From storefront to screen: an in-depth analysis of the dynamics of online for offline retailing

Rapid advances in information technology have transformed retail through models like O2O and omnichannel, which integrate online and offline channels to enhance convenience and customer experience. Despite extensive research on O2O/omnichannel, the emerging online for offline (O4O) model—leveraging online capabilities to augment in-store experiences—remains underexplored. O4O deeply integrates digital data and tools (e.g., analytics, personalization, mobile apps) to tailor offline environments, differing from O2O’s focus on online ordering/reservations for offline services. Notable examples include Amazon Go, Amazon bookstores, Alibaba’s Freshippo, and Musinsa’s linked pickup services. This study applies the Unified Theory of Acceptance and Use of Technology (UTAUT) to investigate determinants of continuance intention and shopping intention in O4O contexts, focusing on performance expectancy, effort expectancy, social influence, facilitating conditions, and the moderating role of consumer innovativeness.

The study grounds hypotheses in UTAUT, which posits performance expectancy, effort expectancy, social influence, and facilitating conditions as core determinants of technology use intention/behavior. In O4O, these constructs are theorized to affect continuance and shopping intentions: - Effort expectancy (perceived ease of use across touchpoints) is expected to positively influence continuance and shopping intentions (H1a, H1b), supported by prior findings in mobile shopping and e-commerce. - Performance expectancy (perceived usefulness) is hypothesized to positively affect continuance and shopping intentions (H2a, H2b), aligning with research linking usefulness to continued use and purchase intentions. - Social influence (perceived expectations of important others) is posited to increase both continuance and shopping intentions (H3a, H3b), consistent with m-commerce adoption literature. - Facilitating conditions (perceived availability of resources/support) are expected to enhance continuance and shopping intentions (H4a, H4b), per studies on shopping app usage. - Innovativeness (openness to new products/experiences) is incorporated as a moderator of the effects of UTAUT constructs on continuance intention (H7a–H7d), with direct effects on continuance and shopping intentions also examined (H5a, H5b). The model also posits continuance intention positively affects shopping intention (H6). The review also distinguishes O4O from O2O/omnichannel and motivates the need to extend UTAUT with consumer innovativeness in this context.

Design: Cross-sectional survey of consumers familiar with O4O platforms. After removing insincere/unreliable responses, the final sample was N=272. Sampling/demographics: Gender: 35.3% male, 64.7% female. Age: 20s 39.7%, 30s 27.9%, 40s 22.8%, 50s 7.0%, 60s 2.6%. Education: High school 23.9%, Undergraduate 61.8%, Graduate 14.3%. Analysis approach: PLS-SEM using SmartPLS was chosen due to the emerging nature of O4O (facilitating robust analysis with moderate sample sizes), predictive goals, distribution-free assumptions, capacity to handle complex models with moderators, and robustness to multicollinearity. Measurement model validation: Factor loadings ranged 0.771–0.945 (p=0.001), indicating convergent validity. Reliability: Cronbach’s alpha and Composite Reliability (CR) for all constructs >0.70. Discriminant validity established via Fornell-Larcker (square root of AVE > inter-construct correlations) and HTMT <0.90. Model fit: χ²=833.223; NFI=0.837 (near 0.90 threshold); SRMR=0.066 (<0.08). Multicollinearity: VIFs from 1.525 (INO1) to 4.200 (COI2), all <5. Structural model: Bootstrapping with 5,000 subsamples tested hypotheses. Endogenous construct variance explained: R²=73.8% (continuance intention) and R²=71.0% (shopping intention).

Hypothesis test results (Table 5): - H1a (Effort expectancy → Continuance intention): β=0.11, T=2.001, p=0.045, Supported. - H1b (Effort expectancy → Shopping intention): β=0.031, T=0.642, p=0.521, Not supported. - H2a (Performance expectancy → Continuance intention): β=0.209, T=2.933, p=0.003, Supported. - H2b (Performance expectancy → Shopping intention): β=0.161, T=2.570, p=0.010, Supported. - H3a (Social influence → Continuance intention): β=0.168, T=3.158, p=0.002, Supported. - H3b (Social influence → Shopping intention): β=0.210, T=1.570, p=0.117, Not supported. - H4a (Facilitating conditions → Continuance intention): β=0.435, T=5.428, p<0.001, Supported. - H4b (Facilitating conditions → Shopping intention): β=0.030, T=0.458, p=0.647, Not supported. - H5a (Innovativeness → Continuance intention): β=0.068, T=1.388, p=0.165, Not supported. - H5b (Innovativeness → Shopping intention): β=0.079, T=1.305, p=0.192, Not supported. - H6 (Continuance intention → Shopping intention): β=0.497, T=6.187, p<0.001, Supported. Moderation (H7 series, moderator=Innovativeness on paths to Continuance intention): - H7a (Effort expectancy × Innovativeness): β=−0.052, T=0.683, p=0.495, Not supported. - H7b (Performance expectancy × Innovativeness): β=0.085, T=1.131, p=0.258, Not supported. - H7c (Social influence × Innovativeness): β=−0.095, T=1.691, p=0.091, Marginally supported (p<0.10). - H7d (Facilitating conditions × Innovativeness): β=−0.024, T=0.337, p=0.736, Not supported. Model statistics: R²=73.8% for continuance intention; R²=71.0% for shopping intention. Measurement model fit: SRMR=0.066; NFI=0.837; χ²=833.223; reliability and validity criteria satisfied; VIFs 1.525–4.200.

Findings show effort expectancy increases continuance intention but not shopping intention, indicating usability helps retain users but does not directly drive purchases in O4O. Performance expectancy positively influences both continuance and shopping intentions, emphasizing perceived usefulness as a key driver of ongoing use and purchase behavior. Social influence significantly affects continuance intention but not shopping intention, suggesting that normative pressures encourage sustained platform usage without necessarily converting to purchases. Facilitating conditions strongly predict continuance intention, highlighting the importance of resources and support for ongoing engagement; however, they do not directly impact shopping intention. Continuance intention strongly predicts shopping intention, indicating that fostering ongoing use can translate into purchase behavior. Innovativeness does not directly affect intentions and only marginally moderates the social influence → continuance intention path, suggesting that highly innovative consumers may be less swayed by social norms in the O4O context. Overall, results refine the applicability of UTAUT in O4O by differentiating antecedents of continuance versus shopping intentions and underscoring the mediating role of continuance intention for conversion.

This study advances understanding of O4O by integrating UTAUT with consumer innovativeness to explain continuance and shopping intentions. Key contributions include: (1) distinguishing the roles of effort expectancy (retention) versus performance expectancy (retention and purchase) in O4O; (2) extending UTAUT with consumer innovativeness, showing only a marginal moderation of social influence on continuance intention; (3) evidencing that continuance intention is a strong predictor of shopping intention in O4O; and (4) revealing that facilitating conditions drive continuance but not shopping intention, highlighting context specificity. Managerially, platforms should prioritize usability and support to build continuance, and enhance perceived usefulness to drive both continuance and purchases; social strategies may aid retention more than conversion. Future research should probe contextual factors behind the limited role of effort expectancy and facilitating conditions in purchase, examine boundary conditions of innovativeness’ moderation, and further explore post-adoption dynamics in O4O.