Dynamics of Online Review Text: What Changes and What Matters Over Time

Abstract

Using 700,000 online reviews related to five product categories on Amazon, we study the dynamics of review text to identify what is discussed in reviews, how this content changes over time, and what remains persistently important over time. Our goal is to understand if different product aspects, including physical and experiential characteristics, are discussed at different points in time and to quantify this temporal dynamic. We call this aspect fluidity. We also study whether certain aspects become more or less important over time, and how sentiments that are expressed about these aspects evolve over time. We call this preference fluidity. We find evidence that aspect fluidity is rather muted, and more interesting is its gradual evolution with small incremental changes over time. The exception is for a few aspects related to new technologies that inherently change faster, making those aspects much more fluid in reviews. On preference fluidity, we find that there is a clear separation between aspect importance and the sentiment expressed in positive reviews compared to negative ones. Additionally, we find that positive reviews tend to focus on different aspects than negative ones, and they tend to view positively all product aspects. On the other hand, negative reviews tend to be persistently dissatisfied with a subset of aspects over time. We find that a few aspects dominate the content of reviews: relatively few aspects remain persistently important, and these have persistently strong positive or negative sentiment. Our findings have clear implications for organizing review information in ways that help potential customers navigate the ever-changing content of reviews, and that help platforms feature reviews. We highlight aspects that remain persistently important to customers, thereby providing a way to identify important reviews from amongst the many available.

Keywords: online reviews, stakeholders, content mining