Impact of smart locker use on customer satisfaction of online shoppers in Vietnam

Rapid e-commerce growth has intensified focus on customer experience and satisfaction. Last-mile delivery plays a pivotal role in online shopping satisfaction, with service reliability and timeliness often most valued. Smart Lockers—self-service, 24/7 pick-up solutions—offer benefits such as reduced failed deliveries, cost efficiencies, environmental gains, and increased consumer flexibility. However, prior research has centered on mature markets, leaving a gap in emerging markets like Vietnam and lacking direct examination of how Smart Locker use affects e-customer satisfaction. This study investigates the relationship between Smart Locker use and overall e-commerce customer satisfaction in Vietnam, proposing and testing a model to assess the impacts of convenience, privacy and security, and reliability via perceived value and transaction costs.

Smart Lockers are automated, cloud-enabled, self-service parcel collection systems, applicable across parcels and temperature-controlled goods, and are conceptualized as self-service technologies (SSTs) where customers co-create value. Prior SST satisfaction research identifies functional service quality attributes and perceived value (emotional, functional, social) as key determinants. Consumer theories applied include resource matching (service design can reduce perceived customer effort), perceived value (functional, hedonic, social benefits), and transaction cost theory (lower complexity reduces costs such as search, learning, time). Building on these theories and past work (e.g., Yuen et al., 2019), the study develops hypotheses that Smart Locker convenience, privacy and security, and reliability increase perceived value and reduce transaction costs, which in turn influence satisfaction with Smart Lockers and subsequently overall e-commerce satisfaction. Tabled prior studies on SST satisfaction (e.g., Wang 2012; Boon-itt 2015; Djelassi et al. 2018; Tang et al. 2021) inform construct selection and model pathways.

Design and analysis: Covariance-based structural equation modelling (CB-SEM) was used to test the hypothesized relationships among latent constructs, consistent with a theory-testing objective. The analysis followed two phases: measurement model assessment (EFA, CFA) and structural model evaluation. Measures: A multi-item questionnaire measured seven latent factors with initially 30 manifest variables adapted from prior literature and expert input: (1) Convenience (5 items), (2) Privacy and security (4 items), (3) Reliability (4 items), (4) Perceived value (4 items, covering functional, hedonic, social value; economic value omitted as service is not surcharged), (5) Transaction cost (4 items: search, learning, locating, travel effort), (6) Satisfaction with Smart Locker (4 items), and (7) Satisfaction with e-commerce (4 items). Items used five-point Likert scales (1=strongly disagree to 5=strongly agree). The instrument was translated and piloted with 12 users, and refined via expert panel. Sampling and data collection: Target population was online shoppers in Vietnam who had used Smart Lockers. A total of 500 questionnaires were distributed via three channels over 30 days (March 5–April 4, 2021): 42% at hubs in Hanoi, 18% by phone, and 40% online; 442 valid responses remained after screening. Sample profile: 54.8% male; 76.7% under 26 years; 62.4% students; 78.9% income <10 million VND. The sample size exceeded EFA/CFA/SEM guidelines (Hair et al., 2010; 2014; Myers et al., 2011). Measurement model: After reliability screening (Cronbach's alpha) and EFA (Principal Axis Factoring, Promax rotation), seven factors were retained with adequate loadings (>0.5) and eigenvalues>1. CFA indicated good fit (Chi-square/df=1.884; GFI=0.907; CFI=0.963; RMSEA=0.045). Composite reliability ranged 0.85–0.91; AVE≥0.58 and exceeded MSV, supporting internal consistency, convergent and discriminant validity. Structural model: SEM estimation in AMOS 20.0 showed good fit (Chi-square/df=1.933; GFI=0.905; CFI=0.960; TLI=0.956; RMSEA=0.046). R-squared for endogenous constructs (perceived value, transaction cost, Smart Locker satisfaction, e-commerce satisfaction) were all >0.5.

All hypotheses were supported (p<0.05). Standardized paths (Table 6):

• Convenience → Perceived value: β=0.441 (p<0.001), strongest positive driver of value.
• Privacy & security → Perceived value: β=0.166 (p<0.001).
• Reliability → Perceived value: β=0.337 (p<0.001).
• Convenience → Transaction cost: β=-0.293 (p<0.001), reduces costs.
• Privacy & security → Transaction cost: β=-0.149 (p=0.005), reduces costs.
• Reliability → Transaction cost: β=-0.431 (p<0.001), strongest cost reduction.
• Perceived value → Satisfaction with Smart Locker: β=0.565 (p<0.001).
• Transaction cost → Satisfaction with Smart Locker: β=-0.326 (p<0.001).
• Satisfaction with Smart Locker → Satisfaction with e-commerce: β=0.753 (p<0.001), a strong effect. Model quality: Measurement and structural models demonstrated good fit; endogenous R² values exceeded 0.5, indicating substantial explained variance.

Findings show Smart Locker attributes improve user outcomes by increasing perceived value and decreasing transaction costs, which in turn elevate satisfaction with Smart Lockers and overall e-commerce satisfaction. This extends prior research beyond usage intention to demonstrate that last-mile delivery experiences materially shape the total online shopping experience. Context matters: unlike findings in more mature markets (e.g., China) where reliability dominated perceived value formation, in Vietnam convenience had the strongest effect on perceived value, likely due to limited Smart Locker coverage and fewer provider options, making location and access critical. Emphasizing placement and accessibility of lockers therefore aligns with consumer priorities in emerging markets. The strong linkage from Smart Locker satisfaction to overall e-commerce satisfaction underlines the strategic role of last-mile delivery design in customer experience management across the purchase journey.

This study contributes theoretically by integrating resource matching, perceived value, and transaction cost perspectives to explain how Smart Locker attributes translate into customer satisfaction with both the delivery service and the broader e-commerce experience in an emerging market. Empirically, it identifies convenience, reliability, and privacy/security as key antecedents that act via perceived value enhancement and transaction cost reduction. Managerially, e-commerce firms should integrate Smart Lockers into delivery options and collaborate with providers to optimize hub placement (convenience), ensure system accuracy and uptime (reliability), and maintain transparent data protection (privacy and security) to boost customer satisfaction. Future research should broaden samples across regions and provider ecosystems, incorporate demographic and contextual moderators, and examine post-pandemic behavioral shifts and other emerging last-mile models to generalize and refine the framework.

Key limitations include: (1) a single-country, convenience-focused sample with limited size and coverage, potentially constraining generalizability; (2) limited inclusion of demographic control variables due to pandemic-related constraints on data collection (e.g., interviews restricted to low-infection zones); and (3) omission of COVID-19-induced behavioral changes and their impacts on last-mile preferences. Future studies should expand sample diversity and size, explicitly model demographic and contextual moderators, and assess evolving delivery preferences in post-pandemic contexts.