
	EMV3: I feel satisfied with using the FDA	0.749
	EMV4: I feel excited when I indulge in using the FDA	0.662
	EMV5: I feel energetic when using the FDA	0.837
	EMV6: I feel free when using the FDA.	0.831
	EMV7: I seem to have control over the FDA	0.758
Conditional value (CDV) (Talwar et al., 2020; Kaur et al., 2021; Tandon et al., 2021)			0.852	0.854	0.630
	CDV1: I would use FDA more frequently if FDA providers reduce the delivery fees	0.818
	CDV2: I would use FDA more frequently if FDA providers offer free-of-charge delivery	0.812
	CDV3: I would use FDA more frequently if better promotional incentives were given	0.799
	CDV4: I would use FDA more frequently if FDA providers shortened the delivery time	0.716
	CDV5: I would use FDA more frequently if more of my favorite restaurants joined these platforms	0.817
Epistemic value (EPV) (Candan et al., 2013; Johnson et al., 2018; Rungruangjit, 2020)			0.815	0.818	0.645
	EPV1: I use FDA for ordering food to provide me with novelty	0.797
	EPV2: I use FDA for ordering food to have a unique experience totally different from eating at a restaurant	0.859
	EPV3: All my friends are making use of FDAs	0.833
	EPV4: I have seen advertisement about FDAs	0.718
Continued usage intentions (CUI) (Venkatesh et al., 2012; Zhao and Bacao, 2020)			0.879	0.880	0.734
	CUI1: I intend to return to my favorite FDA again	0.811
	CU2I: I plan to continue using my favorite FDA frequently	0.902
	CUI3: I will always try to use FDA in my daily routine	0.836
	CUI4: I intend to continue using my favorite FDA in the future	0.875

Source: Created by authors

Table 3

Results of measurement model (Generations X, Y and Z)

		Generation X (242)				Generation Y (215)				Generation Z (288)
Constructs		Loading	CA	CR	AVE	Loading	CA	CR	AVE	Loading	CA	CR	AVE
Functional value (FNV)			0.862	0.885	0.783		0.875	0.876	0.800		0.827	0.829	0.743
	FNV1	0.914				0.887				0.854
	FNV2	0.886				0.917				0.882
	FNV3	0.854				0.878				0.851
Social value (SCV)			0.944	0.948	0.783		0.928	0.928	0.741		0.926	0.927	0.731
	SCV1	0.853				0.819				0.817
	SCV2	0.915				0.914				0.900
	SCV3	0.929				0.903				0.900
	SCV4	0.934				0.929				0.880
	SCV5	0.803				0.704				0.765
	SCV6	0.867				0.877				0.860
Emotional value (EMV)			0.919	0.932	0.674		0.875	0.900	0.575		0.889	0.889	0.601
	EMV1	0.857				0.816				0.801
	EMV2	0.841				0.824				0.799
	EMV3	0.780				0.661				0.770
	EMV4	0.680				0.638				0.705
	EMV5	0.876				0.822				0.830
	EMV6	0.884				0.852				0.785
	EMV7	0.811				0.743				0.729
Conditional value (CDV)			0.885	0.896	0.682		0.810	0.832	0.566		0.854	0.865	0.632
	CDV1	0.839				0.757				0.839
	CDV2:	0.802				0.757				0.840
	CDV3	0.825				0.767				0.796
	CDV4	0.808				0.659				0.684
	CDV5	0.855				0.813				0.807
Epistemic value (EPV)			0.837	0.845	0.676		0.764	0.812	0.584		0.826	0.832	0.660
	EPV1	0.865				0.758				0.782
	EPV2	0.883				0.787				0.874
	EPV3	0.829				0.794				0.854
	EPV4	0.699				0.716				0.730
Continued usage intentions (CUI)			0.885	0.888	0.743		0.848	0.851	0.689		0.888	0.892	0.749
	CUI1	0.822				0.766				0.822
	CU2I	0.882				0.890				0.913
	CUI3	0.852				0.818				0.840
	CUI4	0.891				0.841				0.883

Source: Created by authors

Table 4

Discriminant validity using heterotrait–monotrait (HTMT) ratio

Variable		CDV	CUI	EMV	EPV	FNV	SCV
Discriminant validity (overall model)
CDV
CUI		0.500
EMV		0.555	0.561
EPV		0.640	0.693	0.594
FNV		0.515	0.588	0.664	0.538
SCV		0.344	0.523	0.685	0.484	0.554
Discriminant validity (Generation X)
CDV
CUI		0.656
EMV		0.616	0.647
EPV		0.632	0.645	0.585
FNV		0.647	0.761	0.694	0.564
SCV		0.429	0.457	0.794	0.468	0.642
Discriminant validity (Generation Y)
CDV
CUI		0.445
EMV		0.464	0.531
EPV		0.598	0.747	0.565
FNV		0.443	0.503	0.646	0.562
SCV		0.398	0.571	0.693	0.531	0.555
Discriminant validity (Generation Z)
CDV
CUI		0.464
EMV		0.565	0.565
EPV		0.667	0.707	0.620
FNV		0.501	0.611	0.652	0.533
SCV		0.283	0.541	0.626	0.476	0.514

Source: Created by authors

Table 5

The results of hypothesis testing for overall model

Hypotheses	Path	Path coefficients (β)	t-statistic	p-value	VIF	Results
H1	FNV → CUI	0.201***	3.670	0.000	1.680	Supported
H2	SCV → CUI	0.154**	2.973	0.003	1.749	Supported
H3	EMV → CUI	0.072 ns.	1.130	0.259	2.184	Not supported
H4	CDV → CUI	0.075***	1.963	0.050	1.583	Supported
H5	EPV → CUI	0.355***	6.611	0.000	1.649	Supported

Notes:

***p-value ≤ 0.001, **p-value ≤ 0.01 and *p-value ≤ 0.05; ns. = not-significant

Source: Created by authors

Table 6

The results of invariance measurements testing using permutation

		Compositional (Step 2)		Partial measurement invariance	Equal mean assessment (Step 3)		Equal variance assessment (Step 3)
Constructs	Configurational invariance (Step 1)	Original correlation	5.0%		Original differences	Confidence interval	Original differences	Confidence interval	Full measurement invariance
Generation X vs Y
CDV	Yes	0.998	0.990	Yes	−0.112	[−0.214; 0.215]	0.531	[−0.537; 0.535]	Yes/yes
CUI	Yes	1.000	0.998	Yes	−0.040	[−0.218; 0.224]	0.276	[−0.369; 0.365]	Yes/yes
EMV	Yes	0.997	0.991	Yes	0.126	[−0.198; 0.207]	0.308	[−0.291; 0.280]	Yes/no
EPV	Yes	0.997	0.992	Yes	−0.024	[−0.204; 0.221]	0.450	[−0.352; 0.453]	Yes/yes
FNV	Yes	0.999	0.997	Yes	0.245	[−0.207; 0.208]	−0.150	[−0.155; 0.339]	Yes/yes
SCV	Yes	1.000	0.997	Yes	0.107	[−0.229; 0.215]	−0.020	[−0.281; 0.285]	Yes/yes
Generation X vs Z
CDV	Yes	0.995	0.993	Yes	−0.150	[−0.195; 0.200]	0.355	[−0.434; 0.432]	Yes/yes
CUI	Yes	0.999	0.998	Yes	0.096	[−0.206; 0.200]	−0.359	[−0.330; 0.336]	Yes/no
EMV	Yes	0.998	0.995	Yes	−0.116	[−0.211; 0.203]	0.227	[−0.321; 0.267]	Yes/yes
EPV	Yes	0.998	0.994	Yes	−0.100	[−0.203; 0.187]	−0.006	[−0.339; 0.324]	Yes/yes
FNV	Yes	0.999	0.997	Yes	−0.074	[−0.206; 0.196]	0.050	[−0.307; 0.300]	Yes/yes
SCV	Yes	1.000	0.998	Yes	0.014	[−0.199; 0.202]	−0.093	[−0.296; 0.253]	Yes/yes
Generation Y vs Z
CDV	Yes	0.993	0.989	Yes	−0.055	[−0.200; 0.190]	−0.151	[−0.292; 0.289]	Yes/yes
CUI	Yes	1.000	0.998	Yes	0.135	[−0.172; 0.187]	−0.635	[−0.336; 0.334]	Yes/no
EMV	Yes	0.994	0.991	Yes	−0.265	[−0.185; 0.188]	−0.065	[−0.269; 0.246]	No/yes
EPV	Yes	0.999	0.994	Yes	−0.084	[−0.194; 0.172]	−0.461	[−0.318; 0.324]	Yes/no
FNV	Yes	1.000	0.996	Yes	−0.322	[−0.379; 0.205]	0.195	[−0.317; 0.297]	Yes/yes
SCV	Yes	1.000	0.998	Yes	−0.088	[−0.178; 0.178]	−0.078	[−0.262; 0.234]	Yes/yes

Source: Created by authors

Table 7

The results of hypothesis testing for multigroup analysis

		Path coefficients (β)			Path coefficients differences			p-value differences Henseler’s MGA
Hypotheses	Path	Gen X	Gen Y	Gen Z	Gen X vs Gen Y	Gen X vs Gen Z	Gen Y vs Gen Z	Gen X vs Gen Y	Gen X vs Gen Z	Gen Y vsGen Z	Results
H6	FNV → CUI	0.410***	0.067	0.232**	0.344	0.178	−0.165	0.008**	0.167	0.025*	Supported
H7	SCV → CUI	−0.148	0.207**	0.185**	−0.355	−0.334	0.022	0.008**	0.012**	0.835	Supported
H8	EMV → CUI	0.270**	0.101	0.049	0.170	0.222	0.052	0.154	0.122	0.690	Not supported
H9	CDV → CUI	0.153*	0.051	0.033	0.102	0.120	0.018	0.049*	0.042*	0.826	Supported
H10	EPV → CUI	0.198	0.410***	0.387***	−0.212	−0.189	0.024	0.017*	0.038*	0.815	Supported

Notes:

***p-value ≤ 0.001; **p-value ≤ 0.01 and *p-value ≤ 0.05

Source: Created by authors