Seeing Is Believing? Visual Evidence in Negative Online Reviews and Subsequent Sales-Rank Dynamics: A Matched-Event Research Prototype

Authors

Huili Wang School of Modern Business Technology, Wuhan Business and Trade Vocational College Author
Hui Xu School of Intelligent Economics, Huanggang Vocational College of Science and Technology, Huanggang, Hubei, China Author

electronic word of mouth; negative reviews; visual evidence; consumergenerated content; sales rank; matched-event design; difference-in-differences

Purpose: This study develops a matched-event framework to assess whether negative

online reviews containing images or videos are followed by larger changes in product sales

rank than comparable text-only reviews. Design/methodology/approach: A constructed

prototype dataset comprises 414 focal negative-review events (207 matched visual–text

pairs) across three product categories. The analytical workflow combines difference-in

differences, event-time comparisons, negative-binomial helpfulness models, and placebo

checks. Findings (illustrative, not platform-validated): In the constructed data, the visual

by-post estimate is 0.074 log points, equivalent to a 7.7% increase in numerical sales rank

relative to the text-only group; the simulated event-time difference peaks near day +7. Originality/value: The paper

distinguishes the evidentiary modality of negative reviews from review valence and proposes a replicable bridge be

tween review helpfulness and market-response research. Research limitations: All quantitative findings are generated

from a constructed dataset and cannot substantiate an empirical claim about any real e-commerce platform. External

empirical conclusions require independently sourced observations, documented matching and sampling decisions, and

reproducible estimation.