Can Online Customer Reviews Help Design More Sustainable Products? A Preliminary Study on Amazon Climate Pledge Friendly Products

TL;DR

This study explores how online Amazon reviews can reveal sustainable design insights by analyzing top reviews across product categories, providing a foundation for future automation using machine learning.

Contribution

It introduces a manual analysis method of reviews to identify sustainability-related design insights and discusses potential automation with NLP and machine learning.

Findings

12-20% of reviews mention sustainability aspects

Comparison of certified and standard products reveals design improvements

Provides a baseline for automating review analysis for sustainability insights

Abstract

Online product reviews are a valuable resource for product developers to improve the design of their products. Yet, the potential value of customer feedback to improve the sustainability performance of products is still to be exploited. The present paper investigates and analyzes Amazon product reviews to bring new light on the following question: ``What sustainable design insights can be identified or interpreted from online product reviews?''. To do so, the top 100 reviews, evenly distributed by star ratings, for three product categories (laptop, printer, cable) are collected, manually annotated, analyzed and interpreted. For each product category, the reviews of two similar products (one with environmental certification and one standard version) are compared and combined to come up with sustainable design solutions. In all, for the six products considered, between 12% and 20% of the reviews mentioned directly or indirectly aspects or attributes that could be exploited to improve the design of these products from a sustainability perspective. Concrete examples of sustainable design leads that could be elicited from product reviews are given and discussed. As such, this contribution provides a baseline for future work willing to automate this process to gain further insights from online product reviews. Notably, the deployment of machine learning tools and the use of natural language processing techniques to do so are discussed as promising lines for future research.