Theory vs. data-driven learning in future e-commerce

Abstract

As personalization, adaptation and persuasion are called for in e-commerce and as computational power increases, a question emerges: should companies use theories to develop and run their e-commerce operations or does mere data based optimization do a better job? We explore different types of computer-based learning methods and present two evaluations of primarily data-driven learning applications that advance e-commerce in the direction of interactive e-selling relationships.