Measuring Consumer Experience in Community Unmanned Stores: Development of the ECUS-Scale for Omnichannel Digital Retail

Abstract

As consumer behavior increasingly shifts toward hyperlocal, digitally mediated retail journeys, community unmanned stores have emerged as a transformative model that integrates smart technologies with community proximity services. These fully automated stores offer convenient, contactless shopping and hybrid digital–physical interactions, playing an increasingly important role within broader omnichannel digital retail ecosystems. However, there remains a lack of validated instruments to assess customer experience in such autonomous and locally embedded retail formats. This study develops and validates an ECUS-scale (an experience in community unmanned store scale), a multidimensional measurement tool grounded in qualitative research and refined through exploratory and confirmatory factor analysis. The scale identifies nine key dimensions—convenient service, smooth transaction, preferential price, good quality, safe environment, secure payment, comfortable space, comfortable interaction, and friendly image—across 36 items. These dimensions reflect the technological, spatial, and emotional–social aspects of customer experience in unmanned retail settings. The findings demonstrate that the ECUS-scale offers a robust framework for evaluating consumer experience in low-staffed, tech-enabled community stores, with strong relevance to omnichannel digital retail strategies. Theoretically, it advances the literature on smart retail experience by capturing underexplored dimensions such as emotional engagement with technology and perceptions of safety in staff-free environments. Practically, it serves as a diagnostic tool for businesses to enhance experience design and optimize customer engagement across digital and physical touchpoints.

1. Introduction

In recent years, the retail industry has witnessed two parallel yet interconnected trends: globalization and localization [1,2,3]. On the one hand, globalization in the consumer sector has driven large-scale retailers to expand across borders, standardizing offerings and leveraging economies of scale to meet global demands [4,5]. On the other hand, consumer preferences have shifted toward hyperlocal, personalized services delivered within the “last mile”—or even the “last 100 m” [6,7,8,9]. These contrasting forces reflect a growing tension between globally scaled operations and locally tailored experiences, with digital technologies enabling greater responsiveness and flexibility in localized service delivery [10,11,12].

Emerging at the intersection of these trends is the concept of smart retail, which broadly refers to the digitally mediated transformation of retail operations and experiences through technologies such as artificial intelligence (AI), self-service technologies (SSTs), the Internet of things (IoT), autonomous checkout systems, and big data analytics [13,14,15,16]. It emphasizes not only operational efficiency but also data-driven personalization and seamless omnichannel integration [17,18]. Within this domain, a more specific and localized variant—community unmanned stores—has gained momentum, representing a hybrid model that blends digital automation with community-level embeddedness.

Community unmanned stores are fully automated, staff-free retail spaces [16,19,20] that are situated within defined communities—such as residential neighborhoods, office complexes, or university campuses. They offer 24/7 convenience, intelligent checkout, and self-service shopping experiences while maintaining physical proximity and cultural resonance with local communities. Among recent implementations, stores such as Amazon Go Grocery, BingoBox, and Take Go exemplify this trend by combining the technological infrastructure of smart retail with community proximity—that is, hyperlocal deployment, geographic intimacy, and social familiarity [3,21,22]. While previous studies have explored unmanned retail formats in general [19,20,23,24,25,26], this study builds upon them by introducing community unmanned stores as a concept that emphasizes both technological autonomy and community-level relevance. This framing aligns with recent work on proximity-based consumption [3,12,21,22], the role of retail spaces in fostering social cohesion [27,28,29], and urban last-mile sustainability [30,31].

At the same time, the global economy has moved from a service-based logic to an experience economy, where emotional value, personalization, and immersive engagement are central to consumer decision-making [32,33,34]. Studies show that factors such as familiarity, comfort, and emotional resonance significantly shape consumer satisfaction and behavioral intention [35,36,37]. In community unmanned stores, customers are not merely completing transactions; they are engaging in digital–physical hybrid environments where technology mediates not only service but also affective and social experience. These spaces often serve as informal hubs of community life, contributing to a sense of belonging and community identity [23,38].

Importantly, community unmanned stores are increasingly integrated into omnichannel digital retail ecosystems. As hyperlocal nodes, they extend customer touchpoints beyond traditional online or flagship stores, enabling real-time pick-up, returns, or supplementary shopping within an otherwise digital journey [39,40]. While distinct from O2O models in their autonomous operation and lack of service staff, they still contribute to the seamless continuity of brand experience across physical and digital channels [41,42]. This distinction from traditional O2O models has important implications for how customer experience is conceptualized and measured in these settings. Thus, understanding the experiential logic of community unmanned stores is essential for omnichannel strategy, experience design, and service innovation.

Yet, despite their growing strategic relevance within omnichannel ecosystems, academic research on customer experience in community unmanned stores remains fragmented. While prior studies have examined influencing factors such as servicescape [24], marketing mix factors [25], technical usability [26], spatial layout [43], and experiential psychological states [44], few have developed a validated, multidimensional measurement tool specifically tailored to the community-embedded, technology-mediated nature of this retail model. Some related efforts have been made in omnichannel and O2O contexts. For instance, Pei et al. (2020) developed a customer experience index system for O2O food delivery platforms, incorporating dimensions such as browsing, tangible, empathetic, and assurance experiences [45]. Frasquet-Deltoro et al. (2021) proposed a brand experience scale specific to omnichannel retailing, comprising eight dimensions, including sensory, affective, relational, and social experiences [46]. While these scales are empirically grounded and domain-relevant, they are often tailored to specific service sectors and customer journeys and may not be directly transferable to the operational logic of community unmanned stores due to differences in technical infrastructure, customer interaction modes, and contextual embeddedness.

Moreover, established service quality scales—such as the Retail Service Quality Scale (RSQS) [47], the Integrated Store Service Quality Scale [48], and the Omnichannel Retail Service Quality Scale [49]—have laid foundational groundwork for evaluating perceived service quality across traditional and digitally augmented retail settings. However, these instruments primarily assess attributes such as reliability, responsiveness, and channel integration, often under the implicit assumption of human service involvement or guided omnichannel coordination. In contrast, customer experience is a broader, more subjective construct encompassing consumers’ cognitive, emotional, sensory, and social responses throughout the entire retail journey [50]. This distinction is especially pronounced in fully autonomous retail spaces, where staff absence, technological mediation, and community characteristics co-shape the experiential environment.

As such, while existing service quality scales provide useful benchmarks, they fall short in capturing key experiential dimensions of community unmanned stores—such as emotional engagement with intelligent systems, perceived autonomy during shopping, and comfort within staff-free yet locally embedded environments. A more context-specific, experience-centered approach is therefore needed to understand and evaluate how consumers interact with, interpret, and emotionally respond to this novel retail format. To fill these conceptual and contextual gaps, this study develops and validates an ECUS-scale (an experience in community unmanned store scale), a context-specific, empirically grounded instrument designed to evaluate customer experience in these hybrid spaces. To our knowledge, this is the first study to construct and validate a multidimensional scale specifically tailored to the experiential characteristics of community unmanned stores, highlighting the unique intersection of digital automation, community proximity, and emotional–social engagement. Through grounded theory analysis and exploratory–confirmatory factor procedures, the scale identifies nine key dimensions that reflect the digital, physical, emotional, and social layers of customer experience. In doing so, it contributes to both the theory and practice of omnichannel digital retail, offering a robust tool for assessing experiential quality in emerging, low-contact retail contexts.

The remainder of this paper is organized as follows: Section 2 outlines the research aims and methodological framework adopted for scale development. Section 3 presents the exploratory phase, including qualitative data collection and grounded theory analysis. Section 4 details the process of scale purification and validation through exploratory and confirmatory factor analysis. Section 5 summarizes the main results. Section 6 discusses and concludes the theoretical and practical implications, limitations, and future research directions.

2. Aims and Methods

The main aim of this paper is to investigate the experiences people have when interacting with community unmanned stores. The specific sub-aims are twofold. First, to clarify the concept of consumer experience in community unmanned stores by identifying and distinguishing the various dimensions of experience related to products, services, brands, and store environments. For example, the study explores how the store’s brand image, along with factors such as product quality and store layout, influences consumers’ emotional connection, trust, and overall experience. Second, to develop a scale to measure consumer experience in community unmanned stores. This scale will be straightforward, applicable to any community unmanned store, and will incorporate both the experience enabled by intelligent technology and the experience within the physical service environment. In summary, the research questions (RQs) guiding this study are as follows:

RQ1: What are the key dimensions of consumer experience when interacting with community unmanned stores?

RQ2: How can consumer experience in community unmanned stores be effectively measured, incorporating both technological and physical service elements?

Based on the research’s aims and questions, and by referencing related research methods, such as Kuppelwieser & Klaus (2021) [51], Rejikumar & Asokan-Ajitha (2022) [52], and Zhang et al. (2022) [48], we developed the process and methodology for this study. The development of the ECUS-scale consisted of two phases: an exploratory phase and a confirmatory phase. Each phase involved several studies. An overview of the process is shown in

Table 1. Overview of the phases in the development of the ECUS-scale.

Phase	Stage	Methodology	Sample	Data Collection	Results
1. Exploration of the concept and item generation	Search for dimensions and items	Grounded theory	36 community unmanned store customers	Face-to-face and telephone	12 dimensions of community unmanned stores
	Item generation and pilot test of the ECUS-scale	Pilot test	5 experience experts and 5 customers	Face-to-face and e-mail	Generation and improvement of questions of the concept of the ECUS-scale
2. Scale purification and validation	Item screening and factor structure	Survey, exploratory factor analysis (EFA)	336 community unmanned store customers	Face-to-face, online and paper questionnaire	Validation of the dimensions and condensed version of the ECUS-scale
	Validation of the factor structure	Survey, confirmatory factor analysis (CFA)	336 community unmanned store customers	Face-to-face, online and paper questionnaire	Validation of the dimensions

. The objective of phase one was to explore the consumer experience in community unmanned stores from the customer’s perspective using grounded theory and generate scale items. Phase two involved purifying the scale and validating its factor structure.

3. Phase 1: Exploration of the Concept and Item Generation

3.1. Grounded Theory Deployment

Customer experience in community unmanned stores belongs to the category of customers’ subjective feelings, which are challenging to construct theoretically using quantitative means [53,54]. The grounded theory (GT) approach is a bottom-up qualitative research methodology that allows researchers to develop theory inductively by observing data and building theoretical relationships through reasoning, comparison, and hypothesis testing [55,56,57]. This approach helps discover research clues and propose an abstract analytical framework for a phenomenon, ultimately generating or discovering a theory [55,56]. It is particularly suitable for understanding complex, subjective phenomena such as consumer experience and has been widely applied in the field of experience design [53,58].

This study adopts Straussian grounded theory (SGT), which is an abductive approach that involves observing data that emerge from empirical research, identifying theoretical relationships, and then developing core categories that help understand the research problem [57]. SGT is particularly valuable as it provides systematic research procedures and rigorous analytical tools for data analysis, which enables the construction of well-supported theories [59,60]. In this study, open coding, axial coding, and selective coding were employed to generate initial categories and identify relationships between them, building a three-level conceptual framework for consumer experience in community unmanned stores.

However, while the full grounded theory methodology was applied, including the use of three-level coding and theoretical saturation, the goal of this study was to develop a simple, practical, and effective measurement scale for consumer experience. To this end, the research concentrated on a two-level coding structure for scale development. Accordingly, only the results from open and axial coding were retained in the final scale, as they sufficiently captured the core dimensions of consumer experience aligned with the study’s objectives and enhanced the scale’s applicability in practical settings. A similar approach can be found in [48], where grounded theory was applied to develop a multidimensional measurement scale for omnichannel retail service quality. In this study, the authors used open and axial coding to identify the key dimensions of service quality and created a simplified scale by focusing on essential dimensions [48], just as this study emphasizes a simplified structure to ensure the scale’s practical applicability.

3.1.1. Data Collection

A purposive sampling method was used to identify participants for this study. The sample was selected based on the following criteria: (a) participants lived in communities with access to unmanned stores, (b) participants frequented these unmanned stores and were familiar with the shopping process, and (c) participants demonstrated the ability to observe and reflect on their experiences. For instance, some of the sample’s participants were sourced from Hengshu Community in Huaqiao Economic Development Zone, Kunshan City, China. This community is located near Huaqiao Library, Kangqiao School, Jishan New Village, Changfa Xiangchengwan Residential Area, and other activity centers, and there are also various community convenience stores nearby, such as Moment Home, 52 Week, and Convenient Bee. Based on the above principles, this study ultimately conducted interviews with 36 typical customers of community unmanned stores. The basic information of the interviewees is presented in

Table 2. Basic information of the interview participants.

Variable	Items	Frequency	Percentage (%)
Gender	Male	19	52.78
	Female	17	47.22
Age	≤18	2	5.56
	18–25	8	22.22
	26–30	12	33.33
	31–40	10	27.78
	≥41	4	11.11
Education	High school, technical secondary school, and below	4	11.11
	Junior college	20	55.56
	Bachelor’s degree or above	12	33.33

. Among the interviewees, there were two minors. Consent for their participation was obtained from both the minors themselves and their guardians.

One-on-one semi-structured face-to-face or telephone interviews were conducted with 36 customers. The interview process was designed to be flexible, allowing adjustments based on the flow and dynamics of each conversation. The respondents were encouraged to openly share their genuine feelings, suggestions, and expectations regarding the topic of customer experience in community unmanned stores. Some of the key questions asked during the interviews included the following: “What motivated you to use a community unmanned store for the first time?”; “Do you frequently shop at community unmanned stores?”; “Why do you usually shop at community unmanned stores?”; and “How would you describe your shopping experience at a community unmanned store? Could you share any positive or negative experiences?” Each interview lasted approximately 30 min. With the respondents’ consent, the interviews were audio-recorded, transcribed manually, and the resulting data were imported into NVivo 11 for analysis. In total, 36 valid customer interviews were collected, of which 24 were used for coding, and the remaining 12 were employed to test theoretical saturation.

3.1.2. Findings

Once the data were collected, grounded theory analysis was applied following the basic steps, with the details of open coding and axial coding shown in

Table 3. The findings of open coding and axial coding.

Axial Coding	Open Coding	Excerpts from Answer Statements
Convenient service	Extended opening hours, close location, matching of commodity structure, timely replenishment of commodities, diversified daily services	“This unmanned store is just downstairs of my house. It is open late. It seems that it is 24 h a day. Sometimes I go there to buy some snacks late at night”.
Clear shopping guide	Sufficient shopping guide information, clear and effective shopping guide information, reasonable spatial layout	“The first time I used an unmanned store, there was no one inside, so I knew how to use it by looking at its logo. It was posted on the entrance door. There was also a video guide at the payment place”.
Smooth process	Convenient and fast shopping process, complete auxiliary services, convenient for help, effective service recovery mechanism, other customer impacts	“The shopping process is quite convenient, the checkout is fast, and there is no queue”.
Preferential price	Price discount, preferential activities, regular customer discount	“It is reasonable to say that there is no cost for human services in unmanned stores. Why should I buy a drink bottle here at 0.5 yuan more expensive than in other places? Shouldn’t it be cheaper?”
Good quality	Fresh goods, reliable commodity quality, goods are in good condition	“I feel that the product update in the unmanned store is very slow, so every time I buy something in it, I pay special attention to its production date”.
Safe environment	Atmosphere safety, clear and sufficient safety guidance, easy access to safety equipment, knowing how to escape	“When I go to the unmanned store to buy snacks late every day, I am terrified. There is no one inside. It is quiet around”.
Security payment	Accurate and reliable payment technology, clear feedback, payment procedure security, payment errors are easy to correct, payment guarantee	“I always worry that the machine will miscalculate the number of goods I buy. I must check the shopping list after each purchase carefully”.
Safe information	Information provided with confidence, legal and formal use of information, information security concerns	“With so many cameras in the store, my privacy won’t leak out, will it?”
Comfortable space	Comfortable physical atmosphere, clean and tidy space, appropriate space, environmental esthetics, space privacy, shopping freedom	“It is very cool in the unmanned store, and the temperature is very comfortable when you go to buy drinks after running”.
Comfortable interaction	Easy-to-use equipment, natural comfort, esthetic feeling of form, pleasant sound	“I was shopping when a warning tone suddenly appeared, which frightened me”.
Friendly technology	Content and tone are compatible, proactive and friendly service, close and considerate service	“I can directly ask the store where the shopping guide robot products are and how much they cost. It’s very convenient”.
Affine environment	Other customers are friendly, community characteristics, breath of daily life, strong holiday atmosphere	“The customers in the store are all in the same neighborhood, so we are familiar with each other and will say hello when we meet”.

. Finally, twelve categories were identified: convenient service, clear shopping guide, smooth process, preferential price, good quality, safe environment, secure payment, safe information, comfortable space, comfortable interaction, friendly technology, and affine environment. We will discuss these twelve categories in detail below.

Convenient Service

This item refers to the customer’s experience of the services provided by community unmanned stores that can quickly meet residents’ consumption needs, especially in emergencies. It mainly refers to the customer’s evaluation and feelings regarding the ease of access to the series of services provided. Compared to supermarkets located in urban centers or commercial areas, community unmanned stores are open 24 h a day and have the goods needed for daily life, which can more quickly and conveniently meet the more urgent needs of community residents. It provides customers with a more convenient and timely shopping experience. Convenient service is the main reason customers choose to spend their money shopping at these stores.

Clear Shopping Guide

The customer experience of a shopping guide system in a community unmanned store guides the consumer through the shopping task quickly and smoothly. There is no salesperson or cashier in community unmanned stores; customers must understand the shopping process, product features, equipment usage, etc., through the guidance system to complete the shopping task. Therefore, an easy-to-read and easy-to-understand guidance system is a prerequisite for customers’ successful consumption of services in community unmanned stores.

Smooth Process

Smooth process means that the original processes of entering, shopping, paying, and leaving the store, which need to be handled one by one by customers, become coherent, automatic, and integrated so that customers can focus more on the smooth shopping experience. “No queue”, “grab and go”, “quick checkout”, and “remote emergency service” are the keywords that were mentioned a lot in the interviews with customers. Customers can also quickly solve the problem through intelligent devices or call the company’s emergency service system in case of a service failure during the shopping process, such as remote customer service, further safeguarding and enhancing the smooth shopping experience of customers.

Preferential Price

It is one of the key factors affecting customer experience in community unmanned stores, and it was mentioned many times in the interviews. Common feedback included the following: “There is no cost for human service, why is it more expensive for me to buy goods here than elsewhere, shouldn’t it be cheaper?” The products sold in community unmanned stores are daily consumer goods, which are very similar to the product structure of ordinary supermarkets. Accordingly, customers are more familiar with and sensitive to their prices.

Good Quality

The quality of products is a crucial experiential demand for customers when shopping at community unmanned stores. Customers look for characteristics such as shelf life, place of production, brand, and customer reviews to assess the value of goods. Most products sold in community unmanned stores are fast-moving consumer goods used daily, making it easy for customers to perceive their quality. However, the lack of human interaction in unmanned stores can create the impression that the goods are not updated frequently, potentially leading to concerns about freshness. The freshness of products is an essential factor that affects the customer experience of the quality of goods and their purchase decisions.

Safe Environment

Safe environment refers to customers’ perceptions and reactions to whether their surroundings provide for their safety when shopping in community unmanned stores. Compared to traditional community stores with enthusiastic service staff, community unmanned stores have characteristics such as “unmanned and enclosed space” that make customers feel isolated and helpless. They thus have a more robust demand for the security of community unmanned stores.

Secure Payment

Secure payment is an important aspect of the customer experience in community unmanned stores, as it addresses concerns around protecting personal financial security. Regardless of whether a community unmanned store relies on the RFID tag technology that requires customers to scan a code for payment or biometric and camera capture technology to obtain an automatic deduction, customers may have doubts about the accuracy of these technologies. For instance, customers may fear mistakenly scanning goods and not being able to revoke the payment or experiencing machine recognition errors that result in the loss of personal property.

Safe Information

Safe information refers to customers’ expectations for the protection of their private shopping and behavior data in community unmanned stores. In these stores, cameras, sensors, face recognition, payment settlement, and other devices can monitor customers’ private information such as walking trajectory, shopping demeanor, face portrait, and list of purchased items in the store. While such data can be used to provide personalized services, customers are concerned about the security and privacy of their information. With the rapid development of the Internet technology and the importance individuals place on privacy, customers are becoming increasingly sensitive to their privacy and security.

Comfortable Space

This refers to the atmosphere, privacy, and esthetic design in community unmanned stores, which aim to provide customers with a sense of relaxation and balance. Relevant studies have shown that music, lighting, odor, temperature, cleanliness, and other atmospheric elements that individuals can perceive through their senses have a significant impact on customers’ psychology and behavior [61]. Customers appreciate the privacy offered by community unmanned stores, particularly when buying intimate items, which can be uncomfortable to discuss in public due to cultural norms surrounding sex in the Chinese culture. Furthermore, shopping at community unmanned stores allows customers to escape from interpersonal relationships and shop more freely.

Comfortable Interaction

Customers should feel physically and mentally relaxed and happy when using a series of devices in a community unmanned store. Comfort is a key indicator in evaluating the human–computer interaction experience. As community unmanned stores lack attendants, customers need to complete shopping tasks through self-service devices such as self-scanning payment and self-warming food devices. The visual comfort of external features, such as screen brightness, clarity, and design elements of the service interface, can also affect customers’ comfort level [62,63].

Friendly Technology

Customers in community unmanned stores want smarter, more proactive, welcoming, and humane technology devices. In a traditional community store, customers can chat with familiar store staff, receive warm greetings and recommendations, etc. However, customers can only interact with technology in a community unmanned store to meet their shopping needs. Studies show that anthropomorphism provides the possibility of social connection based on meeting the needs of customers’ cognitive world, and customers can obtain a sense of “partnership” and the pleasure of communication from anthropomorphism, thus forming an emotional bond between customers and brands or products [64,65]. Therefore, friendly technology is a means to make the interaction between customers and technology devices more natural and what customers expect from technology humanization.

Affine Environment

On the one hand, it refers to the friendly attitude, mutual acceptance, and closeness felt while interacting with other customers in community unmanned stores. Jung & Park (2020) concluded through an empirical study that similarities with other customers in unmanned stores positively impact the emotional and social experience [26]. Community customers have similar lifestyles, familiar social relationships, and a common culture. They are more likely to accept and embrace each other. On the other hand, it refers to the sense of self-concept or belonging evoked by the signs, artifacts, or representative decorations, background music, etc. The physical elements in the service place that are rich in social meaning, such as representative background music, poster patterns that match the customer’s image or preference, and humanized signs, can also attract the customer’s attention and feelings [66].

3.2. Item Generation

A comprehensive version of the ECUS-scale was designed based on the grounded theory findings of phase one, as shown in

Table 4. Initial scale of the customer experience of community unmanned stores.

Category	No.	Item	Item Details	References
Convenient service	TS1	Extended opening hours	I feel it is very convenient to be able to shop at this community unmanned store anytime and anywhere.	Retail Service Quality Scale [47]; Smart Service Value Scale [68]
	TS2	Close location	I feel that this community unmanned store is close enough to where I study, live, or work to get there quickly.
	TS3	Matching of commodity structure	This community unmanned store sells a comprehensive range of goods to meet my daily needs.
	TS4	Timely replenishment of commodities	I feel that the number of items sold at this community unmanned store is replenished promptly.
	TS5	Diversified daily services	I feel that this community unmanned store provides diverse services for daily life.
Clear shopping guide	CSG6	Sufficient shopping guide information	This community unmanned store provides comprehensive shopping guide information.	Retail Service Quality Scale [47]; Retail Brand Experience Scale [69]; Omnichannel Retail Service Quality Scale [49]
	CSG7	Clear and effective shopping guide information	I feel that the guidance information within this community unmanned store is easy to understand.
	CSG8	Reasonable spatial layout	I feel that this community unmanned store’s space layout and shelf placement are reasonable.
Smooth process	SP9	Convenient and fast shopping process	I feel that the shopping process of this community unmanned store is efficient.	Retail Service Quality Scale [47]; Omnichannel Retail Service Quality Scale [49]; Smart Service Value Scale [68]; Internet Users’ Information Privacy Concerns Scale [70]
	SP10	Complete auxiliary services	I feel that the services provided by this community unmanned store can support me to complete my shopping smoothly.
	SP11	Convenient for help	I feel it is convenient to ask for help at this community unmanned store.
	SP12	Effective service recovery mechanism	I feel that if shopping is interrupted at this community unmanned store, the system will provide timely remedies.
Preferential price	PP13	Price discount	I feel that the prices of the goods sold at this community unmanned store are reasonable.	Integrated Store Service Quality Measurement Scale [48]
	PP14	Preferential activities	I feel that this community unmanned store has attractive price discounts.
	PP15	Regular customer discount	I feel that frequent shopping at this community unmanned store will be rewarded with corresponding discounts.
Good quality	GQ16	Fresh goods	I feel that the goods sold at this community unmanned store are fresh.	Retail Service Quality Scale [47]
	GQ17	Reliable commodity quality	I feel that the quality of the goods sold at this community unmanned store is reliable.
	GQ18	The goods are in good condition	I feel that the items sold at this community unmanned store are in good condition.
Safe environment	ES19	Atmosphere safety	I never worry about my safety when shopping at this community unmanned store.	/
	ES20	Clear and sufficient safety guidance	I feel that this community unmanned store has sufficient safety guidance information.
	ES21	Easy access to safety equipment	I feel that this community unmanned store’s safety equipment is easily accessible.
Secure payment	PS22	Accurate and reliable payment technology	I feel that the smart technology in this community unmanned store accurately identifies items and calculates prices.	Retail Service Quality Scale [47]; Integrated Store Service Quality Measurement Scale [48]
	PS23	Clear feedback	I feel that the shopping list provided by this community unmanned store was clear.
	PS24	Payment procedure security	I feel that the payment process of this community unmanned store is secure.
	PS25	Payment guarantee	I feel that when I pay at this community unmanned store, even if there is a payment error, I can easily make a complaint to recover the money lost.
Safe information	IS26	Information provided with confidence	When this community unmanned store requires me to provide private information, I feel comfortable providing it.	Omnichannel Retail Service Quality Scale [49]; Internet Users’ Information Privacy Concerns Scale [70]
	IS27	Information security concerns	I feel that my private information is safe when I shop inside this community unmanned store.
	IS28	Legal and formal use of information	I feel that the way this community unmanned store uses customers’ personal information is legitimate.
Comfortable space	CS29	Comfortable physical atmosphere	I feel that the physical environment inside this community unmanned store is comfortable.	Retail Service Quality Scale [47]; Integrated Store Service Quality Measurement Scale [48].
	CS30	Clean and tidy space	I feel that the environment and the arrangement of goods in this community unmanned store are neat.
	CS31	Appropriate space	I feel that the shelves in this community unmanned store are moderately spaced.
	CS32	Environmental esthetics	I feel that the design of this community unmanned store is esthetically pleasing to my eyes.
	CS33	Space privacy	I do not feel embarrassed to buy private items at this community unmanned store.
	CS34	Shopping freedom	I feel that shopping in this community unmanned store is unsupervised.
Comfortable interaction	IC35	Easy-to-use equipment	I feel that the self-service equipment in this community unmanned store is easy to use.	Omnichannel Retail Service Quality Scale [49]
	IC36	Natural comfort	I feel that the interaction with technological devices in this community unmanned store is natural.
	IC37	Esthetic feeling of form	I feel that the interface and shape of the self-service equipment in this community unmanned store are exquisite.
	IC38	Pleasant sound	I feel that the cueing sound in this community unmanned store is pleasant.
Friendly technology	TF39	Content and tone are compatible	I feel that the content and tone of this community unmanned store’s guidance tips and signs are amiable.	Retail Service Quality Scale [47]; Omnichannel Retail Service Quality Scale [49]
	TF40	Proactive and friendly service	I feel that the services provided by this community unmanned store are proactive.
	TF41	Close and considerate service	I feel that the services provided by this community unmanned store are approachable.
Affine environment	EA42	Other customers are friendly	I feel that the other customers in this community unmanned stores are cordial.	Omnichannel Retail Service Quality Scale [49]
	EA43	Community characteristics	I feel that this community unmanned store’s soft decoration aligns with the community characteristics.
	EA44	Strong holiday atmosphere	I feel that this community unmanned store has a strong holiday atmosphere during the holidays.

. For each attribute, a statement was created. Because the target survey respondents were Chinese, we followed the back-to-back translation procedure [67]. To ensure the accuracy of the questions, we also referenced established scales from relevant studies. The main references for these scales are also listed in the Table 4. Additionally, it is important to note that although there are no actual employees in community unmanned stores, the study drew on various scales, such as the employee-related items from the “Retail Service Quality Scale” [47] and the Omnichannel Retail Service Quality Scale [49], for examples. These served as valuable references for the concept of “smart employees”, referring to services provided by intelligent unmanned technologies in this study. Finally, 48 statements with clear and meaningful constructs were confirmed.

To ensure the content validity of the initial ECUS-scale, five individuals with expertise in user research and experience design were invited to participate in a focus group. The group was tasked with reviewing the clarity, conciseness, and semantic expression of the content, as well as the correlation between items and the degree to which latent variables measured by the initial scale were explained. The basic information of the participants is provided in

Table 5. Basic information of focus group members.

No.	Major	Education	Gender	Usage Frequency
1	User researcher	Master	Female	About 2–4 times a week
2	User researcher	Master	Female	About 2–4 times a week
3	Experience designer	Doctor	Female	About 3–5 times a week
4	Experience designer	Master	Male	About 2–4 times a week
5	Product manager	Bachelor	Male	Almost everyday

. Four items with semantic repetition and ambiguous conceptual expressions were deleted, and eight items with unclear semantics and cumbersome expressions were rephrased. Lastly, to ensure the face validity of the initial scale, pre-study and questionnaire feedback interviews were conducted with five people who had used community unmanned stores. They were asked to complete the questionnaire based on their experiences, problems, and feelings while using community unmanned stores. Feedback was also solicited regarding the questionnaire’s clarity, redundancy, and understandability of each measurement question. Based on the feedback, some of the questions were semantically rephrased.

4. Phase 2: Scale Purification and Validation

In Phase 1, we identified the twelve critical dimensions of the experience in community unmanned stores based on the grounded theory. However, can these dimensions truly represent the brick-and-mortar store customer experience quality? How should the community unmanned store customer experience be evaluated? Next, we conducted a questionnaire survey to answer these questions using the measurement scale originating from the qualitative analysis findings and relevant literature.

4.1. First Stage: Item Screening and Factor Structure

For the purification and validation of the ECUS-scale, two quantitative studies were performed. The first stage (field study 1) focused on (1) screening the items to reduce the initial pool to a more manageable size using critical ratio (CR) and (2) exploring the underlying structure of the data using exploratory factor analysis (EFA). The second stage (field study 2) involved validation of the factor structure of the condensed scale by using confirmatory factor analysis (CFA) on a new dataset [71,72,73].

4.1.1. Data Collection and Sample Profile

A questionnaire consisting of three parts was designed for customer experience evaluation of community unmanned stores. The first part eliminated the concerns of the participants by introducing the purpose and motivation of this study and screened out those participants who had experience in using community unmanned stores in the previous six months to ensure the validity of the study sample. The second part was the theme of the questionnaire, which was mainly the specific content of the customer experience evaluation scale of community unmanned stores. The scale adopted a 7-level Likert scale, and customers of community unmanned stores could choose according to their actual experience. The third part was the basic information of the subjects, such as age, gender, education, etc. The purpose of putting the basic information of the subjects at the end of the questionnaire was to reduce the negative emotions caused by filling in the personal information of the subjects and ensure the quality of the questionnaire data as much as possible.

This study employed a convenience sampling method to distribute questionnaires both online and offline. Respondents were only allowed to complete the questionnaire after providing their consent. A total of 758 questionnaires were distributed online and offline using convenience sampling. To ensure the validity of the questionnaires, 82 questionnaires with short response time, serious logical contradictions in answers, and all questions with identical answers were considered invalid and eliminated, and 676 valid questionnaires were obtained. The basic information of the respondents is shown in

Table 6. Basic information of the respondents.

Variable	Items	Frequency	Percent (%)
Gender	Male	351	51.92%
	Female	325	48.06%
Age	≤18	10	1.48%
	18–25	158	23.37%
	26–30	205	30.33%
	31–40	165	24.41%
	≥41	138	20.41%
Education	High school, technical secondary school, and below	193	28.55%
	Junior college	424	62.72%
	Bachelor’s degree or above	59	8.73%

. Anderson & Gerbing (1988) suggested that exploratory factor analysis and confirmatory factor analysis should be based on different sample data for testing [74]. Therefore, the study randomly divided data of the 676 valid questionnaires into two parts of 338 questionnaires each, which were used for exploratory factor analysis (EFA) and confirmatory factor analysis (CFA).

4.1.2. Scale Purification

Following standard scale development procedures, this study first used item analysis based on critical ratio (CR) values to remove items with poor discrimination, i.e., items with a T-value of less than 3.000 between the high and low groups of items [75]. The analysis process was as follows: first, the valid scale data were summed and sorted by SPSS 26.0 software. The data samples were divided into high and low subgroups based on the first 27% and last 27% of the total scores calculated after sequencing. Independent-sample t-tests were performed on the high and low subgroups to calculate the significant differences between the means of the high and low subgroups on each item. The results are shown in

Table 7. Results of the discriminant analysis of the sample data.

No.	t-Value	Sig.	No.	t-Value	Sig.	No.	t-Value	Sig.
TS1	1.945	0.053	GQ16	11.259	0.000	CS31	12.998	0.000
TS2	10.344	0.000	GQ17	13.618	0.000	CS32	12.299	0.000
TS3	10.190	0.000	GQ18	13.540	0.000	CS33	2.562	0.011
TS4	13.041	0.000	ES19	11.380	0.000	CS34	2.826	0.005
TS5	11.989	0.000	ES20	10.082	0.000	IC35	12.820	0.000
CSG6	13.869	0.000	ES21	11.542	0.000	IC36	13.609	0.000
CSG7	12.200	0.000	PS22	11.990	0.000	IC37	12.099	0.000
CSG8	13.557	0.000	PS23	13.311	0.000	IC38	11.384	0.000
SP9	14.295	0.000	PS24	2.803	0.006	TF39	11.598	0.000
SP10	12.725	0.000	PS25	11.210	0.000	TF40	12.228	0.000
SP11	13.847	0.000	IS26	3.669	0.000	TF41	10.965	0.000
SP12	12.847	0.000	IS27	3.183	0.002	EA42	2.310	0.022
PP13	10.294	0.000	IS28	9.205	0.000	EA43	10.101	0.000
PP14	10.518	0.000	CS29	13.462	0.000	EA44	12.063	0.000
PP15	10.435	0.000	CS30	13.219	0.000

. The discrimination of the TS1, PS24, CS33, CS34, and EA42 items in the scale was poor. They were to be deleted, and the remaining items were to be retained.

Secondly, to explore the underlying factor structure and delete some of the less relevant question items, this 39-item scale was submitted to an EFA on dataset 1 (n = 338). The Kaiser–Meyer–Olkin (KMO) value above 0.70 and Bartlett’s sphericity test p-value less than 0.05 indicated that the data were suitable for factor analysis [76]. The data results showed that KMO = 0.937 > 0.7 and Bartlett’s spherical test Sig. = 0.000 < 0.01 met the recommended criteria. The study then analyzed the principal components of the remaining 39 items of the scale. Based on the Kaiser–Guttman criterion (eigenvalues > 1.0), ten public factors were extracted, together explaining 68.91% of the variance.

Generally, when the factor loading of the item is less than 0.4, the cross-loading is more than 0.4, or the common factor is less than three items, the question item should be deleted [75,77]. The study used the maximum variance method for orthogonal rotation. After several explorations, it was found that the factor loadings corresponding to the IS28 measurement questions were below 0.4 in each component column and were to be removed. In this case, the common factor “safe information” contained only two items. Therefore, the common factor and its items were deleted. Convenient service, preferential price, good quality, safe environment, secure payment, comfortable space, and comfortable interaction are consistent with the customer experience evaluation dimensions and items of community unmanned stores constructed earlier. However, clear shopping guide and smooth process were suggested as common factors, which can be attributed to the fact that they are both customer ratings of ease of access to services. Friendly technology and affine environment were suggested as common factors, which can be attributed to the fact that they are both evaluations of the friendliness of the image of community unmanned stores. In addition, the factor loadings of the other nine common factors, except for “safe information”, were all greater than 0.4 in their corresponding dimensions, while they were all less than 0.4 in the other dimensions. The final rotated component matrix was obtained after the exploratory adjustment, as shown in

Table 8. Factor analysis (EFA, promax rotation, n = 338).

Items	Components
	1	2	3	4	5	6	7	8	9
TS2	0.233	0.166	0.062	0.776	0.153	0.071	0.071	0.088	0.096
TS3	0.169	0.109	0.138	0.712	0.103	0.108	−0.017	0.186	0.155
TS4	0.224	0.159	0.201	0.693	0.155	0.110	0.136	0.150	0.065
TS5	0.171	0.157	0.213	0.759	0.074	0.201	0.092	0.075	0.075
CSG6	0.743	0.201	0.061	0.137	0.158	0.049	0.079	0.104	0.139
CSG7	0.764	0.081	0.101	0.207	0.118	0.106	0.110	0.040	0.073
CSG8	0.745	0.142	0.136	0.142	0.061	0.163	0.026	0.080	0.145
SP9	0.763	0.106	0.133	0.112	0.042	0.218	0.125	0.127	0.137
SP10	0.738	0.131	0.135	0.179	0.196	0.009	0.124	0.134	−0.046
SP11	0.764	0.112	0.095	0.064	0.165	0.043	0.112	0.171	0.122
SP12	0.703	0.197	0.091	0.101	0.105	0.125	0.111	0.093	0.123
PP13	0.182	0.112	0.128	0.151	0.197	0.759	0.089	0.086	0.147
PP14	0.119	0.112	0.107	0.180	0.121	0.772	0.062	0.139	0.084
PP15	0.205	0.135	0.139	0.087	0.186	0.754	0.094	0.095	0.045
GQ16	0.239	0.164	0.176	0.087	0.138	0.106	0.01	0.748	0.111
GQ17	0.179	0.122	0.065	0.216	0.121	0.126	0.196	0.779	0.132
GQ18	0.173	0.150	0.133	0.178	0.175	0.122	0.124	0.747	0.151
ES19	0.156	0.148	0.139	0.032	0.189	0.031	0.808	0.130	0.074
ES20	0.169	0.176	0.097	0.108	0.061	0.056	0.786	0.124	0.151
ES21	0.173	0.107	0.182	0.090	0.165	0.177	0.725	0.032	0.171
PS22	0.116	0.193	0.177	0.099	0.199	0.069	0.123	0.135	0.755
PS23	0.166	0.162	0.198	0.199	0.091	0.119	0.145	0.174	0.738
PS25	0.286	0.101	0.121	0.090	0.143	0.117	0.166	0.097	0.718
CS29	0.187	0.204	0.152	0.157	0.641	0.192	0.146	0.066	0.200
CS30	0.197	0.139	0.147	0.114	0.766	0.126	0.139	0.073	0.078
CS31	0.183	0.116	0.163	0.111	0.689	0.138	0.129	0.223	0.179
CS32	0.162	0.170	0.154	0.128	0.741	0.153	0.080	0.131	0.058
IC35	0.145	0.116	0.736	0.145	0.117	0.156	0.109	0.161	0.183
IC36	0.136	0.132	0.691	0.160	0.206	0.114	0.129	0.091	0.218
IC37	0.204	0.122	0.792	0.151	0.072	0.147	0.123	0.095	0.091
IC38	0.123	0.208	0.774	0.137	0.208	0.026	0.108	0.057	0.043
TF39	0.220	0.696	0.075	0.196	0.127	0.159	0.107	0.064	0.064
TF40	0.202	0.729	0.191	0.105	0.194	0.050	0.015	0.097	0.109
TF41	0.089	0.773	0.149	0.067	0.190	0.107	0.085	0.131	0.072
EA43	0.177	0.688	0.046	0.140	0.083	−0.031	0.159	0.131	0.168
EA44	0.146	0.784	0.133	0.103	0.028	0.149	0.133	0.049	0.074

, with the remaining 36 items loaded on the nine factors. To sum up, the study removed three measurement topics through EFA, renamed the common factor containing all the items of “clear shopping guide” and “smooth process” as “smooth transaction”, combined the common factor containing all the items of “friendly technology” and “affine environment” as “friendly image”, and removed the dimension of “safe information”.

Lastly, to further determine the reliability of the scales derived from exploratory factor analysis, a reliability analysis of the remaining 36 items was conducted. The results of the study’s reliability analysis data for the remaining 36 question items are shown in

Table 9. Reliability test results for sample data.

Dimensions	Codes (Variables)	CITC	CAID	Cronbach’s α
Convenient service	TS2	0.711	0.798	0.849
	TS3	0.642	0.827
	TS4	0.682	0.81
	TS5	0.719	0.795
Smooth transaction	CSG6	0.747	0.901	0.915
	CSG7	0.745	0.902
	CSG8	0.739	0.902
	SP9	0.769	0.899
	SP10	0.724	0.904
	SP11	0.750	0.901
	SP12	0.702	0.906
Preferential price	PP13	0.683	0.703	0.806
	PP14	0.638	0.751
	PP15	0.641	0.749
Good quality	GQ16	0.648	0.800	0.829
	GQ17	0.725	0.726
	GQ18	0.689	0.762
Safe environment	ES19	0.695	0.716	0.814
	ES20	0.667	0.745
	ES21	0.639	0.772
Secure payment	PS22	0.664	0.739	0.811
	PS23	0.691	0.709
	PS25	0.631	0.770
Comfortable space	CS29	0.651	0.809	0.841
	CS30	0.702	0.787
	CS31	0.675	0.799
	CS32	0.673	0.800
Comfortable interaction	IC35	0.706	0.824	0.861
	IC36	0.679	0.835
	IC37	0.744	0.808
	IC38	0.704	0.825
Friendly image	TF39	0.671	0.840	0.865
	TF40	0.702	0.832
	TF41	0.714	0.829
	EA43	0.627	0.851
	EA44	0.714	0.829
Overall questionnaire				0.949

. As seen from the data in the table, the composite reliability of the scale is 0.949 > 0.7, which shows that the scale has good reliability. Cronbach’s values for the different dimensional subscales in the scale were 0.849, 0.915, 0.806, 0.829, 0.814, 0.811, 0.841, 0.865, and 0.922, in descending order, indicating good internal consistency and reliability between the measurement items of each dimensional subscale. In addition, each question item’s CITC (Corrected Item–Total Correlation) and CAID (Cronbach Alpha if Item Deleted) values aligned with the requirements; the CITC values were >0.4 and the CAID values were lower than the Cronbach’s alpha values of the scale [75], indicating that the quality of the items of the different dimensional scales was high and no deletion was needed.

4.2. Second Stage: Confirmation of the Factor Structure

For the validation of the nine-factor model that resulted from the EFA, a CFA was conducted on the 338-sample dataset from the other half that was not used for exploratory factor analysis using the software package IBM SPSS AMOS 26.0.

Firstly, the study judged the degree of model fit by absolute fit measures such as CMIN/DF and RMSEA, incremental fit measures such as IFI, TLI, and CFI, and simple fit measures such as PGFI and PNFI. The data showed that CMIN/DF = 1.600 (between 1–3), RMSEA = 0.042 < 0.05, IFI = 0.933 > 0.9, TLI = 0.923 > 0.9, CFI = 0.932 > 0.9, PNFI = 0.743 > 0.5, and PGFI = 0.741 > 0.5, which all met the judgment criteria and showed that the model fit degree was high [78]. Then, the study examined the scales’ internal consistency and convergent validity using standardized factor loading, composite reliability, and average variance extracted.

Table 10. Estimation of loading coefficients and reliability analysis of each indicator on the dimension to which it belongs (*** p < 0.001).

Path			Std.	p	CR	AVE
Convenient service	→	TS2	0.771	***	0.835	0.560
	→	TS3	0.670	***
	→	TS4	0.788	***
	→	TS5	0.758	***
Smooth transaction	→	CSG6	0.737	***	0.902	0.568
	→	CSG7	0.723	***
	→	CSG8	0.786	***
	→	SP9	0.782	***
	→	SP10	0.785	***
	→	SP11	0.725	***
	→	SP12	0.732	***
Preferential price	→	PP13	0.792	***	0.788	0.555
	→	PP14	0.674	***
	→	PP15	0.763	***
Good quality	→	GQ16	0.705	***	0.769	0.529
	→	GQ17	0.653	***
	→	GQ18	0.814	***
Safe environment	→	ES19	0.721	***	0.760	0.515
	→	ES20	0.634	***
	→	ES21	0.790	***
Secure payment	→	PS22	0.719	***	0.751	0.503
	→	PS23	0.630	***
	→	PS25	0.772	***
Comfortable space	→	CS29	0.746	***	0.827	0.545
	→	CS30	0.702	***
	→	CS31	0.726	***
	→	CS32	0.777	***
Comfortable interaction	→	IC35	0.682	***	0.823	0.538
	→	IC36	0.724	***
	→	IC37	0.770	***
	→	IC38	0.756	***
Friendly image	→	TF39	0.643	***	0.851	0.534
	→	TF40	0.738	***
	→	TF41	0.762	***
	→	EA43	0.773	***
	→	EA44	0.729	***

reveals that the 36 items have standardized factor loadings between 0.630 and 0.814 and are significant in the dimension to which they belong, the component reliability being between 0.75 and 0.902, and the average variance extracted being between 0.503 and 0.568, which all meet the criteria proposed by Hair et al. (2010) [79]. It shows that this ECUS-scale is relatively dependable, with good internal consistency and convergent validity.

Finally, to determine the differential validity of different subscales, this study calculated the arithmetic square root of the AVE values of its nine constructs and the correlation coefficients between each construct; the results are presented in

Table 11. Average variance extracted (AVE) and correlation matrix.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	AVE
Convenient service (1)	0.748	-	-	-	-	-	-	-	-	0.560
Smooth transaction (2)	0.301	0.754	-	-	-	-	-	-	-	0.568
Preferential price (3)	0.290	0.170	0.745	-	-	-	-	-	-	0.555
Good quality (4)	0.296	0.231	0.304	0.727	-		-	-		0.529
Safe environment (5)	0.102	0.140	0.186	0.174	0.718	-	-	-	-	0.515
Secure payment (6)	0.285	0.346	0.104	0.348	0.191	0.709	-	-	-	0.503
Comfortable space (7)	0.211	0.223	0.212	0.211	0.178	0.220	0.738	-	-	0.545
Comfortable interaction (8)	0.208	0.267	0.165	0.263	0.160	0.201	0.199	0.733	-	0.538
Friendly image (9)	0.302	0.249	0.275	0.304	0.298	0.395	0.331	0.205	0.731	0.534

. The arithmetic square root of the AVE values on the diagonal in Table 11 for the nine dimensions is greater than the absolute value of the Pearson correlation coefficient (values in the lower triangle) corresponding to each dimension, which meets the criteria suggested by Hair et al. (2010) [79]. It indicates that the ECUS-scale has good discriminant validity.

5. Results

The ECUS-scale was developed based on grounded theory analysis, which identified key dimensions of customer experience in community unmanned stores. In the exploratory phase, twelve dimensions and 48 initial attributes were identified. Following content validation and exploratory factor analysis (EFA), several items were removed or revised based on empirical data and their alignment with consumer experiences in unmanned retail environments.

Specifically, dimensions such as safe information were discarded due to insufficient factor loadings, while others, such as clear shopping guide and smooth process, were combined into the smooth transaction dimension. Similarly, the friendly technology and affine environment dimensions were merged into the friendly image dimension. Ultimately, the scale was refined to 36 items across nine core dimensions: convenient service, smooth transaction, preferential price, good quality, safe environment, secure payment, comfortable space, comfortable interaction, and friendly image.

The specific measurement items for each dimension can be found in Appendix A. These dimensions represent the key aspects of customer experience in community unmanned stores, combining both technological and emotional elements that shape consumer engagement in this unique retail setting.

6. Discussions and Conclusions

This study aimed to explore consumer experiences in community unmanned stores, with a focus on identifying and conceptualizing the key dimensions of experience and developing a reliable scale for measurement. In response to RQ1, the study identified nine core dimensions of consumer experience: convenient service, smooth transaction, preferential price, good quality, safe environment, secure payment, comfortable space, comfortable interaction, and friendly image. These dimensions provide a comprehensive framework for understanding consumer experience in community unmanned stores, where digital technology, physical space, and emotional–social factors are intertwined.

Regarding RQ2, the development of the ECUS-scale effectively captures the consumer experience in community unmanned stores by measuring both technological and physical aspects. The scale was developed through a rigorous process that included grounded theory analysis, expert validation, and exploratory factor analysis. The final ECUS-scale comprises 36 items across nine dimensions, providing a comprehensive tool for assessing consumer experience in community unmanned stores.

6.1. Theoretical Contributions

This study provides theoretical contributions by addressing a gap in the literature on customer experience in community unmanned stores, a novel retail model that merges online–offline convergence, intelligent automation, and local community proximity. Similar to findings from previous studies on traditional retail, this research reaffirms that even in community unmanned stores, customer expectations continue to revolve around foundational service qualities such as convenience [47,80], preferential price [80,81], good quality [47,80] and the esthetics and comfort of the physical space [47,48,49]. These dimensions reflect the basic needs that consumers expect retail formats to fulfill.

With the introduction of digital retail technologies, many studies have focused on aspects such as perceived ease of use and system efficiency [48,49]. Building on this foundation, our study extends the discussion by emphasizing smooth transaction, a dimension that reflects the elevated expectations for seamless technological performance in environments where most interactions and services are automated. Although remote assistance is sometimes available, customers primarily rely on technology to complete tasks independently, making transaction fluency critically important. Furthermore, as seen in service quality scales such as the retail service quality scale [47], Integrated Store Service Quality Measurement Scale [48], and the omnichannel retail service quality scale [49], the role of employees is traditionally central to perceived service quality. In community unmanned stores, however, intelligent technologies take over many of these service functions. As a result, consumer expectations of human-like responsiveness and support are transferred onto the technology itself—an evolution this study captures through the lens of digital interaction quality. In this context, the importance of natural interaction with technology—one that mimics human engagement—is emphasized as a key element in enhancing the experiential quality of unmanned service encounters.

In addition, this study identified safe environment as a distinct and underexplored dimension of customer experience in unmanned retail. While earlier studies acknowledged financial security [47,48,49] and information privacy [49,70], this research also explicitly highlights the importance of secure payment as a standalone dimension. This emphasis may stem from consumer concerns about the reliability of unmanned payment systems and the difficulty of obtaining immediate assistance when payment errors occur. However, few studies have examined the issue of physical safety in environments without staff presence. In community unmanned stores, where customers often shop alone, the lack of human supervision can make some feel vulnerable. To mitigate this, community unmanned stores must adopt alternative strategies such as surveillance systems and clear spatial design to enhance perceived safety. This research brings attention to this overlooked dimension and argues for its inclusion in future retail experience frameworks.

Beyond technological expectations and physical safety, emotional and social dimensions also play a pivotal role in shaping customer experience. In the retail brand experience scale [69], the importance of emotional event experiences was particularly emphasized. While the ECUS-scale developed in this study does not explicitly measure brand experience, it similarly foregrounds the affective and relational aspects of customer experience—an aspect often underrepresented in earlier research. In traditional neighborhood retail settings, consumers frequently engage in brief, familiar interactions with store employees, fostering a sense of warmth and belonging. In contrast, community unmanned stores lack such human contact, making emotional engagement even more critical in cultivating customer satisfaction and loyalty. This study further highlights the role of friendly image as a dimension that reflects how unmanned stores can embody the cultural and social atmosphere of their surrounding communities. Since each community may carry its own rhythm, values, and lifestyle patterns, community unmanned stores are often more closely embedded in consumers’ daily lives than standardized chain supermarkets. The sense of familiarity and place-specific identity embedded in these stores enhances their relevance to local consumers.

In sum, this study contributes to advancing experience research by bridging technological, spatial, and emotional–social aspects in unmanned retail settings. It offers a timely conceptual response to the rise of community proximity-based, omnichannel digital retail by extending the literature on customer experience in smart, low-contact environments. Conceptually, the ECUS-scale is positioned at the intersection of customer experience theory and digital retail transformation, offering a context-specific instrument for capturing experience in community unmanned stores—hybrid retail spaces that fuse intelligent automation with local embeddedness. Unlike traditional service quality or experience scales developed for physical or omnichannel environments with human service agents, the ECUS-scale addresses the unique experiential logic of low-contact, community-based digital commerce. It is particularly suited for evaluating customer experience in retail scenarios characterized by minimal staff interaction, spatial proximity to consumers’ community, and technology-mediated service delivery.

However, while the ECUS-scale captures the core experience dimensions during the in-store encounter, it does not currently assess long-term outcomes such as trust, repurchase intention, or customer lifetime value. Future research may consider extending the scale to broader smart retail environments or adapting it for use in other AI-intensive service contexts.

6.2. Practical Implications

From a practical standpoint, this study provides valuable insights and actionable tools for businesses operating or planning to implement community unmanned stores. The development of the ECUS-scale offers retailers a comprehensive, reliable, and flexible instrument for assessing and enhancing customer experience in these stores. With nine key dimensions—ranging from convenient service to friendly image—the scale allows businesses to evaluate the multiple aspects of customer experience in a nuanced and detailed manner.

One of the key strengths of the ECUS-scale is its modularity. Retailers can tailor their approach to address specific strategic needs by focusing on particular dimensions based on their operational priorities. For instance, if security concerns are a priority, businesses can use items related to the safe environment and secure payment dimensions to evaluate and improve customer perceptions of safety. Similarly, if the goal is to improve emotional engagement with consumers, elements related to friendly image and comfortable interaction can be prioritized. This flexibility enables businesses to target particular areas for improvement, ensuring more efficient resource allocation and a more personalized customer experience.

The scale also has significant implications for community unmanned store design. Understanding the importance of both physical and emotional–social aspects allow retailers to optimize store layouts, technology interfaces, and customer service strategies. For example, the comfortable space dimension highlights the importance of creating a pleasant and inviting physical environment, while smooth transaction emphasizes the need for seamless, efficient technological interfaces that enhance the shopping experience. By addressing both physical and technological components, retailers can create an environment that not only meets the practical needs of consumers, but also fosters positive emotional connections.

Moreover, this study provides guidance for customer relationship management (CRM) in the context of unmanned retail. By using the ECUS-scale to regularly measure and track customer experiences, retailers can identify trends in consumer satisfaction and engagement over time. This allows for continuous experience optimization, helping businesses stay competitive in an increasingly digital and community-focused retail landscape.

Overall, the ECUS-scale serves as a diagnostic and strategic tool that not only provides insights into how customers experience community unmanned stores, but also helps businesses make data-informed decisions to improve overall service quality and consumer satisfaction. The ability to focus on specific dimensions in response to customer needs gives retailers the flexibility to refine their offerings in line with evolving consumer expectations in the digital retail space.

6.3. Limitations and Suggestions for Future Research

Despite its contributions, this study is subject to several limitations that should be acknowledged. First, the sample was primarily composed of younger consumers, with relatively limited representation of elderly individuals and children. This demographic skew may restrict the generalizability of the findings across diverse age groups. Second, although the study achieved initial validation of the ECUS-scale, the overall sample size was modest, which may influence the robustness of certain statistical results. Third, all data were collected within a single cultural context—China—which may limit the cross-cultural applicability of the scale. Since customer experience is influenced by cultural values and social norms, variations in perception and behavior may arise in other countries or regions. Finally, while this study proposes a comprehensive measurement instrument, it does not yet assess the relative importance or weighting of each experience dimension in shaping customer outcomes.

To build on the current research, future studies are encouraged to adopt a more inclusive sampling strategy that incorporates vulnerable or underrepresented groups, such as children, the elderly, and individuals with disabilities. This is particularly important as urban communities evolve toward more age-friendly and inclusive service environments. Comparative cross-national studies could also be conducted to examine the scale’s cross-cultural validity and contextual adaptability. Additionally, future research may apply hierarchical analytical methods (e.g., analytic hierarchy process) to assign weights to each dimension of the ECUS-scale, thereby prioritizing key experience components. This would enable firms to allocate resources more efficiently when designing or optimizing community unmanned store experiences. Finally, longitudinal research could be valuable in tracking how customer experience evolves over time with repeated exposure to intelligent retail environments and changing digital marketing strategies.