Substitute and Complementary Recommendations in E-Commerce

Problem and Motivation

In today’s fast-paced e-commerce landscape, recommender systems play a crucial role in shaping customer journeys, driving product views, and influencing sales. When consumers browse a product page they often encounter multiple recommendations simultaneously. These typically fall into two categories: substitute recommendations (suggesting alternative products, like other sneakers when viewing a shoe) and complementary recommendations (offering products that pair well with the focal product, like matching apparel). While both types of recommendations aim to guide consumers, their combined effects on product views and sales remain poorly understood.

Aim and Method

Our project seeks to uncover how substitute and complementary recommender systems interact and influence consumer behavior when used simultaneously on e-commerce platforms. To answer this, we will conduct a randomized between-subject experiment within the apparel industry. This study will uncover how consumers respond to various combinations of substitute and complementary recommendations at different stages of their purchase journey — from early-stage product exploration to final purchase decisions.

Expected Insights

Our research is expected to offer valuable insights for various stakeholders:

• For Marketing Managers: We will clarify how combining substitute and complementary recommendations influences both product views and sales. This will help managers design more effective recommendation strategies tailored to specific purchase stages — boosting product engagement without unintentionally lowering conversions.

• For Consumers: Understanding how recommendation systems interact will shed light on how product suggestions are structured. This empowers consumers to navigate online shopping more confidently, recognizing when they're being steered toward alternatives versus complements.

• For Society: As personalized technology becomes more embedded in daily life, our research will highlight the broader implications of algorithmic decision-making. Promoting transparency in recommender systems fosters informed consumer choices and supports responsible e-commerce practices.