From Interest to Intent: The Dual Role of AI-Generated Images in Property Search Decision-Making

This study examines the impact of AI-generated design renderings on consumers’ evaluations of roughcast (i.e., unfinished) properties in the real estate industry. While roughcast properties face inherent presentation limitations due to their lack of interior elements, AI-generated images may overcome this by enhancing information delivery through property presentations. Drawing on the structural versus symbolic information framework and two-stage shopping goals theory, this study investigates how AI-generated images influence consumers’ initial interest and purchase decisions at different evaluation stages. Using data from a randomized field experiment conducted on a leading real estate platform in China (Study 1) and two follow-up online experiments (Study 2), we find that AI-generated images significantly boost initial interest in roughcast properties. However, their impact on purchase decision-making varies with the degree of AI’s creative freedom: lower creative freedom effectively promotes purchase decisions, while higher creative freedom can backfire by increasing perceived gaps in feasibility. This research extends the literature on applying generative AI in marketing and product presentation by revealing how AI-generated images influence consumers in different purchasing stages. It also introduces the concept of AI’s creative freedom and its role in shaping consumers’ perceptions. The findings provide actionable insights for marketers and platform providers to address product presentation challenges and optimize the use of AI in enhancing engagement and decision-making.