Quantifying the effect of eWOM embedded consumer perceptions on sales: An integrated aspect-level sentiment analysis and panel data modeling approach

This paper proposes a text-analytics framework that integrates aspect-level sentiment analysis (ASLSA) with biascorrected least square dummy variable (LSDVc) - a panel data regression method - to empirically examine the influence of review-embedded information on product sales. We characterize the online perceptions as consumer opinions or sentiments corresponding to the product features discussed within the review. While ASLSA discovers key product features and quantifies the opinions in corresponding content, the LSDVc-based panel data regression analyses the consumer sentiments to explore their influence on product sales. The proposed framework is tested on the mid-sized car segment in India. Our findings suggest that review volume and the sentiments corresponding to the exterior and appearance significantly influence the mid-size car sales in India.

Automated summary

This paper proposes a text-analytics framework integrating aspect-level sentiment analysis with bias-corrected least square dummy variable method to examine the influence of review-embedded information on product sales empirically. The findings suggest that review volume and the sentiments corresponding to the exterior significantly influence mid-size car sales in India.