The rapid advancement of the low-altitude economy has positioned drone-based food delivery as a transformative force in China’s urban logistics landscape. As a novel service model within the realm of smart logistics, the commercial success of drone food delivery critically hinges on user acceptance. This study, grounded in the Unified Theory of Acceptance and Use of Technology (UTAUT), investigates the determinants of user adoption behavior. We extend the foundational model by incorporating two pivotal variables: trust and perceived risk, to construct a comprehensive theoretical framework tailored to the unique context of drone food delivery in China. An empirical investigation was conducted using a questionnaire survey, yielding 498 valid responses. Data analysis was performed using structural equation modeling and the Bootstrap mediation effect test. Our findings indicate that performance expectancy, effort expectancy, social influence, facilitating conditions, and trust significantly and positively influence users’ behavioral intention to use drone food delivery services. Conversely, perceived risk exerts a significant negative impact on this intention. Furthermore, behavioral intention acts as a full mediator in the relationship between these core antecedents and actual usage behavior. All research hypotheses were supported. This study not only extends the application of the UTAUT model to an emerging technological scenario but also provides actionable insights for operators of China drone delivery services to enhance user acceptance, optimize service design, and formulate effective marketing strategies.
1. Introduction
As an emerging growth point of the low-altitude economy, the drone industry in China is experiencing unprecedented development opportunities. In 2021, the “National Comprehensive Three-dimensional Transportation Network Planning纲要” issued by the Central Committee of the Communist Party of China and the State Council proposed the development of the low-altitude economy. By 2023, the low-altitude economy was established as a strategic emerging industry, and it experienced explosive growth in 2024. The focus of the drone industry’s development has shifted from the manufacturing of drone platforms towards specialized services across various sectors of the economy and society. This shift is particularly evident in applications such as drone food delivery. Internationally, drone delivery projects have proven effective in high-timeliness scenarios like emergency medical supply delivery. In China, Meituan had launched 55 routes by July 2025, completing over 520,000 orders. Through the “drone + smart locker” model, it has achieved large-scale operations in cities like Shenzhen. Despite the rapid development of drone food delivery, academic research on users’ continued usage intention remains relatively underdeveloped.
As an extension of the low-altitude delivery domain to the general public, drone food delivery exerts a significant influence on residents’ daily lives. Current research in the low-altitude delivery field primarily focuses on collaborative delivery, path planning, and risk profile construction, with limited studies on public acceptance of this new service. Moreover, the rapid growth of the low-altitude economy has led to more frequent use of drone food delivery, yet the factors determining public acceptance remain unclear. This gap underscores the need for a systematic investigation into the antecedents of users’ intention to use China drone food delivery services.
To address this issue, this study employs the UTAUT model to explore the factors influencing users’ adoption behavior and their underlying mechanisms. Empirical evidence has shown that this model explains up to 70% of the variance in individual user usage behavior, offering greater explanatory power than other models. To better align with the context of drone food delivery in China, we incorporate two additional variables—trust and perceived risk—into the existing framework. This extension allows for a more nuanced analysis of user behavior, capturing the complex psychological processes involved in adopting this novel technology.
The key innovations of this study are threefold. First, traditional technology acceptance models may have limited explanatory power for contemporary technological applications; hence, a new theoretical model is required. Our proposed model extends the conventional UTAUT framework by adding two variables to elucidate the complex user psychology behind ordering meals via drones. Second, while most research on China drone delivery addresses technological perspectives, this study adopts a user-centric approach, thereby enriching the research angle. Third, this research ranks the importance of various influencing factors, providing a solid empirical basis for future studies on drone food delivery.
2. Literature Review
As a segment of the low-altitude economy most closely related to public life, the user acceptance of drone food delivery in China is influenced by multiple factors. Previous research has explored these influences from various angles. For instance, one study, grounded in the Technology Acceptance Model (TAM), found that perceived usefulness and perceived risk directly affect residents’ usage attitudes. Other researchers identified age, education level, and monthly income as primary factors influencing the willingness to use drone delivery. While these perspectives have enriched our understanding, their differing viewpoints make it challenging to gain a comprehensive picture of the factors affecting drone food delivery usage behavior.
The UTAUT model, developed by Venkatesh et al. (2003), integrates eight distinct theories into a unified research framework for technology acceptance. Its core constructs include performance expectancy, effort expectancy, social influence, and facilitating conditions. The model also incorporates moderating variables such as age, gender, and experience, effectively explaining users’ intention and behavior towards new technologies. Researchers have increasingly applied this model to analyze users’ intention and behavior regarding novel services. For example, one recent study integrated the UTAUT2 model with a structural equation modeling-artificial neural network (SEM-ANN) approach to study users of digital currency, incorporating trust as a core variable. Another study added perceived risk and trust theories to the UTAUT model to construct a model for user behavior on UGC-based smart tourism service platforms, thoroughly analyzing key factors influencing usage behavior.
We believe that trust and perceived risk play indispensable roles in shaping user behavior towards China drone delivery services. Trust is fundamental to commercial transactions as it influences consumers’ subjective initiative and behavioral intentions, thereby affecting usage behavior. In the era of the internet economy, trust has garnered increasing attention. Furthermore, studies have found that perceived risk serves as a key indicator for effectively assessing consumers’ willingness to accept new technologies. Although drone food delivery can offer significant convenience, user concerns about the associated risks are also growing.
As a novel delivery method, drone food delivery in China offers advantages such as bypassing traffic congestion and reducing labor costs. However, it also faces challenges related to insufficient infrastructure coverage and varying levels of social recognition. Therefore, investigating the continued usage intention for drone food delivery is a crucial prerequisite for its trial operations and widespread adoption.
While the UTAUT model has been widely used to explain technology acceptance, there is a pressing need for research that applies this model specifically to the acceptance of drone food delivery. This study, based on the UTAUT model, combines the technical characteristics of drone delivery with user experience pain points. By rationally introducing the new perspectives of trust and perceived risk, we aim to provide a more comprehensive and systematic exploration of the influencing factors and their effects on the usage behavior of China drone delivery services.
3. Research Hypotheses and Model Construction
3.1 Research Hypotheses
3.1.1 Performance Expectancy and Behavioral Intention
In the UTAUT model, performance expectancy refers to the degree to which an individual believes that using a technology will help attain gains in job performance. If users perceive that adopting a new technology will enhance their performance, this increased expectancy will positively influence their behavioral intention. In the context of drone food delivery in China, this is reflected in the user’s assessment of aspects like delivery speed and service reliability. For example, drone delivery can enable rapid service, and its enclosed cargo compartments can better ensure food safety and hygiene. Previous research on online learning behavior has confirmed this relationship. Hence, we propose:
H1: Performance expectancy positively influences behavioral intention.
3.1.2 Effort Expectancy and Behavioral Intention
Effort expectancy is defined as the degree of ease associated with the use of the technology. In this study, effort expectancy for drone food delivery encompasses the entire operational experience, from placing an order on an app to receiving and verifying the delivery. This directly impacts the user’s actual experience. Studies on technology acceptance models have shown that perceived ease of use significantly affects users’ behavioral intention. If the process is cumbersome or the interface is complex, requiring significant user effort, this psychological burden can become a barrier to adoption. Thus, we propose:
H2: Effort expectancy positively influences behavioral intention.
3.1.3 Social Influence and Behavioral Intention
Social influence is a critical factor in the literature on new IT adoption. Many studies have found that social influence significantly affects an individual’s intention to use new information technology, as people often adjust their attitudes, beliefs, and behaviors based on their social network. Research on public acceptance has shown that social influence positively affects users’ behavioral intention. If friends and family are using drone food delivery, and the media actively promotes its convenience, an individual’s need for social integration may increase their frequency of use. Hence, we propose:
H3: Social influence positively influences behavioral intention.
3.1.4 Facilitating Conditions and Behavioral Intention
Facilitating conditions refer to the degree to which an individual believes that an organizational and technical infrastructure exists to support the use of the technology. This is a key factor influencing usage intention. In the context of drone food delivery in China, this refers to the external support required, including the coverage of smart pickup lockers, technical reliability, and policy guarantees. Empirical evidence has demonstrated a significant positive effect of facilitating conditions on behavioral intention. Since drone delivery is still in its early stages, users may have doubts about its ability to complete the service, affecting their usage intention. Therefore, we propose:
H4: Facilitating conditions positively influence behavioral intention.
3.1.5 Trust and Behavioral Intention
Trust is defined as a psychological state involving a willingness to accept vulnerability based on positive expectations of the technology’s behavior. It not only directly influences behavioral intention but also serves as a core cognitive and emotional basis for reducing perceived uncertainty and promoting technology adoption. Research indicates that trust can effectively lower users’ risk perception, thereby directly and positively driving behavioral intention. Trust is also critical in the drone food delivery scenario. When users have confidence in the system’s delivery security and data protection mechanisms, their behavioral intention is more likely to be positively influenced. We thus propose:
H5: Trust positively influences behavioral intention.
3.1.6 Perceived Risk and Behavioral Intention
Perceived risk refers to a consumer’s subjective feeling of uncertainty about the potential negative consequences of a purchase or service decision. This mainly stems from concerns about negative outcomes, such as financial loss. Studies have shown that various types of perceived risk lower target users’ acceptance of drone delivery technology. Users might worry about delivery failures (e.g., damaged food, wrong address) or have concerns about safety and privacy breaches. These risk perceptions can significantly reduce their behavioral intention. Thus, we propose:
H6: Perceived risk negatively influences behavioral intention.
3.1.7 Behavioral Intention and Usage Behavior
Behavioral intention is the subjective probability or tendency of an individual to try or continue using a product or service. A user’s intention can serve as the internal motivation that drives their actual usage behavior. In the case of drone food delivery, even with a strong intention, the actual translation to behavior can be influenced by various factors (e.g., seeking others’ opinions, promotional activities). However, in general, a stronger user intention correlates with a higher probability of conversion into actual behavior. Therefore, we propose:
H7: Behavioral intention positively influences usage behavior.
3.1.8 The Mediating Role of Behavioral Intention
Behavioral intention often acts as a mediating variable in acceptance models, linking external variables to actual usage behavior. Factors like performance expectancy, effort expectancy, and social influence may have direct effects on behavior but, more importantly, they influence actual behavior indirectly through behavioral intention. Based on our preceding hypotheses, trust and facilitating conditions positively influence behavioral intention, which in turn positively affects usage behavior. Consequently, we further propose that behavioral intention mediates the relationship between trust and usage behavior. Similarly, perceived risk negatively influences behavioral intention, while behavioral intention positively influences usage behavior. Therefore, we hypothesize a mediating role for behavioral intention in the path from perceived risk to usage behavior. We propose the following mediation hypotheses:
H8a: Behavioral intention mediates the relationship between performance expectancy and usage behavior.
H8b: Behavioral intention mediates the relationship between effort expectancy and usage behavior.
H8c: Behavioral intention mediates the relationship between social influence and usage behavior.
H8d: Behavioral intention mediates the relationship between facilitating conditions and usage behavior.
H8e: Behavioral intention mediates the relationship between trust and usage behavior.
H8f: Behavioral intention mediates the relationship between perceived risk and usage behavior.
3.2 Model Construction
Based on the above hypotheses, we constructed a research model where performance expectancy, effort expectancy, social influence, facilitating conditions, trust, and perceived risk serve as independent variables. Behavioral intention towards drone food delivery acts as the mediating variable, and usage behavior is the dependent variable. This model is designed to explain the complex decision-making process behind the adoption of China drone delivery services.
4. Research Design
4.1 Questionnaire Design
This study collected data through a questionnaire survey. Given that drone food delivery is not yet universally implemented, public acceptance is more reliant on subjective judgments. The availability of objective data from public sources is limited. Therefore, the questionnaire survey is the most reliable method for obtaining data on public acceptance of drone food delivery. To ensure data validity, the questionnaire was divided into three parts. The first part collected respondents’ basic demographic information. The second part measured the core influencing factors derived from our model. The third part assessed respondents’ behavioral intention and actual usage behavior. The measurement items were adapted from existing validated scales. A five-point Likert scale was used for scoring, with scores from 1 to 5 representing “strongly disagree” to “strongly agree.”
4.2 Data Collection
The survey was conducted using a combination of online and offline methods, targeting four cities in the Yangtze River Delta region with high consumer spending: Ningbo, Hangzhou, Nanjing, and Shanghai. By July 4, 2025, a total of 530 questionnaires were collected. After screening for invalid responses, 498 valid questionnaires were retained, yielding an effective response rate of 93.96%. The demographic analysis revealed that 50.20% of the respondents were aged between 26 and 35, indicating a concentration of young to middle-aged adults. This demographic is typically more sensitive to new technologies, possesses a good level of scientific knowledge, and has ample life experience, which ensures the authenticity and validity of the data for subsequent analysis.
5. Data Analysis
5.1 Reliability and Validity Test
To assess the reliability of the sample data, we performed reliability and validity tests using SPSS 26.0 software. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was 0.927, indicating that the data was highly suitable for factor analysis. Bartlett’s Test of Sphericity was significant (approx. Chi-Square = 8197.325, p < 0.001), confirming that the variables were sufficiently correlated for factor analysis. The overall Cronbach’s alpha for the entire scale was 0.900, and the Cronbach’s alpha for each individual variable exceeded 0.7, demonstrating excellent internal consistency and convergent validity for our measurement model.
5.2 Discriminant Validity Test
To test the discriminant validity among the latent constructs in our model, we calculated the Pearson correlation coefficients between all pairs of variables. The analysis results showed that all correlation coefficients were well below 0.7, indicating that the core variables in the model exhibit a moderate degree of intercorrelation. This confirms they are related but distinct concepts. Notably, perceived risk was negatively correlated with all other variables, and all absolute correlation values were significantly below 0.7, which aligns with our theoretical expectations. The strongest negative correlation was between perceived risk and trust, which is theoretically sound. Furthermore, the square root of the Average Variance Extracted (AVE) for each construct was greater than its highest correlation with any other construct. This pattern provides initial support for adequate discriminant validity among the constructs in our study.
5.3 Mediation Effect Test
To formally test the mediating role of behavioral intention, we conducted a mediation effect analysis using the Bootstrap method. The results are summarized in the following table.
| Path | c (Total Effect) | a | b | a*b (Indirect Effect) | Boot SE | z | p | 95% BootCI | c’ (Direct Effect) |
|---|---|---|---|---|---|---|---|---|---|
| 1: PE => UI => UB | 0.597** | 0.614** | 0.474** | 0.291 | 0.032 | 9.003 | 0.000 | 0.112 ~ 0.238 | 0.306** |
| 2: EE => UI => UB | 0.328** | 0.372** | 0.568** | 0.211 | 0.025 | 8.546 | 0.000 | 0.105 ~ 0.201 | 0.117 |
| 3: SI => UI => UB | 0.607** | 0.338** | 0.351** | 0.119 | 0.022 | 5.320 | 0.000 | 0.055 ~ 0.143 | 0.488** |
| 4: FC => UI => UB | 0.278** | 0.437** | 0.620** | 0.271 | 0.025 | 10.700 | 0.000 | 0.107 ~ 0.207 | 0.007 |
| 5: TR => UI => UB | 0.549** | 0.645** | 0.510** | 0.329 | 0.032 | 10.346 | 0.000 | 0.126 ~ 0.251 | 0.220* |
| 6: PR => UI => UB | 0.074 | -0.117** | 0.675** | -0.079 | 0.019 | -4.268 | 0.000 | -0.114 ~ -0.042 | 0.153** |
Note: * p<0.05, ** p<0.01. Bootstrap type = percentile bootstrap. PE=Performance Expectancy, EE=Effort Expectancy, SI=Social Influence, FC=Facilitating Conditions, TR=Trust, PR=Perceived Risk, UI=Behavioral Intention, UB=Usage Behavior.
The results indicate that for paths 1, 3, and 5 (Performance Expectancy, Social Influence, and Trust), behavioral intention plays a partial mediating role, as the direct effect (c’) remains significant. For paths 2 and 4 (Effort Expectancy and Facilitating Conditions), behavioral intention plays a full mediating role, as the direct effect (c’) becomes non-significant. For path 6 (Perceived Risk), the indirect effect is negative and significant, while the direct effect (c’) is positive and significant. This pattern, where the total effect (c) is small and non-significant, is indicative of a suppressor or mediating effect where the indirect path carries the negative influence.
5.4 Structural Equation Modeling (SEM) Analysis
Given the satisfactory reliability and validity of our measures, we proceeded with a structural equation model analysis. The model was specified according to our hypotheses. The key regression coefficients are presented in the following table.
| Path (X -> Y) | Unstandardized Coefficient | SE | z (CR) | p | Standardized Coefficient |
|---|---|---|---|---|---|
| PE -> UI | 0.360 | 0.045 | 7.916 | 0.000 | 0.311 |
| EE -> UI | 0.109 | 0.037 | 2.942 | 0.003 | 0.113 |
| SI -> UI | 0.136 | 0.030 | 4.526 | 0.000 | 0.163 |
| FC -> UI | 0.090 | 0.045 | 2.006 | 0.045 | 0.075 |
| TR -> UI | 0.349 | 0.048 | 7.338 | 0.000 | 0.285 |
| PR -> UI | -0.051 | 0.020 | -2.555 | 0.011 | -0.072 |
| UI -> UB | 0.720 | 0.057 | 12.705 | 0.000 | 0.495 |
As shown in Table 2, all proposed paths were statistically significant. The critical ratio (CR) values for all hypotheses (H1 through H7) were greater than 1.96, with most being significant at the p < 0.001 level. The path from behavioral intention to usage behavior (H7) showed the highest standardized coefficient. The results support all our direct effect hypotheses: H1, H2, H3, H4, H5 are positively related to behavioral intention, H6 is negatively related, and H7 is positively related. The strength of influence on behavioral intention, from strongest to weakest, was: Performance Expectancy (0.311), Trust (0.285), Social Influence (0.163), Effort Expectancy (0.113), Facilitating Conditions (0.075), and Perceived Risk (-0.072).
5.5 Summary of Results and Discussion
All hypotheses in our study were supported. Performance expectancy emerged as the strongest predictor of behavioral intention, indicating that users of China drone delivery services are most concerned with the technology’s ability to improve timeliness and provide functional value. Trust was the second most important factor, underscoring that reliability is a prerequisite for user adoption. Effort expectancy had a moderate but significant impact, highlighting the importance of a seamless user experience for user retention. Social influence and facilitating conditions also played significant roles. Perceived risk, while having a smaller coefficient, exerted a negative effect, which operators must address. The full mediation of behavioral intention for effort expectancy and facilitating conditions suggests that these factors primarily influence actual use by first shaping the user’s intention. The findings align with the core propositions of Venkatesh et al. (2003) and demonstrate the unique explanatory power of trust and perceived risk in the specific context of China drone delivery. These results suggest that operators should prioritize performance enhancement, trust-building measures, and process simplification to foster sustained user adoption.
6. Implications
6.1 Theoretical Implications
First, this study extends the applicability of the UTAUT model to the specific context of the low-altitude economy, confirming that its core constructs are significant predictors of behavioral intention in the case of drone food delivery in China. Second, by integrating trust and perceived risk, we constructed a theoretical model that is more appropriate for studying the acceptance of China drone delivery services. This addresses the theoretical call to understand the “risk-trust paradox” in emerging technology adoption and provides a scalable variable framework for future research. Third, by adopting a micro-level user decision-making perspective, this study enriches the empirical literature on technology acceptance within the low-altitude economy domain. The findings reveal that in the early stages of technology diffusion, performance expectancy, effort expectancy, and trust are key drivers, while perceived risk acts as a significant barrier.
6.2 Practical Implications
Based on our findings, we propose several practical recommendations for stakeholders in the China drone delivery ecosystem. First, companies should focus on optimizing operational processes and simplifying interface design to reduce the user’s learning cost. Simultaneously, they must improve the efficiency and stability of drone delivery to enhance performance. Second, fostering collaboration with government bodies is crucial. Through policy advocacy and co-development of regulations, a favorable social and legal environment can be created to promote public acceptance. Finally, companies should strengthen equipment management and safety protocols. Marketing efforts should center on the themes of “stability” and “efficiency” to build user trust. Implementing these recommendations can positively influence behavioral intention, which in turn will drive actual usage behavior for drone food delivery.
7. Limitations and Future Research
This study has several limitations that suggest avenues for future research. First, the sample was concentrated in the Yangtze River Delta region. Future studies should expand the geographic scope to enhance the generalizability of the findings. Second, our model did not include all potential influencing variables, such as environmental awareness or hedonic motivation. Future research could explore the differential impact of these variables on various user segments. Third, future studies could employ complementary qualitative methods to delve deeper into users’ underlying needs and barriers, providing a richer complement to the quantitative findings presented here.