Loading...
Paper Number
1813
Paper Type
Complete
Description
The proliferation of video data in retail marketing brings opportunities for researchers to study customer behavior using rich video information. Our study demonstrates how to understand customer behavior of multiple dimensions using video analytics on a scalable basis. We obtained a unique video footage data collected from in-store cameras, resulting in approximately 20,000 customers involved and over 6,000 payments recorded. We extracted features on the demographics, appearance, emotion, and contextual dimensions of customer behavior from the video with state-of-the-art computer vision techniques and proposed a novel framework using machine learning and deep learning models to predict consumer purchase decision. Results showed that our framework makes accurate predictions which indicate the importance of incorporating emotional response into prediction. Our findings reveal multi-dimensional drivers of purchase decision and provide an implementable video analytics tool for marketers. It shows possibility of involving personalized recommendations that would potentially integrate our framework into omnichannel landscape.
Recommended Citation
Ghose, Anindya; Li, Beibei; Li, Rubing; and Xu, Kaiquan, "Real-Time Purchase Prediction Using Retail Video Analytics" (2022). ICIS 2022 Proceedings. 6.
https://aisel.aisnet.org/icis2022/data_analytics/data_analytics/6
Real-Time Purchase Prediction Using Retail Video Analytics
The proliferation of video data in retail marketing brings opportunities for researchers to study customer behavior using rich video information. Our study demonstrates how to understand customer behavior of multiple dimensions using video analytics on a scalable basis. We obtained a unique video footage data collected from in-store cameras, resulting in approximately 20,000 customers involved and over 6,000 payments recorded. We extracted features on the demographics, appearance, emotion, and contextual dimensions of customer behavior from the video with state-of-the-art computer vision techniques and proposed a novel framework using machine learning and deep learning models to predict consumer purchase decision. Results showed that our framework makes accurate predictions which indicate the importance of incorporating emotional response into prediction. Our findings reveal multi-dimensional drivers of purchase decision and provide an implementable video analytics tool for marketers. It shows possibility of involving personalized recommendations that would potentially integrate our framework into omnichannel landscape.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
13-DataAnalytics