Visual servicescape analytics: design style and demand in Airbnb from unstructured data

Purpose This study aims to examine the impact of design style on property occupancy rates in home-sharing platforms using the Stimulus-Organism-Response framework, focusing on its interaction with key service attributes such as host service, price, facilities and location. By integrating unstructured data to identify design style-related variables, this research offers practical insights into how design style shapes occupancy rates. Design/methodology/approach Using a mixed-methods approach, this study analyzes Airbnb data from 562 properties across 28 cities in the United States. It incorporates 5,620 evaluations by human coders of Airbnb property designs and integrates image-based feature extraction to identify design style characteristics based on shape, texture and color. The study also leverages text-based analysis, applying word embeddings to 6,706,913 online reviews to construct a comprehensive design style-related lexicon. Findings Empirical findings indicate that design style plays a significant role in shaping property occupancy rates, with modern-style Airbnb properties achieving higher occupancy rates than classic or traditional-style Airbnb properties. Design style also moderates the effects of host service and price on occupancy rates. Furthermore, image- and text-based features related to design style substantially contribute to occupancy rates. Originality/value This study contributes to the service marketing literature by empirically validating the impact of design style using real-market data from Airbnb. It further identifies nuanced mechanisms through which design style influences the evaluation of other service attributes, such as host service and price. By leveraging unstructured data, including images and text reviews, this research deepens understanding of aesthetic value in digitalized servicescapes and demonstrates its implications for platform-based services.