Leveraging Large Language Models and Machine Learning to Tackle Challenges in E-Commerce Catalog Management and Its Implications on Advertising

Abstract

This paper highlights the challenges e-commerce platforms face in managing large catalogs and how Large Language Models (LLMs) can be leveraged to drive operational efficiency, customer engagement, and personalized digital Ads. By analyzing current research and industry practices, we showcase that well-structured catalogs can significantly reduce cart abandonment rates and improve customer satisfaction, since the vast majority of shoppers consider product content crucial in purchasing decisions. The paper details how LLMs and Machine Learning Models can transform traditional catalog management by automating product classification, optimizing attribute extraction and content accuracy, while enhancing search accuracy. We also present a framework for leveraging enhanced catalog data in advertising through three primary techniques: personalized product recommendations, focused retargeting campaigns, and concurrent dynamic pricing. Our findings suggest that businesses implementing LLM and ML-driven catalog management solutions can achieve superior customer engagement and operational efficiency. Our objective with this paper is to contribute to knowledge on catalog management and suggest some scalable catalog management strategies.