Enhancing Consumer Experience through Development of Implicit Attitudes Using Food Delivery Applications

: In the context created by the COVID-19 pandemic, the aim of the research in this article is to highlight how the consumer experience is constructed, starting from the consumer’s attitudes— which we consider to be implicitly positive—and from the formation of habits of using mobile applications for food supply. Data were collected from 610 people—users of applications for food delivery in urban areas—in the spring of 2021, using a quantitative type of research in the form of a ﬁeld survey based on questionnaire completion; for the model validation, we used conﬁrmatory factor analysis. The online questionnaire link was sent out in April 2021, followed by the consumers’ completion of the questionnaire in the same month. The proposed theoretical model highlights the inﬂuence of variables that are speciﬁc for technology acceptance models on the perceived utility and development of implicit attitudes based on it, as well as the habits of using the application. The results of the research validate the proposed model, showing that attitudes in the form of an implicit-type attitudes developed in time are inﬂuencing habits and aid the development of a certain usage behavior depicted by a consumption experience that has speciﬁc content. Our paper contributes to a better understanding of the consumer experience development process within the context of technology acceptance in the ﬁeld of food delivery services, emphasizing the importance of satisfying hedonic motivations as an integral part of the compensatory mechanisms determined by the effects of the pandemic.


Introduction
The effects of the COVID-19 pandemic crisis are visible in all human activities, both individually and collectively, with production, distribution and consumption changing significantly. One of the deeply affected areas is the HoReCa (hotel, restaurant and catering) industry. Complete closure during repeated emergencies and, subsequently, drastic limitation of activity in the case of classic restaurants, cafes, fast food units, etc., strongly affect buying and consuming behavior. Against this background, various mobile applications for food delivery have begun to develop and proliferate, with this new mode of distribution being considered by most consumers and authorities to be a viable solution in terms of hygiene and access to the user's home. For example, in the US alone, 41.7% of consumers used home delivery services during home isolation [1].
Food provision has become something beyond simple nutrient supply. Under the conditions imposed by the pandemic, food acquired the role of an emotional regulator, being able to talk about a way or even an emotional regulation strategy. Cooking in their personal space allows individuals to turn this into a special "ritual" capable of recovering emotions and experiences from the past [2]. Thus, food preparation becomes a way to overcome the negative emotions generated by isolation, as well as to prevent them within a family [2,3].
In less than two years after the outbreak of the pandemic, many papers that analyze this new trend in food consumption, the factors that influence consumption, the ways in which consumers adopt the new technology and the change in their consumption patterns have been published. The present research integrates various variables from well-known models regarding the type of technology adoption, such as TAM 1 [4] and TAM 2 [5], supplemented with UTAUT [6], but is not limited only to explaining the factors that contribute to the acceptance of technology or change in behavior regarding food consumption. Without limiting itself to explaining the factors that determine the adoption of mobile applications, our paper highlights the correlation between perceived utility and the formation of a positive attitude (an implicit type of attitude of the individual) and also presents the importance of a certain consumption experience.
Through their features, such as an intuitive menu, interactivity and interaction, restaurant evaluation possibilities and a short response time, mobile applications must offer consumers the opportunity to obtain long-term satisfaction. In our opinion, however, the conditions of the pandemic require special attention to be paid to the elements that meet the psychological needs of individuals-human contact after a period of isolation and the satisfaction of hedonic needs, seen as a mechanism to compensate for anxiety and frustration caused by restrictions and the constant fear of possible negative pandemic situations (loss of a loved one, danger of infection, lack of "normalcy", etc.). Loyalty to this type of application was considered an issue even before the pandemic, with Google reporting that 25% of mobile applications are never used by consumers and 26% of applications are abandoned after a single use [7]. With an appropriate approach to employee involvement, these services have every chance of becoming an optimal solution for customers. Statistics indicate that the majority of new users (57%) are likely to use them again [8]. In terms of options, as the market will be populated with increasingly differentiated offerings, variables such as perceived customer pleasure and social presence are expected to affect the attitude that has driven the behavior towards the delivery brand [9]. Consumer behavior can be considered from the perspective of a process embedding purchase, evaluation, use and disposal of products or services, a process that is influenced by endogenous or exogenous factors from different categories, such as social, economic, psychologic, contextual or personal factors. These factors affect consumer behavior on a regular basis, but there can be situations in which changes at the behavioral level are profound and at a high level. Such a situation presupposes the occurrence of extreme events: disasters or crises such as the one we are confronting right now. The COVID-19 pandemic is a complex phenomenon that has altered the mindsets of the people, not only a healthcare/biological crisis.
Thus, in addition to "classic" variables that define consumer behavior in a traditional way, such as product quality, satisfaction and learning, the situation caused by the pandemic puts particular pressure on consumers and producers/providers alike, with other different variables coming into light, variables capable of measuring factors that are specific to the new types of needs and consumption motivations. Among these, we can cite safety measures-elements found to be very important for retaining customers-and illicit loyaltytype behaviors [10], or variables which are part of a specific group of variables (those of a psychological nature)-fear, anxiety, stress, depression, self-justifications and personality traits [11]. Different authors try to grasp certain features and characteristics of consumers' behavior during the COVID-19 pandemic. From a general perspective, we can talk about panic buying behavior, correlated with impulse buying behavior, as previous studies have shown that dramatic events that are perceived by individuals as being a direct threat to their lives and safety can motivate compensatory responses able to alleviate fear and anxiety [11][12][13].
Additionally, a high level of stress induced by dramatic events can cause individuals to acquire a certain depressive emotional state that can be translated into impulsive buying behaviors or compulsive buying behaviors. Actually, all these dysfunctional buying behaviors are more or less a kind of self-protective strategy meant to manage depressive states and negative emotions through restoring the positive image of self and a sense of control [14].
In the context of the COVID-19 pandemic, consumers also tend to purchase larger than usual quantities of food, compulsive behavior that is also triggered by the high level of stress associated with the need to have a daily basket of products assured in the face of an uncertain future capability of buying those products [15]. During the pandemic, the personality traits of individuals become a source of change at the level of consumer behavior, with some of them influencing the predisposition to consume more unnecessary products, usually associated with hedonic-type needs and consumption motives [16,17].
Major players within the catering industry worldwide have to identify and adopt the best solutions that allow their businesses to develop and remain profitable. Within the context of the pandemic, orientation towards integrated solutions regarding the sales and delivery process optimization, such as online food delivery platforms, becomes a natural choice. The COVID-19 pandemic poses a huge challenge in terms of business models, supply chain management, emergency responses, staff training and incident management for any organization in the field of food processing [18].
In the face of this complex and prolonged challenge, small restaurants and food processing businesses are the most vulnerable, due to the limited amount of resources and sales capacity. In this category, online food delivery platforms are the key for survival and to meet the customers' requirements in terms of hygiene and delivery safety. Despite these net advantages that can be highlighted very clearly in the short term, after the initial phases of pandemic lockdowns, an interesting hypothesis is the fact that in the long run, in a future post pandemic stage, it is actually possible that the propensity of people to use online delivery platforms may decrease as traditional dining options will be revitalized. In the same direction, home cooking trends that have seen a particular development during emergency periods, will revert to the situation of before the pandemic once people return to a normal working program that implies again less time spent within their homes [19].
Still, the future holds a great amount of uncertainty as the current evolution of the pandemic still remains unpredictable for specialists, and the possibility of having emergency lockdown periods again is a cruel reality.
In the second part of this article, we will present the conceptual framework, which includes the conceptual model and the development of hypotheses, in the third part we will present the research model and methodology, and in the following sections we will present the empirical results, discussions and implications. At the end of the paper, the conclusions will be presented.
The originality of our approach consists of the different view that we bring to the table for the technology acceptance process, with five different factors that are considered antecedents for perceived utility (expected effort, facilitating conditions, social influence, hedonic motivation and expected performance) and the effect of perceived utility being manifested directly upon mobile application use habits, mediated through attitude. Within our model, we have employed original variables that measure the consumption experience of the users of food delivery applications, directly connected with utilitarian and hedonic aspects alike (quality of food, speed of delivery, standardization of delivery, comfort, staff interaction, positive emotions). The novelty and value of the model are closely related to the fact that the main findings are centered on the idea that attitudes, in the form of an implicittype attitude developed over time, are influencing habits and help the development of a certain usage behavior depicted by a consumption experience that has a specific content. Taking this approach into account, we stress the correlation between hedonic motivation, positive attitudes, habits and positive emotion, as these have to respond to the very specific consumer behavior within the COVID-19 pandemic.

Theoretical Framework and Hypothesis Development
The analysis of the level of adoption of a technology has the way in which the attitude towards technology can be measured and the intention to adopt it as reference elements. The framework of an appropriate analysis is given by the theory of motivated action (TRA) developed by Ajzen and Fishbein [20], which clarified the idea that a certain intention determines a certain behavior; meanwhile, the attitude can influence the intention and then the behavior. Subsequently, the theory of planned behavior (TPB) model proposed by Azjen [21] demonstrated that the degree of acceptance of technology is influenced by beliefs that have behavioral, normative and control dimensions, with the adoption process taking the form of a certain behavior. The TAM (technology acceptance model) mentioned above is derived from TRA and comes with a new set of variables to clarify the behavior corresponding to technology acceptance (perceived utility, perceived ease of use, attitude towards use, behavioral intention to use, etc.) [7,[22][23][24][25]. There are other approaches that analyze the adoption of technology that can be taken into consideration: task-technology fit (TTF) [26], the motivational and self-determination theory [27], the social cognitive theory (SCT) [28] and the technology-organization-environment framework (TOE) [29]. One of the last and most comprehensive theories is UTAUT (unified theory of acceptance and use of technology), an approach based on the combination of eight previous models of technology acceptance [30]. The UTAUT, and afterwards UTAUT-2, has been considered a more integrative model than TAM because of its greater predictive capacity [31,32].
Within the technology adoption process, one of the most important variables is referring to ease of use or expected effort. This can be defined as the degree of ease which can be associated with the usage itself of a technology system [29]. Generally, any kind of technology is perceived as useful if the individuals are capable of using it with ease, or the functioning itself is free from effort [33]. Mobile software applications that are easy to use are more susceptible to being adopted, especially by groups that are not technologically knowledgeable, such as older people [34]. The perceived ease of use can elicit positive emotions and a high degree of user satisfaction, with a tendency to be frequently used if the mobile application is less confusing and easy to use [35]. Both the TAM model and the TAM2 model, and later the UTAUT model, consider perceived ease of use or expected effort as a determinant for perceived utility, as evidence suggests that the easier it is to use a system, the more its use can increase performance in the workplace. The expected effort will be directly and indirectly linked with intention, because of its impact on perceived utility [4,5]. The perceived ease of use is related to the auto-efficacity of a person when it comes to computer use, taking into account the objective utility of a system for the person in place. The expected effort is conceptualized having complementary social influence and cognitive instrumental processes [5]. So, we can formulate the first hypothesis:

Hypothesis 1 (H1). Expected effort positively influences perceived utility.
Another variable that shapes the acceptance of technology is that of facilitating conditions. This variable refers to the situation in which users have all the resources and knowledge that allows them to actually use a certain technology [36,37]. There is a distinction between the individuals that are considering accepting and/or using a technology within an organizational environment and the ones that are using it outside such an environment. In an organizational context, it is possible that facilitating conditions (such as training and support) are actually available for free and invariably across employees. In such a situation, facilitating conditions directly influence behavior and determine a kind of predisposition for accepting technology. In different personal environments, users will interact with very different facilitating conditions because of different vendors, types of the technology, characteristics of mobile devices, etc. Within these contexts, facilitating conditions will more likely influence both intention and behavior [21,38]. As the consumer has access to more facilitating conditions, the intention to use technology becomes higher [39]. In a very simple example-the adoption of mobile internet, for instance-consumers can have access to different amounts of information and resources, such as different phone models, internet subscription plans, online tutorials, etc. In this case, facilitating conditions are related to behavioral intention and behavior [28].
Thus, taking account all of the above, we can issue the second hypothesis: Hypothesis 2 (H2). Facilitating conditions positively influence perceived utility.
When we talk about technology-related decisions, we may find social pressure around the consumer to be of a great importance. In various social contexts, the degree of acceptance for any kind of concept is given by the opinions and reviews of other persons that are important to the individual, elements that ultimately will influence the decision regarding the application acceptance [31]. In terms of mobile application usage, social influence represents a crucial factor related to consumers' behavior. Because of the motivations and interest aroused by this kind of application, consumers feel the need for social confirmation of their own decisions [40].
According to the TRA (theory of reasoned action), which stands as an important theoretical background for the TAM-1 model, social influences can be analyzed through the subjective norm concept as the perception of an individual referring to the fact that the majority of the other persons from their important personal group expect a certain behavior to be taken into consideration [20]. The same concept can also be found within the context of TBP (theory of planned behavior); the subjective norm is considered a factor that is capable of influencing intention because individuals may choose to adopt a specific behavior even if they do not have a favorable opinion about it, just because one or more important referrals to them approve of it.
The TAM-2 (technology acceptance model-2) takes into consideration another two elements, internalization and identification. Internalization refers to the process by which a user will incorporate in his own belief system the opinion of a referent regarding the use of a certain system [41].
In the case of internalization, subjective norm comes with an indirect effect on intention through perceived usefulness; the user perception about the utility can still increase as a response to the social persuasion [5].
Social influences also have to be considered in the context of maintaining a favorable image within a reference group. The use of innovation can enhance one's social status [42].
The TAM-2 theorizes that subjective norm will positively influence image because, if important members of a person's social group at work believe that he or she should perform a behavior (e.g., using a system), then performing it will tend to elevate his or her standing within the group. The effect of subjective norm on image, captured at the level of TAM-2, shows an increased effect on perceived usefulness as long as group norms continue to favor the usage of the technology. Therefore, taking into consideration all of the above, we can issue the following hypothesis: Hypothesis 3 (H3). Social norms, referring to social influence, positively influence perceived utility.
Another variable that is essential to ascertain the use of technology refers to hedonic motivation. In a general approach, hedonic motivation can be defined as the fun or enjoyment that is derived from using a technology, being associated with inner beliefs of the user regarding a kind of immediate gratification. Reasons to consider hedonic motivation among the variables that are determinant in the first place for the acceptance of technology processes can be found within the scientific literature concerning consumer behavior and acceptance technology models alike [8,9,30,[43][44][45].
In addition, we consider it opportune to take into consideration hedonic motivation as a variable that can directly influence perceived utility and further the acceptance of technology because of the specific conditions developed within the COVID-19 pandemic.
From this point of view, there are specific mechanisms that can be used by individuals to overcome anxiety, panic and depression caused by the isolation measures, mass-mediaexaggerated messages and family members or friends already infected and affected by the disease. Among these psychological mechanisms, we may find impulsive consumption and/or hedonic-type consumption behavior [46,47].
In the case of food product delivery applications, the convenience of consuming preferred food in a person's own intimacy, having assured all the safety and hygienic conditions already, can be an important element for hedonic-type motivation. Ordering online and consuming food products from a whole variety of offers has all the advantages to turn into a consumption experience full of many hedonic values. Therefore, we can highlight the following hypothesis: Hypothesis 4 (H4). Hedonic motivation also positively influences perceived utility.
The expected performance can be defined as the degree to which any system would increase user productivity or contribute to performance gains in the workplace [30]. As time goes on, the consumer becomes more and more attracted by the application and more confident in their capacity of using it, therefore capable of obtaining a superior productivity [48,49].
In different research contexts, studies have confirmed the fact that expected performance has a direct effect on users' satisfaction when the considered application is used on a continuous basis [50,51]. Male consumers have correlated perceived expected performance with the effort to attain their objectives into the context of continuous usage and adoption [52]. As a consequence, we can issue another hypothesis according to all of the above:

Hypothesis 5 (H5). Expected performance positively influences perceived utility.
The perceived utility of technology is determined mainly with the help of cognitive processes and inferences able to determine the degree of achievement of the objectives through the applications' usage. Combining elements from work motivation theory, action theory from social psychology and task-contingent decision making from behavioral decision theory, the TAM model proposes a framework that assesses the mental representation of the perceived technology utility and instrumental behavior regarding the use of a certain information technology system. In the case of food product delivery applications, users will analyze the utility of such applications in the context of consumer behavior usage habits. Habits will comprise the automating behavior, from the initial learning process to regular use of the technology [53]. For such applications, habits that can be associated with application consumption refer to different types of habits. The main cues are internal ones derived from the actualized needs of individuals-a regular need for food consumption. At the same time, the interesting situation is related to a different category of habits-"checking habits", triggered by different cues from the device itself, such as different situations or emotional states, or by the application itself if it has "the reminder" functionality [54]. Although in the UTAUT model, the habit was considered as a variable that was part of the group of antecedents of the perceived utility, in the case of food delivery applications, we consider the situation in which the perceived utility will influence the birth and development of the mobile application use habit to be more appropriate.
These arguments can be used to formulate the following hypothesis of the proposed research model: Hypothesis 6 (H6). Perceived utility positively influences mobile application use habits.
Implicit attitudes have the quality of being especially evident in situations that involve routine and a high degree of familiarity with technology. The implicit attitude is formed over time, requiring a series of repeated interactions with technology to shape an attitude model. By contrast, explicit attitudes show their influence, especially when users engage in the planned use of a system, and especially in cases where individuals interact with new systems [55].
This stems from the fact that the cognitive processing quantity needed to develop an attitude depends on the degree of familiarity of the user with a certain system. When individuals interact with new systems, they are motivated to engage deliberately in cognitive processing in order to develop an attitude towards the system [56].
Every individual will carefully weigh the advantages and disadvantages of the system and make a careful and planned decision. When an interaction is made with already familiar systems, it is possible that people do not make a conscious effort to cognitively process the system characteristics and evaluate their actions; in exchange, they will try to minimize their cognitive task.
As a result, instead of engaging themselves in deliberated thinking in order to develop or reflect upon an explicit attitude, users can instantaneously invoke the implicit attitude that is capable of triggering the system usage and evades the rational decision-making processes [57].
In conclusion, people's natural inclination to save mental effort can lead them to rely more on implicit attitudes when they usually use familiar computer systems, while the explicit attitude will be obvious when it comes to planned use for a new system [58].
This situation imposes the following hypothesis to be issued: Hypothesis 7 (H7). Perceived utility or usefulness of the application positively influences the positive attitude, seen as a dimension related to implicit attitude.
Habit has been defined as the extent to which people tend to perform behaviors automatically because of a previous learning process, behavior that can be equated with automaticity [59].
Referring to internet mobile applications, habit and intention exert a direct effect on actual behavior. Actually, habit and intention can act in tandem as predictors for actual food consumption. Consumption experiences give birth to a knowledge continuum and can form the basis of a behavior that expresses habit. Experience is necessary, but not a sufficient condition for the formation of a habit. In time, individuals can form different levels of habit depending on the extent of familiarity that is developed with the specific technology. Feedback gained over time from previous experiences will determine the manifestation of different beliefs and, because of that, different future behavior. So, a habit may have the qualities of a perceptual construct that is capable of reflecting the outcome of previous experiences. Previous experiences can be highlighted by positive implicit attitudes that have a stable form; so, acquiring a positive attitude becomes a precondition for the manifestation of habits seen as automatic and intuitive continued usage behaviors based on the frequency of prior behavior, comprehensiveness of usage and repeated satisfaction [60]. The connection between the two variables permits us to issue the following hypothesis:

Hypothesis 8 (H8). Positive attitude influences mobile application use habits.
The content of habits represents much more than satisfaction. Both intentions and habits are rooted in satisfaction. Thus, personal habits will influence continuous usage of technology, having at the same time an interactive effect that can be measured by the changing of intention in usage behavior. In terms of mobile internet applications, prior experience with the technology is a significant factor, with behavioral intention being strongly related to actual use of the technology in comparison with inexperienced users [61].
Internal and external cues that influence the usage of technology can be linked through a positive conceptualization of the individual regarding the main characteristics of their own habits [62].
In order to develop a habit, there are three main elements that have to come together: repetition of a behavior, automaticity of the specific repeated behavior and different contextual cues [63,64]. Habits will be acquired through an incremental strengthening of the association between a situation (cue) and a behavior. At the same time the automaticity of the behavior is a key characteristic of habitual behaviors. The automaticity is developed when a particular behavior is repeated in a stable context [62]. Therefore, developing a habit needs the continuous intention of using the technology and for the consumer to manifest the automaticity of his actions in the process. In light of these considerations, we can highlight the following hypothesis: Hypothesis 9 (H9). Application use habit positively influences usage intention.
The direct connection between usage intention and usage behavior is postulated within social psychology theories, such as the theory of reasoned action (TRA) and theory of planned behavior (TPB). Thus, according to them, usage intention presupposes the manifestation of the motivational factors which are the defining ones for an individual to perform a given action in a particular situation. Actually, in the TRA setting, it has been stated that intention has the power to predict a specific behavior suggested by the intention to do so. Within the IT application-specific field, this correlation has already been highlighted regarding usage behavior of the internet stock trading system [65].
Additionally, other research has found that intention is clearly correlated with technology usage behavior .
Taking the above into consideration, we can issue the following hypothesis:

Hypothesis 10 (H10). Usage intention positively influences usage behavior.
The consumption experience within the field of food products implies a complex, multidimensional process of self-evaluation of individuals' consumption acts. Moreover, the complexity of the consumption experience today is given by changes and transformations that take place regarding collaborative consumption, sharing economy, evolution of social and digital media, and even the need for a sustainable approach of consumption in general [68][69][70].
The consumption experience scientific literature considers the work of Holbrook and Hirschman that, in 1982, proposed a new model of consumption behavior that emphasized the experiential dimension of the behavior [71]. From that moment until the present, the body of literature concerning consumption experience has grown exponentially, with different authors trying to integrate different views and ideas into a single model. We can cite the works of Cohen and Areni [72]-the dynamic model of affective response, Holbrook [73]-the consciousness emotion value model and Kwortnik and Ross [74]-the experiential decision model. The integrative models have a common view characterized by three different phases of the models: the first phase is comprising the antecedents of experience (individual-or environment-related factors that drive consumption); the second phase refers to the content of the experience itself; the third phase deals with the consequences of experience at the individual and inter-individual levels [69].
The relationship between consumer behavior and consumption experience starts from the idea that a consumer passes through a variety of stages, during which he gains experience on consumption and share it with his peers. A synthetic view of modern theories on consumer behavior suggest that emotion can be at the same time a cause and an effect for information processing and experiences of the consumers. Consumer behavior is formed in time through the interaction of different internal and external factors [75]. During this interaction, perceptions and the learning process develop in attitudes that become stable over time. The pattern of consumption can be assessed through different consumption experiences, during which consumers are reacting in line with the pattern of consumption already in place. The different dimensions of consumption experience-cognitive, affective or social-are integrating within the behavior and are evolving along with the consumer [76].
In the following section, we will highlight certain determinants for the consumption experience in the case of food delivery applications and we will point out the corresponding hypothesis regarding the relationship with usage behavior. In the field of food products, the cue utilization theory can be used in order to ascertain different levels of quality. Thus, different types of cues can be used, such as intrinsic cues-taste, flavor, color, shape, structure and composition-and extrinsic cues that are not related to the physical composition of the products, such as brand, nutritional information, country of origin, price, etc. Another view regarding the perception of food product quality as an attribute of consumption experience starts from the perspective offered by S-O-R (stimulus, organism, response) theory, as a stimulus can trigger an internal evaluation process of the organism that will influence the response [77]. Perception of the quality of the food products is essential for the consumption experience assessment. As an attribute that can be related to satisfaction and an important source of trust, perception about food product quality is an important attribute of the consumption experience [78].
Taking the above into account, we can state the following hypothesis: Hypothesis 11 (H11). Usage behavior positively influences consumption experience from the point of view of perception of ordered food product quality.
Another attribute of consumption experience relevant in the context of food product delivery applications refers to the perception of the delivery speed. Different studies highlight the fact that delivery represents a very important element of the logistics of online services [79].
Purchase decisions and consumer loyalty are directly influenced by shipping fees and delivery quality [80]. Within the online environment, online consumers are sensitive regarding the delivery time, any delays decrease the propensity of consumers to become loyal. The delivery time has a direct impact over the consumer behavior from the point of view of satisfaction levels and repurchase intentions [81,82]. Different authors studied different consumption contexts and confirmed the idea that online consumers are especially sensitive to on-time delivery, considering it to be a special factor that can influence their satisfaction in the long run [83][84][85].
In the case of hedonic products, a shorter delivery time (under 24 h) can increase the level of satisfaction of consumers, as they are motivated by the impulse-buying process associated with immediate gratification through hedonic product consumption [86].
A study conducted on UK consumers in the case of online grocery products shows that from the point of view of decision criteria, the delivery speed and on-time delivery are among the most important criteria [87]. Additionally, from the perspective of value drivers that are motivating consumers to choose a specific online provider, total price was followed by delivery speed and delivery method [88].
With reference to the online buying of specialty foods, a short delivery time contributes to the quality of the received food, which is one of the most valuable attributes for consumers [89].
Taking the above into account, we can issue the following hypothesis: Hypothesis 12 (H12). Usage behavior positively influences consumption experience from the point of view of speed of delivery perception.
The consumption experience is built over time through the involvement of consumers in multiple consumption acts as a part of their consumer behavior. In the case of the Internet and mobile-related applications, the consumption experience implies a certain degree of standardization as the interface and the modalities that are used to interact with applications are built, taking specific technical features into consideration. The proper implementation of a service presupposes the capacity to adapt the service provision to particular demands that can be expressed by the consumers in specific settings, at the same time still maintaining the quality standards required by the assumed positioning of the brand. Standardization can be considered a voluntary process made to develop a series of specifications that are common for the companies and their stakeholders; meanwhile, innovation is considering new applications for knowledge, methods and skills capable of generating enhancements to products and services in the same direction as the market and customer needs evolution [90]. Provision of the food product delivery application services is a mixture of innovation and standardization. Despite the fact that food products are mainly hedonic-type products, the customers are very sensitive about the reliability promised, so they need at the same time a kind of assurance regarding the process of food provision and delivery. Some aspects regarding the list of restaurants used by the application and the menu that is available can be tailored taking into account the consumer needs; meanwhile, other aspects, such as delivery time and the process of delivery itself, have to be standardized, in order to offer to the customers enough reliability. When the customer focus on efficiency and functionality, the organization should provide standardized services; when the same customer is focused on fun and novelty, the organization will provide standardized services [91].
The consumption system that is required to consume online and mobile food product delivery services comprises a bundle of information, processes and goods with which the consumer comes into sequential contact during several stages. We can distinguish three basic elements within such a system [92]: attribute-level evaluations, in-process satisfaction and behavioral intentions. In online markets we can see that is a certain spatial and temporal separation between buyers and sellers; therefore, exchanges of money and goods are not performed simultaneously. This means that for some of the customers, the risk associated with delivery is very present. The reliability of such a service is based on delivery time fulfillment, the capacity to deliver the food product as promised and the consistency of customer service (order tracking, on-time delivery, customer support and product expectations being met). The consumer behavior will be built through a specific consumption experience from the point of view of standardization of the delivery process in order to maintain and emphasize the positive perception of the customers. Taking all of the above into account, we can issue the following hypothesis:

Hypothesis 13 (H13). Usage behavior positively influences consumption experience from the point of view of delivery standardization perception.
Within the food product delivery applications setting, consumption experience can also be interpreted from the point of view of the user's level of comfort. People tend to retain pleasant experiences that can elicit positive emotions. Hedonic-type consumption is essential in order to understand and assess the consumption experience related to food products. As the very meaning of food delivery applications is built around the idea that the applications have to simplify access to food products, save time and contribute to the wellbeing of the individual, the consumption experience around them falls into the category of a hedonic one. In addition to these aspects, one can also find some rational elements that are important for customers from the point of view of the functionality of the application and the ergonomics of it, elements that can also contribute to the proper level of comfort. The consumption of food itself represents another source of comfort, at least at the individual level, and specialists have found that as the consumers' age increased, "comfort food" is associated with greater positive emotions prior to consumption [93].
In this context, in the case of the food product delivery applications, we can issue the following hypothesis:

Hypothesis 14 (H14). Usage behavior positively influences consumption experience from the point of view of comfort.
The nature of customer interaction experiences in virtual environments is as critical as their offline experience related to a product. Studies may suggest that beliefs concerning potential benefits will impact future customer participation in product support, while affective reactions will primarily influence consumer attitudes regarding the provider of the service. The benefits that customers can expect to obtain are cognitive (related to information acquisition and are strengthening the understanding of the environment), social integrative (are strengthening consumer ties with other relevant consumers), personal integrative (are strengthening the credibility, status and confidence of the individual) and hedonic (are strengthening aesthetic or pleasurable experiences) [94]. At the same time, aside from these potential benefits, customers' interactions with the applications might generate immediate positive and negative feelings, changes that may shape their attitude towards the provider. Consumer experience related to food consumption can be associated with two different types of variety-seeking behaviors: the first refers to the routine rotation of various ingredients across time, so it can be called routine variety seeking, the second can be called novelty seeking [95].
Regardless of the type of seeking behavior employed, customer service and the contact with home delivery staff has a particular significance because of the human relationship that can be developed. In particular, during periods such as the present that imply a general crisis, with the anxiety and pressure of social isolation, contact with employees gains a particular significance as a modality to overcome all these negative conditions. In such a context, we can formulate the following hypothesis:

Hypothesis 15 (H15). Usage behavior positively influences consumption experience through interaction with customer service/home delivery staff.
A loyal-type behavior can be developed if a certain provider builds emotional connections in addition to positive attitudes and behaviors. The vast majority of consumers that feel emotionally bound to a certain provider actually have memorable experiences with that company [96].
Emotions have been found to be very important in the process of performance appraisals. In order to create the conditions necessary for a loyal customer, the company has to deliver a special delight state for them [97,98]. In the case of the food product delivery applications, the consumption experience associated with these applications is mainly a hedonic-type one correlated with the consumption of the food itself. Users will search for positive emotion or gratification, opposing these to the pressure of anxiety and social isolation caused by the pandemic crisis. So, taking into account the practical conditions, we can issue the last hypothesis:

Hypothesis 16 (H16). Usage behavior positively influences consumption experience through positive emotion/gratification.
Taking account of all the above hypotheses, we can propose the following conceptual model of the research, as can be seen in Figure 1 below.
The advanced hypotheses and the related constructs are summarized in the form of the table that can be found in Appendix A. The advanced hypotheses and the related constructs are summarized in the form of the table that can be found in Appendix A

Sampling and Data Collection
In order to characterize the link between the consumption experience variables and the ones corresponding to perceived usefulness, implicit positive attitude, usage habit, intention to use and behavior of use, we conducted a study in the spring of 2021 among students and graduates from three universities: a private one in the capital, and two state universities in Romania's development region "Center".
We considered it appropriate to choose a statistical population from which to extract the sample, consisting of students and college graduates based on considerations found in the literature that specifically addresses the topic of our research-the consumption of online food distribution services. Thus, it is highlighted that, in different research contexts, the population used for sampling in the case of online food delivery services is dominated mainly by young consumers, with large numbers of students that associate a high frequency of online food delivery consumption with medical issues like overweight and obesity [99][100][101][102]. Other studies show that from a demographic point of view, young couples are more inclined towards the adoption of online food delivery services in their quest to fulfill the need for convenience due to their dynamic program [103][104][105]. The propensity to share food ordered with the help of applications designed for online distribution was found in the case of close family members or work colleagues and students who live together in dormitories [106]. At the same time, other studies conducted during the COVID-19 pandemic have shown that people with higher education levels tend to buy larger quantities of goods than people who have a lower level of education [11,107]. Young people are individuals with the propensity to use digital applications more and post images of food items on their social networks, a tendency that is amplified in the context of online food delivery application usage [108]. Another interesting piece of information that comes from previous research is that consumers' intentions to use OFDSs (online food delivery services) are associated with their perceived self-image [109], with young people, especially students, being predisposed to give more importance to this than other segments of the population.
The questionnaires were sent over email to a total of 2500 students and graduates that were asked to help by sending the questionnaire to relatives and friends. The number of returned questionnaires was 625, of which the number of valid, fully completed

Sampling and Data Collection
In order to characterize the link between the consumption experience variables and the ones corresponding to perceived usefulness, implicit positive attitude, usage habit, intention to use and behavior of use, we conducted a study in the spring of 2021 among students and graduates from three universities: a private one in the capital, and two state universities in Romania's development region "Center".
We considered it appropriate to choose a statistical population from which to extract the sample, consisting of students and college graduates based on considerations found in the literature that specifically addresses the topic of our research-the consumption of online food distribution services. Thus, it is highlighted that, in different research contexts, the population used for sampling in the case of online food delivery services is dominated mainly by young consumers, with large numbers of students that associate a high frequency of online food delivery consumption with medical issues like overweight and obesity [99][100][101][102]. Other studies show that from a demographic point of view, young couples are more inclined towards the adoption of online food delivery services in their quest to fulfill the need for convenience due to their dynamic program [103][104][105]. The propensity to share food ordered with the help of applications designed for online distribution was found in the case of close family members or work colleagues and students who live together in dormitories [106]. At the same time, other studies conducted during the COVID-19 pandemic have shown that people with higher education levels tend to buy larger quantities of goods than people who have a lower level of education [11,107]. Young people are individuals with the propensity to use digital applications more and post images of food items on their social networks, a tendency that is amplified in the context of online food delivery application usage [108]. Another interesting piece of information that comes from previous research is that consumers' intentions to use OFDSs (online food delivery services) are associated with their perceived self-image [109], with young people, especially students, being predisposed to give more importance to this than other segments of the population.
The questionnaires were sent over email to a total of 2500 students and graduates that were asked to help by sending the questionnaire to relatives and friends. The number of returned questionnaires was 625, of which the number of valid, fully completed questionnaires was 610. The structure of the sample can be found within Table 1, presented below. Taking into account the fact that the sample was extracted from a statistical population comprising only students and graduates, it can be observed that we did not take into consideration entire age range usually associated with generation Z.

Measures
Responses to the questionnaire were measured with a seven-point Likert scale. In order to validate the measurement instruments, we have first tested the viability of the model by computing Cronbach alpha and composite reliability (CR). Both have values above the threshold of 0.7 for each of the constructs [111][112][113][114][115][116][117]. These results can be found in Table 2. As regarding the use of the Kaiser-Meyer-Olkin (KMO) test in order to measure the suitability of the sample, it must have a minimum value of 0.5 for the sample size to be considered suitable to perform the factor analysis [118][119][120].
A confirmatory factor analysis was then performed in order to assess the relevance of the relationships that exist between the model variables using the IBM-SPSS AMOS 24.0 program.
As can be seen in Table 3, where the values obtained from the research are presented as well as the appropriate theoretical statistical values for a goodness of fit, indices of the structural model presented good values for the variables of the model.

Results
Test results can be declared eligible by the criteria in the previous Table 3, namely the criteria of P, GFI, AGFI, PGFI, NFI, RFI, IFI, TLI, CFI, PNFI, PCFI, RMSEA and PCLOSE so that all hypotheses proposed in this study can be explained. Table 4 shows the structural model results demonstrating the validity of the hypotheses put forward.
According to Table 4, it can be seen that the hypotheses advanced in the proposed conceptual model have been validated. Thereby, variables that are antecedents for perceived utility (EE-expected effort, FC-facilitating conditions, SI-social influence, HMhedonic motivation, EP-expected performance) have a positive and significant effect on PU-perceived utility. It can be observed that facilitating conditions have the greatest effect on perceived utility (β = 5.319; p < 0.01; critical ratio test = 8.663 > 1.96); therefore, H2 is validated. Regarding expected effort, this in turn strongly and significantly influences perceived utility (β = 4.168; p < 0.01; critical ratio test = 8.472 > 1.96), which means that H1 is also validated. Social influence has a strong and significant influence over perceived utility (β = 0.241; p < 0.01; critical ratio test = 11.476 > 1.96), meaning hypothesis 3 is validated.
Hedonic motivation shows the same type of influence over perceived utility with values of (β = 2.749; p < 0.01; critical ratio test = 4.874 > 1.96); thus, hypothesis H4 from the proposed model is verified. The last variable from the initial group of antecedents for perceived utility, expected performance, also has a positive and significant effect on perceived utility (β = 1.699; p < 0.01; critical ratio test = 4.144 > 1.96) confirming hypothesis H5. The following hypotheses from the model, H6 and H7, refer to the influence exerted by the perceived utility over mobile application use habit and positive attitude. In the case of each relationship, we can see values that are showing the positive and significant effect of perceived utility over the two variables considered, (β = 5.528; p < 0.01; critical ratio test = 8.263 > 1.96) in the case of influence over mobile application use habit and (β = 6.724; p < 0.01; critical ratio test = 7.764 > 1.96) in the case of influence over positive attitude.
Positive attitude in turn has a strong and positive effect on mobile application use habit (β = 4.168; p < 0.01; critical ratio test = 8.472 > 1.96), meaning hypothesis H8 is validated. Mobile application use habit (β = 1.011; p < 0.01; critical ratio test = 9.191 >1.96) has a positive and significant effect on usage intention, meaning hypothesis H9 is validated. Usage intention has the same type of relationship with usage behavior (β = 0.089; p < 0.01; critical ratio test = 5.235 > 1.96), confirming hypothesis H10. The final part of the model comprises the relationship between usage behavior and variables that are describing the consumption experience of the users in the case of food product delivery applications: perception of ordered food product quality, perception of speed delivery, perception of delivery degree of standardization, perception of comfort, perception of customer service interaction and positive emotion. Thereby, usage behavior has a positive and significant effect on perception of ordered food products quality (β = 1.689; p < 0.01; critical ratio test = 4.120 > 1.96), meaning hypothesis H11 is confirmed. Usage behavior will also have a positive and significant effect on perception of speed delivery (β= 1.015; p < 0.01; critical ratio test = 126.875 > 1.96), confirming hypothesis H12. Usage behavior has a positive and significant effect on perception of delivery standardization (β = 1.013; p < 0.01; critical ratio test = 112.556 > 1.96), confirming hypothesis H13. The same positive and significant effect was confirmed in the case of usage behavior upon perception related with comfort, seen as a consumption experience element (β = 0.973; p < 0.01; critical ratio test = 64.867 > 1.96), which confirms hypothesis H14. Usage behavior has a positive and important relationship confirmed with perception of interaction with customer service or home delivery staff (β = 0.861; p < 0.01; critical ratio test = 43.050 > 1.96), which corresponds with and confirms hypothesis H15. Finally, usage behavior also shows a positive and significant effect on positive emotion, seen as a consumption experience element (β = 1.687; p < 0.01; critical ratio test = 4.115 > 1.96), confirming the last hypothesis of the model, H16.

Discussion over the Main Results
Finding ways to develop a sustainable supply chain within the pandemic consumption and production framework resides in bringing together various resources at the level of both producers and consumers alike [127,128]. The fast development of technology and, at the same time, the speed of diffusion among users are positive factors that can contribute to the sustainability of supply value chains, especially within fields such as food product delivery.
In order to assure more objectivity in our endeavor, we have to specify once more that our proposed model intends to explain and capture the relationship and effect of different variables considered suitable to explain technology acceptance and usage within a very specific context-that of online and mobile applications for food product delivery. All the considered hypotheses and the results obtained must be interpreted from this perspective.
Our proposed research model begins with the relationship between five variables that are also considered in other contexts (four variables contained by both UTAUT 1 and UTAUT 2 models: expected effort, facilitating conditions, social influence and expected performance, and one variable contained only in the UTAUT 2 model: hedonic motivation) as being antecedents for the perceived utility of the technology. From all the five variables, the one that refers to facilitating conditions has stood out with the greatest influence upon perceived utility in the case of our model. This means that in the case of food product delivery applications, users consider facilitating conditions an important issue and determinants for the effective use of the application.
All the other variables have a significant and positive effect on perceived utility, having strong explanatory power as a group of variables, interconnected between them. Modern consumers have the willingness to test and adopt technology, the above variables depicting a group of factors that are essential for technology adoption behavior.
The first hypothesis ascertains the fact that perceived ease of use positively influences perceived utility. The perceived ease of use has been defined within our research with the help of four items referring to food delivery application easiness (the application is clear, understandable, does not require a lot of the user's mental effort, etc.).
A particular approach has been used within the context of our research regarding the influence of perceived utility over mobile food delivery application use habits, and later in the proposed model, the influence of positive attitudes on the same mobile food delivery application use habits. In the latest version of the UTAUT-2 model, the use habit is incorporated in the first group of variables that are determinants for the behavioral intention (performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation, price value, habit), and is also considered to have a direct influence over use behavior [30].
We consider that, in the case of food product delivery applications, use habit is more suited to be considered after the effect of perceived utility is visible, the explanatory power of its variable being employed simultaneously with the formation of positive attitude toward the usage of the applications. Food product delivery applications are part of a category of applications developed both for use in the online environment and for the mobile phone, which have been reported as applications to which users have struggled to develop a loyalty-type behavior. Specifically, the situation of food product delivery applications may be slightly different than the others because of their potential to be connected with hedonic motivations. In this context, the importance of habit in order to have usage intention and later usage behavior becomes essential in order to develop a behavior oriented toward a loyal-type state of mind for the consumers. The data obtained regarding usage habits show that 78% of respondents between 22-39 years (millennials (Gen-Y)) and 86% of men use online food delivery applications on a regular basis.
Validation of hypotheses H11-H16 represents a confirmation of a particular consumption pattern that can have as specific important elements: food product quality, speed of delivery, delivery process standardization, consumption comfort, human-staff interactions (customer service and delivery staff) and positive emotions (pleasant experience, joy, selfgratification state of mind). All these elements lead to making consumption experience content in the case of food product delivery applications, which implies a certain replication of the behavior, meaning thus the development of a loyal-type consumer behavior.

Theoretical Implications
The results of this paper validate the application of variables specific for the TAM model and the UTAUT-2 model within the food product delivery application consumption context. Our proposed model offers another approach for the technology acceptance process, in which five factors are considered antecedents for perceived utility (expected effort, facilitating conditions, social influence, hedonic motivation and expected performance), and the effect of perceived utility is manifested directly upon mobile application use habits and also mediated through attitude. Additionally, the model proposes variables that measure consumption experience of food delivery application users, directly connected with utilitarian and hedonic aspects alike (quality of the food, speed of delivery, standardization of delivery, comfort, staff interaction, positive emotion).
The originality and value of the model stem from the main findings that are centered on the idea that attitude in the form of an implicit-type attitude developed in time influences habit and helps the development of a certain usage behavior depicted by a consumption experience that has a specific content. The model presents the correlation between hedonic motivation, positive attitude, habit and positive emotion as a red wire, as it has to respond to the very specific consumer behavior within the COVID-19 pandemic.

Managerial and Policy Implications in the Field
Food product delivery services have the potential to be true hedonic-type services that elicit a behavior that can compensate the negative effects of the pandemic (strong anxiety, depression, perpetual uncertainty regarding his own, etc.). Thus, the managerial implications can be considered in the first place, the ones referring to the need of food delivery companies' managers to become aware of the importance of human interactions in the context of pandemic. This is a field of "almost infinite" possibilities in order to differentiate from competition and to fulfill the need of warm communication with the customers. Another clear managerial implication of our research results are the ones referring to the hedonic motivation stated above: managers have to take into consideration the development of the applications in the line of increasing the functionality from the perspective of satisfying this type of motivation-the possibility for the customers to have a personal profile with predefined orders or menu selections, preferential discounts from the favorites restaurants, etc.
The future evolution of the pandemic will influence policy decisions in the field. Our study emphasizes the fact that consumers, especially young people, are developing a consumer behavior that is characterized by the avoidance of anxiety, stress and depression caused by the pandemic by using new technologies that can allow a proper consumption experience based on safety being free of infection risk. Policy makers in the field should pay further attention to the legal framework that can help companies to develop the safest delivery circuits and sustainable activity during eventual future lockdowns imposed by the negative evolution of the pandemic.

Research Limitations and Future Directions
The research has certain limitations, such as the one due to the type of the research employed-a cross-type quantitative research, as from the point of view of broadly evaluated consumption experience, longitudinal repeated research is much more appropriatecapable of offering a better view of the concept of consumption experience. Another limit is given by the structure and size of the sample-consumers only from the urban areas in Romania, as well as limits specific to email-based empirical research that we can highlight below. The risks associated with a questionnaire administered over email is the reluctance of some individuals that they will have their identity disclosed. Another limit may be the fact that respondents could send the questionnaire to be answered in their place, or to send to other persons, and in this case the researchers have serious difficulties in assessing the real response rate, which in general is low. Another issue, this being a self-administered questionnaire, is the tendency of certain categories to be more receptive in answering to the questionnaire, a fact that can affect the representativeness level of the sample.
Future research can try to validate the proposed model within other consumption contexts, on different scales using a combined qualitative type and quantitative type research method. Due to the importance of the topic, we consider that such efforts are welcomed as these types of services are representative of the consumer behavior evolution which has been deeply affected on multiple layers by the pandemic, thus being a field of research that can bring a lot of new ideas and testing possibilities useful for the scientific community.

Consumption experience-CE
Usage behavior positively influences consumption experience from the point of view of perception of ordered food product quality.