Robot acceptability in quick service restaurants: a customer’s point of view

Many sectors in the service industry increasingly turn to intelligent automation to improve service consistency, save costs, and bolster their service offerings in light of labor shortages. This trend began before the COVID-19 pandemic, but the pandemic accelerated efforts to automate (Brengman & Willems, 2023). The pandemic hit the hospitality industry particularly hard. As the hospitality industry recovers, the quick-service restaurant sector is emerging as an innovator in service robot implementation. Popular quick-service restaurant (QSR) chains, such as McDonald’s, Chipotle, and Chick-fil-A, have launched robotic devices to perform various tasks, including cooking, taking orders, and serving food and beverages (Halpern, 2024). Robot service undoubtedly creates a new customer experience (Seo & Lee, 2021), leading to two questions –will customers accept service robots? And what influences robot acceptance by restaurant customers?

Existing research unveiled stakeholders’ mixed feelings about service robots. Interviews with business managers suggest that managers see both opportunities and challenges in adopting service robots on the front line (Meyer, Roth, & Gutknecht, 2023). Similarly, diners at QSRs report that they are uncertain about the impact of robotics on the workforce and are concerned about the absence of human interactions in a restaurant setting (Zemke, Tang, Raab, & Kim, 2020).

Most previous research on service robots applied the Technology Acceptance Model (TAM), using the original constructs -- perceived usefulness, perceived ease of use, and acceptance. These studies found that they positively impact customer revisit intentions in a robot service setting. Recently, however, Seo and Lee (2021) tested this model and found no impact of “perceived ease of use” on consumers’ intentions to revisit a restaurant that incorporates robotics. These contradictions, along with a lack of new, innovative factors that explore the influences on robot acceptance provide fertile ground for further investigation. Previous researchers (Savela, Turja, & Oksanen, 2017; Zemke et al., 2020; Wan, Chan, & Luo, 2021; Guan, Gong, Li, & Huan, 2021) all suggested future research on the topic of robot acceptability.

A few studies have investigated the factors impacting consumers’ behaviors and attitudes toward robots in restaurants. These studies mainly explored the benefits of using robots, such as reduced labor and training costs (Kuo, Chen, & Tseng, 2017; Zemke et al., 2020). Lu, Zhang, and Zhang (2021) conducted a comprehensive literature review of the service industry’s deployment of service robots. They concluded that most of the existing research is of a conceptual or experimental nature through a scenario-based research design. Therefore, there is a need for research on experience-based customer perceptions and motivations. This study addresses this gap and investigates which factors impact QSR customers’ intentions to accept service robots in QSRs. The robots referred to in this study are non-anthropomorphic machines that can substitute for human employees to perform cooking tasks in the kitchen or serving tasks in the front of the house.

To bridge the gap, the current study extends the body of literature on the application of robotics in the restaurant industry. First, this study pioneers the investigation of QSR customers’ attitudes and intentions to patronize restaurants using robots by creating a model that integrates the service-dominant logic theory and the theory of planned behavior. This research also answers calls from previous research to further investigate robot acceptability in restaurants (Savela et al., 2017; Zemke et al., 2020; Wan et al., 2021; Guan et al., 2021). In addition, the study extends on Zemke et al.’s (2020) findings of restaurant customers’ concerns about the advantages and disadvantages of using robotic technology. Finally, this study explores additional critical factors that influence customers’ acceptance of using robotics in QSRs.

A service robot is defined as “a robot that operates semi- or fully autonomously to perform services useful to the well-being of humans and equipment” (IFR International Federation of Robotics, n.d.). Service robots can create a competitive advantage for industry players who adopt advanced technologies (Ivanov, Webster, & Berezina, 2017), stimulating discussions of the pros and cons of implementing the technology in the hospitality industry. Service robots could enhance the customer experience by delivering benefits such as speed and convenience (Lu et al., 2021). They also co-create service value and provide social interaction, such as when robotics assist in senior living environments (Čaić, Odekerken-Schröder, & Mahr, 2018).

Before the COVID-19 pandemic, hospitality businesses used robotics for their novelty value and to cut costs. However, some studies revealed that customers often prefer to be served by humans (Zemke et al., 2020) and view the lack of the “human touch” as a negative effect of robot implementation. Another study found that customers sometimes view a robot as an intruder and a threat to human identity (Mende, Fischer, & Kühne, 2019). Restaurant customers’ encounters with service robots have social and emotional aspects. Therefore, it is crucial to investigate customer perceptions of service robots to gain a comprehensive understanding of how robotics affect the customers’ experience and behavior.

Previous research investigated the impact of service robots on customer behavior from three perspectives (Guan et al., 2021). One perspective focused on customer attitudes toward service robots, which are determined by the degree of anthropomorphism and the level of customer trust towards service robots (Tussyadiah, 2020). For example, Li, Zhou, Jiang, Fan, and Song (2024) and Qin, Li, Zhu, and Qiu (2025) conducted scenario experiments and found that service robots with human-like appearances boost consumer trust through increased perceived warmth and perceived interpersonal attraction. Another perspective concentrated on the factors affecting customer perceptions of the value created by service robots (Choi, Liu, & Mattila, 2019), such as perceived ease of use, perceived credibility, and perceived usefulness. A third stream of research studied the variables that are important in determining customers’ intentions to use robots (Huang, Cheng, Sun, & Chou, 2021), such as perceived value, trust, and empathy.

There is a paucity of research focusing specifically on restaurants. El-Said and Hajiri (2022) tested the impact of service robots on restaurant customers’ satisfaction. They found that perceived usefulness, service speed, and novelty experience positively impacted experience satisfaction. Cha (2020) established that restaurant customers’ hedonically-motivated and socially-motivated innovativeness positively impacted their attitudes and intentions to use robots. A field study exploring the effects of the restaurant servicescape and robot competency found that both constructs positively impacted consumer dining and word-of-mouth intentions (Guan et al., 2021). Qualitative research deepened the understanding of consumer perceptions about service robots. Zemke et al.’s (2020) formative qualitative study found that quick-service restaurant (QSR) customers expressed major concerns about robotics. They were concerned that implementing service robots could adversely impact restaurant food safety, communication quality, and the sense of the “human touch” in the service setting. Zemke et al. (2020) concluded that further exploration was needed.

The overarching theories applied here are the service-dominant logic theory and the theory of consumer attitude. This theoretical framework explores service innovation with constructs related to service outcomes, service innovation process, and customer experience. Service-dominant (S-D) logic contends that customers are the value co-creators, and they acquire value through shared stakeholder norms and the service exchange process. Drawing on the S-D logic, Helkkula, Kowalkowski, and Tronvoll (2018) summarized four dimensions of value cocreation that are triggered by service innovations, including output (performance indicators), process, experience, and system. Liu, Henseler, and Liu (2022) explored technological innovation with Macau hotel guests. They found that perceived benefits and perceived ease of use are two major performance areas that customers value.

Process-based factors refer to changes in the service process that impact the guests’ perceived value of the service. An example is the employee’s role and any behavior shifts required to accommodate the robotic technology. Restaurants use service robots to reduce repetitive tasks so that employees can better focus on applying their soft skills, such as responding to customers’ emotional needs (Lu et al., 2021). Consequently, the employee’s role shifted from a “server” to a “special request handler.” In addition, job insecurity can harm employees’ psychological well-being and lead to less satisfying work performance (Lu et al., 2021).

The use of robotics also triggers social concerns among consumers. S-D logic emphasizes the social aspects of value cocreation and argues that consumers are influenced by their social network (Rubalcaba, Michel, Sundbo, Brown, & Reynoso, 2012). Likewise, Zemke et al. (2020) identified negative societal change as one of the major concerns held by QSR guests regarding the increasing use of service robots. To account for the social influence and societal concerns perceived by consumers, this study incorporated ethics and social influence into the theoretical framework to study consumer intentions to use service robots.

The creation and the success of service innovation also depend on users’ subjective experiences with the new service. QSR guests value multiple aspects of their service experience, so they have mixed feelings about service robots. They found service robots to be a novelty but were concerned that potential communication issues, such as mishaps with speech recognition systems, might ruin their service experience (Zemke et al., 2020). Wu, Fan, He, and Her (2021) conducted scenario experiments in a dining service context. They discovered that while some consumers were willing to pay more for robotic service than human service, other consumers had opposing attitudes. Individual characteristics, such as the tendency to innovate and cultural background, are critical in determining consumers’ willingness to accept service robots (Wu et al., 2021). Based on prior findings, this study integrated novelty-seeking tendencies and the need for human interaction as two focal constructs in the theoretical framework.

The current study examines how consumers’ beliefs on six dimensions affect their acceptance of service robots. The six dimensions were developed from three archetypes of service innovation according to the S-D logic. They include perceived benefits, perceived ease of use, ethics, social influence, novelty-seeking tendency, and human interaction. This study adopted the intention to use service robots as the response variable and attitude as the mediating variable between consumer beliefs and intentions. It was proposed that individuals who value human interaction are less likely to be interested in interacting with service robots because it contradicts their beliefs. On the other hand, individuals with stronger beliefs on the other five dimensions are more motivated to adopt the new technology, as the behavior is consistent with their beliefs, and they acquire psychological benefits by doing so.

The novelty-seeking theory suggests that customers seek a change from routine and boredom (Ha & Jang, 2015). Retail literature shows that consumers consider service robots to be interesting, and humanoid service robots outperform employees in attracting the consumer’s attention (De Gauquier, Willems, Cao, Vanderborght, & Brengman, 2023; Wang, Kang, Zhou, Dong, & Liu, 2022). Similarly, Zemke et al. (2020) established that study participants were intrigued by robots in restaurants as a novelty. Consumers will not be satisfied if perceived novelty is low; the opposite is true if perceived novelty is high (Toyama & Yamada, 2012). The search for a balance between security and novelty-seeking is also defined in the Optimal Arousal Theory (Plog, 2001). According to Plog (2001), most people seek an equilibrium between the predictable and the new. Therefore, restaurants using service robots are ideal for novelty-seeking customers who want to explore the new and different in a familiar environment. While what qualifies a product or service as a novelty continuously evolves, it was explored in this study as an antecedent of customer attitude and robot acceptability. Therefore, it can be hypothesized that:

Zemke et al. (2020) found that restaurant guests were apprehensive about robot use, mainly because they worried that robots might reduce the chance of human interaction. Liao and Huang (2024) found that guests experience a narrower spectrum of emotions when interacting with service robots compared to interacting with human employees. This suggests that service robots used in the hospitality industry today have not achieved the same level of communication as humans and, therefore, cannot replace human employees. In addition, industry news and research in hospitality have documented the difficulty of achieving customer satisfaction through service robots because of ineffective communications (Arikan, Altinigne, Kuzgun, & Okan, 2023; Park, Yoo, Cho, & Park, 2023). Service robots have a limited capability to understand human emotional states, which often results in misunderstandings (Söderlund, 2022). Guests respond to employees with empathy, which is not the case with service robots, where it leads to a lower tolerance for service failure (Chen, Mohanty, Jiao, & Fan, 2021). Walter, Edvardsson, and Öström (2010) found that one of the reasons for customers to frequent restaurants is, in fact, their desire for human interaction. Consequently, decreased human interaction that results from robot usage will negatively affect customers’ attitudes toward robot acceptability. However, Wan et al. (2021) discovered that the perception of reduced personal interactions with humans, achieved through the use of robotics, significantly increased intentions to visit restaurants and hotels during the COVID-19 pandemic. Nevertheless, most of the previous research emphasizes the importance of person-to-person interaction in restaurants. Therefore, this study hypothesizes that:

Customers weigh the risks and benefits of using technology when they decide whether to accept a new technology. If the benefits outweigh the risks, they will start using or continue to use the technology. There are two key benefits to be considered: functional benefits and emotional benefits. This study focuses on functional benefits, such as improved order accuracy and quality, as well as reduced waiting time and costs. A consistent positive relationship exists between the functional benefits, the ease of use, and customer attitudes and acceptance of service robots (Wirtz et al., 2018). Lin and Mattila (2021) found that service robots (e.g., delivery robots in hotels) were more reliable and more efficient than human workers. Consistent with previous service robot research, their study found positive functional benefits of service robots.

Consequently, perceived benefits are expected to positively affect attitudes toward robot acceptance. This study hypothesizes that:

Though activities that are complex and highly emotional by nature are best carried out by humans, robots will be increasingly sophisticated and capable of taking on more complex roles in social interactions. As such, the use of artificial intelligence and robots will continue to advance in the hospitality industry. However, hotel guests perceive various ethical issues regarding service robots, such as job replacement and the potential misuse of technology. In addition, ethical considerations arise during service interactions, with consumers expressing concerns about surveillance and privacy infringement (Lin, Lee, Wise, & Choi, 2024). Job replacement is rated the most critical issue among emerging ethical considerations (Etemad-Sajadi, Soussan, & Schöpfer, 2022). Some people believe that the increased use of robots will make life easier for humans, while others are concerned that robotics will replace humans and only benefit a small proportion of the population (Kumar et al., 2022). Zemke et al. (2020) identified QSR consumers’ concerns about the societal change that robot implementation may cause. For example, respondents feared that an increased presence of robots in QSRs would decrease employment for young adults or other entry-level workers.

Trevino (1986) determined that there is a cultural impact on ethical behavior. Hunt and Vitell (1986) agreed that beliefs about what is culturally acceptable play a major part in the ethical evaluation process. Reidenbach and Robin (1990) conclude that beliefs about moral equity are influenced by people’s experiences based on traditions and culture. Ferrell and Gresham (1985) add to the notion that society and culture are determinants together with knowledge and attitudes. Therefore, this study hypothesizes that:

Most previous research (Jeon, Lee, & Jeong, 2018) on service robots applied the Technology Acceptance Model (TAM). They generally find that TAM’s original constructs positively impact revisit intentions. However, this study is only concerned with one specific dimension of the model – perceived ease of use (PEOU). The TAM was initially developed to assess a consumer’s likelihood of accepting and using a device, such as a computer, an order kiosk, or a tablet – in other words, it assesses the consumer’s likelihood of adopting information technology (IT). The consumer will touch, feel, hold, and/or interact with the device – there is usually a learning curve as the consumer learns to use the device, becomes proficient, and then actively implements the device.

The current study, however, does not examine the consumer’s interactions with devices such as kiosks or tablets. Interacting with a service robot is not the same as interacting with information technology unless the consumer is programming instructions to operate the robot. It is unlikely that a QSR customer would arrive at a restaurant to order from a robot at this time – which is distinctly different from interacting with an ordering kiosk (which is not a robot). If the customer does encounter a robotic device to place an order, the instructions are most likely to be provided verbally, which will replicate the traditional practice of placing an order with a human. It is unlikely that the customer would need to learn to program or manipulate the device in some way. Therefore, some of the constructs included in the original version of the TAM are not applicable to a study of consumer interactions with robots that operate independently. The construct of perceived usefulness was excluded, as diners cannot assess the usefulness of a service robot until they learn how to operate it. PEOU was the one construct applicable for inclusion in this study.

TAM suggests that PEOU acts as a predictor of a user’s intentions to adopt a system, which in turn serves as a good indicator of the user’s actual adoption. Therefore, it is hypothesized that:

Aronson (2005, p. 6) defined social influence as “the influence that people have upon the beliefs, feelings, and behavior of others”. The mere presence of others affects customers’ behaviors (Jeon et al., 2018), and social influence is magnified if one’s reference group is present. Social interaction theory suggests that each member of society is subjected to constant judgment and that the role of social influence can be complex. Previous research suggests that social influence plays a role in forming customers’ attitudes toward technology, where the customer assumes the reference group’s beliefs and acts accordingly (Jeon et al., 2018; Lu, Cai, & Gursoy, 2019).

In the current study, social influence represents the extent to which customers conform with their social network’s beliefs concerning frequenting restaurants that employ robotics (Lu et al., 2019). An individual’s reference group tends to influence his or her attitude and behavior, and thus their behavioral intentions. This study hypothesizes that:

Attitude is formed by the combination of cognition/thought, values/beliefs, and affect/emotion toward a specific activity, person, or tangible item. Managers should understand how attitude and attitude formation processes influence customers’ preferences to engage in specific behaviors. In this study, the topic of attitude will be approached through the lens of the theory of consumer attitude, which is frequently applied through the Expectancy Value Model (Ajzen & Fishbein, 2008). The expectancy-value model suggests that an individual forms an attitude about an object by evaluating each of its attributes, combining the evaluation, and then determining how well the aggregated evaluations compare with expectations. Previous research suggests that positive attitudes significantly affect consumer intentions to use (Cha, 2020). In the current study, attitude is proposed to mediate the relationship between the antecedents of attitude formation and behavioral intentions (robot acceptance).

This study hypothesizes that attitude has a positive effect on robot acceptance intentions:

The conceptual framework of this study is presented in Figure 1. It proposes that perceived ease of use, novelty-seeking tendency, human interaction, social influence, perceived benefits, and ethics are antecedents of consumer attitudes toward robot acceptance. Meanwhile, attitudes directly impact the intention to accept the usage of robots in QSRs.

A questionnaire was distributed to an online survey panel provided by Qualtrics. Two screening questions on age and quick-service restaurant (QSR) dining experience were implemented at the beginning of the questionnaire to identify the target sample. Qualified respondents were US-based adults (age 18 years and older) who eat at QSRs at least once a month. Frequent QSR diners were chosen because QSR owners and operators are the most likely U.S. restaurant segment to purchase or lease robotic equipment at this time. Automation is expensive, and smaller operators are likely unable to afford the necessary equipment. In addition, nearly all media reports on restaurants adopting robotic technology reference chain restaurants, particularly QSRs. Therefore, specifying frequent QSR visitation provides the most rational sampling frame for a U.S.-based sample at this time.

In order to ensure participants’ understanding of service robots, a concise definition was presented: “What are robots? They are machines that perform tasks, such as cooking food or serving your food order. They are not touch pads or kiosks that take your order. Robots can move and may be able to interact with you verbally.” A photo/illustration of a robot was not provided since the study focuses on robots that the participant has either encountered first-hand or has heard about either from acquaintances or via some type of media. Photos showing a specific brand of robot may have biased the respondent’s top-of-mind concepts of a robot used in a QSR setting, such as showing a hamburger-flipping robot overriding the participant’s past experience with a food-delivery robot. The study also did not assess the level of anthropomorphism posed by the first-hand or second-hand knowledge of QSR robots in an effort to gain insight into more top-of-mind responses to robotic technology.

The questionnaire then proceeded with five questions about the respondents’ novelty-seeking tendencies, adapted from Lee and Crompton (1992) and Kim and Kim (2015). Next, respondents’ perceived importance of human interaction in the dining experience was reflected via a 5-item construct developed by Zemke et al. (2020, 2023). Items assessing perceived ease of use (PEOU) (Jeon et al., 2018), perceived benefits (Wu and Wang, 2006), moral equity if human workers are replaced with robots (Reidenbach & Robin, 1990), attitude (Ajzen & Fishbein, 2008), social influence (Lu et al., 2019), and intentions to visit a QSR that employs service robots (Kim & Qu, 2014) followed. Respondents assessed each question on a 7-point Likert-type scale, from 1 = strongly disagree to 7 = strongly agree.

The questionnaire then presented two open-ended questions about the potential benefits and problems of using robots in QSRs to gain a deeper understanding of the perceived benefits and ethical considerations. The questionnaire concluded with demographic questions asking respondents’ gender and age. Participants who completed the questionnaire received monetary compensation through the Qualtrics panel recruitment system. The questions are displayed in Supplement 1.

A pilot study was conducted to validate and establish the reliability of the construct, and the main study collected 615 usable responses. The demographic background of the sample is presented in Table 1. The sample represents the U.S. population in terms of sex, age, and race distribution, with slightly higher education levels (US Census Bureau, 2023).

Confirmatory Factor Analysis (CFA) and structural equation modeling (SEM) within Mplus were used to analyze the final sample. First, the proposed measurement structure embedded in the data was assessed using the CFA model. After confirming an acceptable fit for the measurement model, the SEM model was then specified. Fit indices used in the model fit evaluation included the Comparative Fit Index (CFI) and Tucker-Lewis Index (TLI), with a criterion of ≥0.900 for acceptable fit (Hu & Bentler, 1999). In addition, Root Mean Square Error of Approximation (RMSEA) with thresholds of ≤0.050 for a close fit and ≤0.080 for a reasonable fit were used. Bootstrap resampling was used to strengthen the test of mediation effects.

Two open-ended questions were included in the questionnaire, which sought participants’ opinions on the potential benefits and problems of using service robots. Thematic content analysis in MAXQDA 2020 was performed on the collected comments to identify the emerging patterns within the data. Comments not written in English or irrelevant to the questions were screened out in the coding process. Some examples included “I would like it” and “I would not eat there.” As a result, 597 comments related to the potential benefits of service robots in QSRs and 598 comments related to potential problems with service robots in QSRs were then included in the content analysis. Three researchers manually coded the data individually, discussed them to resolve discrepancies identified in the cross-comparison, and reached a consensus. Themes and subthemes were generated from the data through classifying codes. The frequencies of the occurrences of emerging themes were then analyzed and reported.

The CFA model demonstrated an acceptable fit when applied to the quantitative dataset, yielding χ² = 1681.730 (df = 566), p < 0.001, CFI = 0.952, TLI = 0.947, and RMSEA = 0.057 with 90% confidence interval between .054 and .060. As detailed in Table 2, all factor loadings obtained from the CFA were no lower than 0.643 (p < 0.001), exhibiting moderate-to-strong correlations. Additionally, all indicators were appropriately loaded onto their respective proposed constructs, indicating that the latent constructs were adequately represented by the observed variables. An item with a factor loading above 0.500 is acceptable under the corresponding factor and is considered good for the factor if the loading value equals or exceeds 0.700 (Hair, Black, Babin, & Anderson, 2009).

Table 3 presents the reliabilities, Average Variance Extracted (AVE), and correlations pertaining to the constructs within the measurement model. The reliability values for the constructs surpassed the recommended threshold of 0.700 (Hair et al., 2009), indicating adequate internal consistency. Examining the correlations presented in parentheses, it is evident that all factor correlations were significant and aligned with the anticipated directions. The AVE values for all constructs exceeded the 0.500 threshold proposed by Bagozzi and Yi (1988), indicating sufficient convergent validity. In addition, the AVE estimates for each pair of constructs exceeded the square of their correlation, which supports the discriminant validity of the constructs.

The analysis of the structural model indicates a good fit, where χ² = 1947.387, df = 572, p < 0.001, CFI = 0.941, TLI = 0.935, and RMSEA = 0.063 with a 90% confidence interval between 0.059 and 0.066). Figure 2 illustrates the model with unstandardized and standardized coefficients (in parenthesis). Supporting H1 and H3 through H6, novelty-seeking, perceived benefit, ethics, PEOU, and social influence directly impact attitudes towards using robots in quick-service restaurants (QSRs). Consistent with predictions, attitude positively affects the intention to visit QSRs that employ robots. Contrary to expectations, human interaction positively affects attitude and, consequently, intention to visit the restaurant. As a result, H2 is rejected. Overall, the model explained 75.1% of the variance in attitude and 75.0% in intention to use. Among all the antecedents specified in the model, perceived benefit exerts a strong impact on attitudes (β = 0.348). Specifically, each unit increase in perceived benefit results in a 0.348 unit increase in attitude, assuming other predictors held constant. Social influence and perceived ease of use show moderate influence on attitudes, with β = 0.247 and β = 0.221, respectively. Human interaction (β = 0.185) and ethics (β = 0.133) exhibit weak effects on attitudes. Novelty-seeking demonstrates a significant impact but has the lowest effect on attitudes (β = 0.079), suggesting a minimal practical influence on attitude towards the use of service robots.

We hypothesized the mediating effects of attitude in the relationship between the exogenous variables specified in the model and the consumers’ QSR visit intentions. The bootstrapping technique texted the significance of these indirect effects. The results in Table 4 reveal significant indirect effects of all exogenous variables on visit intention via attitude. The standard coefficients in Table 4 suggest that the relative strength of the indirect effects from exogenous variables aligns with their direct effects on attitude. Consequently, H7 was supported. Table 5 presents a summary of the hypotheses testing results.

Respondents’ answers to the open-ended questions further investigating their perceptions of the potential benefits and potential problems with the use of service robots were coded and classified, subject to the researchers’ interpretations. Themes that were mentioned fewer than three times were not included. Examples of the benefits of using service robots, such as “fast delivery of food”, “fast response”, and “fast service”, are all coded as “improved service speed”, which is identified as a subtheme under “improve service quality”. Tables 6 and 7 list overarching themes, subthemes, frequencies, and sample quotes.

Overall, respondents reported that the benefits of using service robots could be reflected in four major themes, including improving service quality, enhancing operational efficiency, increasing the competitive advantages of QSR, and “none” (i.e., no potential benefits). A total of 360 comments summarized that service robots could enhance the quality of service by minimizing service failures through increased order accuracy, increasing service speed, providing better service through achieving consistency in service quality, improving the hygiene of the dining environment by reducing the chance of germ spreading when cleaning the table or delivering the dishes to the tables, and reducing human error, such as cashier mistakes. A few respondents also mentioned that service robots improve food hygiene/safety due to reduced human contact. Among these aspects of service quality, service speed and order accuracy were prominent, with frequencies of 150 and 119, respectively.

In addition to providing customer benefits, service robots benefit business operators by enhancing operational efficiency and competitive advantages. In particular, respondents reported that service robots help the QSR owners reduce labor costs, such as wages and benefits. Other benefits for the operation include improving efficiency by doing more work in a shorter time, minimizing management effort due to robots not needing attention, rewards, recognition, or scheduling, and reducing or eliminating absenteeism issues. Service robots also help the company avoid labor shortages when employees call in sick. Some respondents believe a reduced labor cost will lead to a lower product price offered by the QSR. A few participants also mentioned that using service robots is a novel idea that will attract customers. Interestingly, 99 respondents believe using a service robot does not provide any benefit at all, and another 20 respondents reported that they are unsure of the benefits a service robot can bring.

Respondents’ comments regarding the problems with using service robots fell into four major categories: the limitations of technology, increasing unemployment, downgrading customer experience, and social/societal concerns. Two hundred fifty-four comments summarized various limitations of technological devices that negatively impact QSR operations, including machine malfunctions, the inability to process complicated tasks or provide personalized service, the inability to cook as well as a human, the inability to remedy a service failure when a customer is not satisfied or frustrated, high maintenance costs, and possible downtime when there is a power outage. A notable 175 respondents expressed concerns about humans losing jobs due to service robots. In addition, 87 comments indicated that the customer service experience would be downgraded because service robots are extremely new and may scare guests. The respondents suggested that people, especially seniors, may have barriers to learning how to use this new technology, and more importantly, robots cannot provide the same level of interpersonal interaction with the guests as human employees. Twelve respondents expressed concerns regarding social issues, such as the injustice of robots taking people’s jobs away. Interestingly, 44 respondents believed that using service robots did not have any problems, while another 21 respondents held the opposite opinion that there was nothing right about using service robots.

This study employed a survey method to examine the antecedents of diners’ intentions to frequent quick-service restaurants (QSRs) that use robots. It also tested the mediating role of consumer attitude between the proposed determinants and the consumers’ restaurant visit intentions. The structural equation modeling results support the proposed hypotheses by showing that attitude is directly impacted by novelty-seeking tendencies, the need for human interaction, the perceived benefits, the ease of use of the service robots, the consumer’s beliefs about moral equity, and the consumer’s social influencers. The results revealed that consumer attitude positively influences diners’ intentions to dine at restaurants that use service robots in a QSR setting. The results implied that individual preferences, the features of the new technology, and social norms jointly shaped consumers’ attitudes toward service robot usage. More specifically, respondents with strong novelty-seeking tendencies and those who believed that using service robots was ethical (fair and just) showed favorable attitudes towards using service robots in QSRs. In addition, positive attitudes towards robot usage increased when respondents believed that service robots were beneficial to them and/or were easy to use. Finally, diners whose peers favored the idea of service robots were more inclined to accept the usage of service robots.

However, Hypothesis 2 (the desire for human interaction negatively influences attitudes) was rejected. This finding suggests that QSR guests may not see service robots as a replacement for human employees but as tools that enhance the quality of human interaction. Respondents perceived benefits of service robots, such as improving service speed and order accuracy, may allow human staff to focus more on personalized service and meaningful interactions, which could explain why those who value human interactions exhibit a positive attitude toward robots. This finding contradicts Zemke et al. (2020), wherea negative effect of human interaction on intention to accept service robots was found. However, their study was conducted before the COVID-19 pandemic and consumer beliefs about the functionality of service robots may have changed over time and due to the pandemic. It is possible that the respondents accept the deployment of robotics in QSRs, but they would not be as amenable to this technology at a more formal level of restaurant service. It could also be that restaurant customers will accept robot applications in their daily lives as long as their needs for human interactions are appropriately met in other ways. Finally, the results further indicated that respondents’ favorable attitudes positively impacted their overall intentions to visit QSRs that employ robots. These findings provide valuable insights to researchers, as well as hospitality industry practitioners.

Two open-ended questions provided qualitative data that supplement the quantitative findings. The questions revealed the scope of consumer-perceived benefits and ethical considerations. Prevalent benefits include increasing operational efficiency and improving service quality. The majority of respondents who described potential problems referenced the limitations of the current technological capabilities, such as potential software and hardware malfunctions and a lack of personalized service resulting in lowered service quality. The respondents identified operational efficiencies, including reduced labor costs, increased general efficiency, and a solution for labor shortages. Nearly one-third of the survey respondents suggested that they could not identify any benefits of deploying robots in QSRs, while 68% expressed concerns about robotic technology leading to increased unemployment rates.

This study expands on previous research by investigating the aspects that influence QSR customers’ attitudes toward the application of robots in QSRs. The conceptual framework of the current research incorporates elements of the service-dominant logic theory (S-D Logic) and the theory of planned behavior (TPB). While recent research has primarily studied service robots using scenario experiments, where the type or function of the service robots is manipulated (Lu et al., 2021), the current study draws conclusions based on consumers’ actual beliefs and perceptions of service robots. This research contributes both qualitative and quantitative empirical evidence to the literature. The findings expand the body of knowledge by demonstrating how restaurants can use technology as a source of support and competitive advantage. The study expands on S-D Logic in general (Helkkula et al., 2018) by exploring customers’ involvement in innovation and by providing specifics on what influences customers’ attitudes so robots can be implemented in restaurants in a way that contributes to improving customer value cocreation. Therefore, this study also contributes to strengthening the fundamental premises of S-D Logic – that stronger relationships between companies and customers emerge as the outcome of value cocreation (Lusch & Vargo, 2014). This research also added to and reaffirmed the Expectancy Value Model (Ajzen, 2008) by identifying attributes influencing restaurant customers’ attitudes toward the application of robots.

It is also interesting to see that nearly 24% of the respondents hold opposing attitudes toward service robots. Around 16.56% of the respondents did not believe there was any benefit associated with QSRs using service robots. By comparison, 7.36% did not believe there was any problem deploying service robots in QSRs. This phenomenon can be explained by consumer innovativeness tendencies, which reflect the degree to which customers are amenable to new ideas (Wu et al., 2021).

The benefits that robots provide customers are the most important factor significantly influencing their attitudes towards service robot usage and, consequently, their intentions to visit restaurants that employ service robots. This result supports the findings of Wirtz et al. (2018) and Lin and Mattila (2021). It also complements previous S-D Logic research (Lusch & Vargo, 2014), which describes S-D Logic as beneficiary-specific. This study added to the literature on robot use in restaurants by showing results contrary to findings from previous research (Zemke et al., 2020, 2023; Wirtz et al., 2018) in which respondents expressed concerns about a possible lack of human interaction, or the “human touch”, in service encounters. Thus, today’s restaurant customers value human interaction and also accept the application of robots. This may be situationally dependent, as Moriuchi and Murdy (2024) found when they explored customer acceptance of robots in restaurants. Their study revealed that restaurant customers preferred human-delivered services but would accept robot-based service at casual dining establishments.

It is worth noting that the effect size of novelty seeking is small, indicating a weak connection between the application of service robots and the concept of novelty in consumers’ cognition. This phenomenon may attributed to the increasing application of robotic technology’s effect on diminishing its attachment to novelty. The effect size of ethics is also relatively small, which implies that ethical considerations might not be a primary driver of QSR consumers’ attitudes and intentions toward service robots.

This research also identified the effects of social influence on customers’ attitudes. It supported previous research by revealing that consumers’ behavioral intentions are significantly affected by their peers and families. “Social influence” directly impacted customers’ attitudes about robots. However, there may be significant additional influences from the technology itself. For example, a recent study by Choi, Wan, and Mattila (2024) used experimental design and found that when a more humanoid-looking robot offered an array of choices to a study participant, the participant was more likely to select a more healthful option, whereas when the robot had a less humanoid appearance, the participants were more likely to select a more “self-indulgent” and less healthful option. As the robot became more human-like, the authors speculate that the study participants may have felt the urge to conform to social desirability norms and select healthier options despite the fact that the robot (theoretically) does not judge behavior like a human might.

Extending the theory of consumer attitude, this study uniquely examined “attitude” in mediating the relationships between the antecedents (novelty seeking, human interaction, ethics, social influences, perceived benefits, and perceived ease of use) and customers’ intentions to frequent restaurants that employ robots. Moreover, the positive influence of perceived ease of use (PEOU) supports the widely applied Technology Acceptance Model (TAM) (Davis, 1989; Jeon et al., 2018) by revealing that functionality significantly impacts the intention to adopt new technology.

Finally, the results showed that positive customer attitudes significantly influence the intention to visit, supporting previous research (Cha, 2020). This supports the findings of a recent study by Santiago, Borges-Tiago, and Tiago (2024), who explored consumer attitudes toward technology, interest in robots, artificial intelligence, and service automation (i.e., RAISA), enjoyment, and the technology’s relevance to survey respondents’ needs. These dimensions contributed to a strong positive relationship with their intentions to use RAISA-type technologies in restaurant situations.

Artificial intelligence and robotic technology may permanently change the operating dynamics of QSRs (Meisenzahl, 2022). Practitioners must continue to learn how consumer perceptions of these technologies evolve. This study has several managerial implications.

Emphasize benefits. First, it reveals what influences restaurant customers’ attitudes and intentions to visit QSRs that employ robots. Most importantly, customers’ perceived benefits of robot use drive acceptance. Managers should emphasize and continually improve such benefits, including highlighting the benefits of reduced human interactions in a restaurant that uses robotic technology. The benefits can include increased product quality, sanitation, reduced waiting times, and the experiential aspects of robotics, particularly for guests with high novelty-seeking tendencies. These efforts should be designed to overcome the hesitation of customers who do not trust the technology yet.

While the “novelty seeking” and “ethics” constructs positively influenced QSR customers’ attitudes and visiting intentions, their effects are not as strong as other variables examined. The small effect size suggests that robotic-related marketing initiatives targeting novelty-seeking consumers or those with strong ethical values are unlikely to significantly improve consumer intentions to visit restaurants. Managers should recognize that customers’ perceptions have changed, indicating that robots used in restaurants are no longer a great novelty and are not much of an ethical concern anymore. In other words, business operators should appeal to consumers based on factors with stronger relationships, such as highlighting the benefits of using service robots and emphasizing their ease of use.

Social influence. In addition, this study shows managers the importance of the influence of customers’ social networks on their attitudes. The social influence construct has the second highest impact on customers’ attitudes about robots being present in QSRs. Managers should make an effort to understand their customers well in general along with who influences their behaviors. However, while determining which groups have social influence on a customer may be difficult, managers could engage in frequent informal communication with customers to shed light on what types of social influence groups the customers have.

For example, a QSR manager could (and should) casually interact with dine-in guests. This can be done in person by the manager, or frontline employees can be instructed to engage in casual conversation with guests while taking orders or delivering orders. The goal of the interaction can be to identify which social media platforms the guest is on, whom the guests are dining with (if there are two or more at a table), why the guest(s) chose to dine there, and other socially-focused questions. These are commonsense activities, but if an operator relies solely on survey-based data and operational data, the operator misses an important opportunity to interact with guests to gather intelligence. In addition, the QSR owner or operator should actively monitor the brand/property’s social media activities, loyalty program data, or the sources for coupons and other promotions that the guest might use. It seems like an obvious recommendation, but we find (anecdotally) that many businesses collect data without analyzing it and using the results to improve operations.

Effective surveying. Most people today are inundated with post-service experience surveys. Again, anecdotally, we have not observed any recent surveys that ask questions about how decisions are made to dine. For example, we do not see questions about why the restaurant was selected, the occasion for the visit, other people at the diner’s party, etc. Larger restaurant brands may want to consider rotating social influence questions through the short post-visit satisfaction surveys distributed to guests.

Collaboration versus replacement. It should also be noted that some respondents were concerned that humans might be replaced by robots needlessly. To address such concerns, restaurants may launch marketing initiatives highlighting the benefits of collaboration between employees and service robots. In fact, some customers have directly observed restaurant robots successfully serving in roles supporting human workers, rather than replacing them. For example, the restaurant chain Kura Sushi, a US-based brand, uses service robots to deliver drinks to tables during peak hours, which reduces the intense workload of human employees (Kelso, 2022). In addition, robots permit humans to perform different and more complex tasks and allow humans to be more productive, which is contrary to Zemke et al.’s (2020) findings, where customers were worried about robots replacing humans. For example, the Dawn Cafe in Japan hires disabled staff to control robots that serve guests, providing social and employment opportunities to individuals with physical limitations (BBC, 2023). This phenomenon may also explain the result that, while customers do value human interaction in restaurants, they still hold a positive attitude toward robot usage in QSRs. Management should ensure that human employees create and maintain the human connections their customers desire for this specific style of restaurant service.

Service failure. Management must actively emphasize, maintain, and improve the benefits of robots identified in this research, including improved quality, cleanliness, accuracy, and food safety. Finally, management can address concerns expressed in this study, such as robots’ inability to provide personalized service and appropriate service recovery, as well as respondents’ apprehensions about malfunctioning technology. Recent research found that ensuring information transparency and providing compensation options contribute to successful recovery in the context of robot service failures (Yu, Liu, He, Huang, & Li, 2024). By comparison, offering financial compensation is more effective than apologies that focus on restoring the relational tie when the service recovery is led by the robot (Nguyen, Nguyen, & Hancer, 2025).

This is particularly relevant in the event of a service failure. A recent study by Ryoo, Jeon, and Kim (2024) found that when a human employee makes a mistake, the customer blames the human employee. However, if a robot makes a mistake, the customer does not blame the robot but instead assigns the blame for the problem to the company as a whole. Therefore, restaurant companies that are considering deploying robot technology that guests may be aware of, such as a robot taking the customer’s order or where the customer can see a robotic device preparing food in the kitchen, any service failure that occurs may tarnish the reputation of the company as a whole, rather than simply assigning the blame to the individual human employee.

This study, like all studies, has limitations that could be addressed in future research. Respondents in this study answered questions related to their past experiences with QSRs, which may result in inaccurate information recall. A future study may utilize a field experiment instead, where the subjects respond to questions that assume that they are physically present in the service environment with the robot. The researchers might also employ an observational technique. The added advantage of an observational study is that it can help removing the influence of recall bias or social desirability bias.

Next, this study sampled only adults who frequently dine at QSRs, regardless of their prior experience with service robots. Therefore, participants’ responsesdepended on anecdotal rather than first-hand evidence. In future studies, scholars can focus specifically on individuals who have interacted with service robots to examine their intentions in revisiting restaurants that utilize this technology. This could also be further organized into front-of-house and back-of-house robotic applications.

This study’s sample consisted of adults in the United States, which limits the generalizability of the findings. Future studies could include a multi-national sample to provide direct comparisons between different countries and cultures. Alternatively, there is a growing body of knowledge related to the use of robotics in the hospitality industry. Researchers could conduct a literature review summarizing customer attitudes, awareness, and usage/acceptance of robot technology on a region-by-region basis.

This study focused on frequent QSR diners and did not consider other restaurant categories, such as fast-casual, full-service, and fine-dining restaurants. Future research could examine if the type of dining service affects the guest’s perceptions of the usage of robotics in the business, as alluded to in Zemke et al. (2020). This might explain the competing and seemingly disparate phenomena in which consumers value human interactions while maintaining positive attitudes toward the application of robots in restaurants.

In addition to the variables examined in this study, future research can explore the connections between other factors and consumer attitudes toward service robots. The current research identified variables with limited practical influence on consumer attitudes and intentions. These findings may be context-dependent, and future research could examine other antecedents alongside the weak-effect factors identified in this study (i.e., novelty seeking, ethics) to determine if the results differ across contexts.

Finally, this study provided a definition of what a robot is, but did not describe a robot’s features/appearance beyond the fact that a robot performs a task and can move autonomously. The purpose was to ensure that study participants understood that the questions would not focus on the use of ordering apps or kiosks (which are not robots). The definition did not discuss appearance or the level of anthropomorphism. Future research may go beyond the functional aspects of the benefits measured in this study and address the emotional impacts of service robots on customer attitudes, behavioral intentions, and actual behavior. For example, very recently published research now evaluates consumer attitudes toward robots with different levels of anthropomorphism. Moriuchi and Murdy (2024) found that the more human-like a robot restaurant server appeared to be (compared to a human server), the greater the study respondent’s intention to visit a restaurant that employs robots. Similarly, Choi et al. (2024) used a scenario-based approach to explore a human customer’s tendency to be self-indulgent. The scenario presented a situation that included Pepper, a slightly humanoid-looking robot manufactured by SoftBank Robotics, offering a welcome beverage versus a non-humanoid robot. The level of humanoid-like characteristics influences the selection of beverages. It should be noted that the Pepper robot cannot, at this time, physically move objects; it is supposed that the Pepper scenario in the study simply verbally communicated the choice but did not physically deliver the selection.

The robotic technology used in other countries, particularly in Asia, is more prevalent and often more human-looking than the devices currently used in the US. Most of the technology in the US (where the current study was conducted) consists of robotic assembly/heating equipment (including cooking burgers and fries), or food delivery and table bussing in the form of rolling carts. These devices may have a cartoon-like appearance but are decidedly not human-looking. The scholarly research and the actual technology development are moving toward assessing the effects of increasingly human-like devices, which should provide fertile ground for researchers for some time to come.

The availability of robotic technology in restaurants is rapidly increasing. While the technology could offer many benefits to restaurant owners/operators, including reduced labor costs, increased productivity, and improvements to service and food quality and consistency, it is still unclear how well the restaurant guests will view this technology. This study examined the quick-service restaurant (QSR) guests’ perceptions of robotic technology and the guests’ intentions to patronize restaurants that deploy the technology. It identifies various factors impacting attitudes toward this technology, including novelty-seeking tendencies, human interaction, social influences, perceived benefits, ethical concerns, and perceived ease of use. While the desire for human interaction did not reduce intentions to patronize a QSR, the desire for human interaction did form a strong factor that can be used for explanatory power in future studies. QSR guests may be simply not looking for strong human interactions when visiting QSRs; they may look for more human connections when visiting full-service establishments. Finally, this study did identify an emerging concern regarding robotic technology replacing humans, which points to a universal emerging challenge of balancing the benefits and the ultimate dangers of using AI and AI-driven technology.