Examining the characteristics and effectiveness of online employee reviews

Abstract

Employee reviews on platforms like Glassdoor and Indeed significantly influence organizational attractiveness of millions of prospective applicants. To deepen our understanding of this phenomenon, we examined the effects of employee review characteristics on perceived helpfulness – a proximal indicator of adopting shared information. Specifically, we investigated the relationship between the sentiment of organic Glassdoor reviews (ranging from positive to negative attitudes) and their helpfulness ratings. Additionally, we explored the moderating roles of overall corporate ratings and employee status in shaping the impact of employee reviews.

Employing automated text analysis with Latent Dirichlet Allocation (LDA) and Structural Topic Modeling, we further delved into employee review content to extract the topics discussed and how their attributes (e.g., the extent to which the topic is discussed) influence perceived helpfulness. Drawing insights from an extensive analysis of 24,687 Glassdoor reviews, our findings revealed that negative reviews of lower rated organizations tend to receive higher helpfulness ratings, particularly when provided by former employees. The topics identified through LDA encompassed both instrumental and symbolic aspects of organizations, with their extent of discussion uniquely interacting with sentiment. Our study sheds light on the profound impact of employee satisfaction on the perceived helpfulness of online reviews. By presenting a comprehensive analysis of online reviews, this research offers valuable insights for businesses to enhance their organizational attractiveness and better understand the dynamics of online reputation management.