Timing is everything: Segmenting OTA's cancellations behavior through the BCR model
International Journal of Hospitality Management, vol.135, 2026 (SSCI, Scopus)
- Publication Type: Article / Article
- Volume: 135
- Publication Date: 2026
- Doi Number: 10.1016/j.ijhm.2026.104605
- Journal Name: International Journal of Hospitality Management
- Journal Indexes: Social Sciences Citation Index (SSCI), Scopus, Hospitality & Tourism Complete, Hospitality & Tourism Index, Psycinfo
- Keywords: Bayesian Network analysis, BCR segmentation model, Cancellation Behavior, OTA booking cancellations, Risk Perception, Signaling Theory
- Akdeniz University Affiliated: Yes
Abstract
This study examines consumer risk perception and cancellation behaviour signals in the hotel industry by analyzing OTA reservation cancellations. Customer behaviors were clustered using the BCR segmentation model, consisting of booking window (BW), cancellation window (CW), and risk window (RW) variables. Five distinct segments were identified using Mini-Batch K-Means clustering: Moderate Risk Planners (8 %), Strategic Cancellers (23 %), Impulsive Cancellers (58 %), Risk-Averse Early Planners (7 %), and Long-Term Cancellers (5 %). A key finding highlights a significant paradox within the largest segment, “Impulsive Cancellers” (58 %): this group exhibits spontaneous decision-making with short temporal windows, yet paradoxically has the highest non-refundable booking rate (21.3 %). Bayesian Network and Markov Blanket analyses revealed probabilistic dependencies between temporal parameters and other booking variables, with BW acting as a primary driver influencing subsequent cancellation behaviors. The BCR model provides hoteliers with a data-driven method for analyzing cancellation behavior, while the risk window concept offers a new analytical perspective for minimizing revenue losses and understanding customer risk tolerance.