This study aims to examine how cognitive systems overload (i.e. system feature overload and system complexity) in online travel agencies (OTAs) shapes travelers’ perceived OTA risks (i.e. performance, privacy, psychological and time) and, in turn, their behavioral outcomes (decision postponement and switching intentions). Drawing on cognitive load theory, the authors also test whether time pressure moderates the relationships between cognitive systems overload and perceived OTA risks.
This paper surveyed 604 US OTA users and analyzed the data with partial least squares structural equation modeling (PLS-SEM) (SmartPLS 4) using a two-step procedure (base model and moderation), 5,000-sample bootstrapping and collinearity checks. This paper assessed out-of-sample predictive power (PLS-Predict; CVPAT) and unobserved heterogeneity (FIMIX-PLS; multigroup analysis) (SEM).
Both system feature overload and system complexity significantly increase all four perceived risks and directly raise postponement and switching intentions. All four risks increase postponement; performance, time and privacy risks also increase switching intentions, whereas psychological risk does not. Postponement itself raises switching intentions. Time pressure does not moderate the relationships between cognitive systems overload and perceived OTA risks. Predictive assessments indicate meaningful out-of-sample performance and heterogeneity tests suggest no meaningful segmentation.
This study integrates Cognitive Load Theory with perceived-risk theory in an OTA context, reframing risk as interface-activated and positioning postponement as a pivotal mechanism linking design-induced risk appraisals to switching. The null moderation result delineates an important boundary condition for time pressure.
