Scholarlink

<p>This study examines customer review impact on jello shot sales, leveraging a natural experiment from production changes in the U.S. beverage market. Analyzing over 15,000 reviews using Natural Language Processing (NLP) across NRC Emotion, Bing, and AFINN lexicons, the research identifies significant correlations between review valence and sales. Specific sentiments, particularly anger, disgust, joy, and trust, emerge as key sales predictors, with negative emotions exerting a disproportionately strong influence. While star ratings correlate positively with sales, platform biases render them less reliable predictors than nuanced sentiment. Generalized linear modeling confirms review sentiment affects sales magnitude and direction. Findings advance literature on review valence (Chevalier and Mayzlin in J Market Res 43(3):345–354, 2006) and negativity bias (Baumeister et al. in Rev Gen Psychol 5(4):323–370, 2001), highlighting the need for granular sentiment analysis beyond aggregate ratings (Mudambi and Schuff in MIS Q 34(1):185–200, 2010a). Methodologically, it demonstrates NLP’s utility in linking sentiment to sales outcomes (Cui et al. in Int J Electron Commer 17(1):39–58, 2008). For practitioners, results advocate for sentiment monitoring to inform customer-centric strategies. For academics, it provides empirical grounding for consumer behavior and eWOM theories, considering platform effects.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Review valence impact on jello shot sales

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月4日

查看全文
  • James Christopher Westland

摘要

This study examines customer review impact on jello shot sales, leveraging a natural experiment from production changes in the U.S. beverage market. Analyzing over 15,000 reviews using Natural Language Processing (NLP) across NRC Emotion, Bing, and AFINN lexicons, the research identifies significant correlations between review valence and sales. Specific sentiments, particularly anger, disgust, joy, and trust, emerge as key sales predictors, with negative emotions exerting a disproportionately strong influence. While star ratings correlate positively with sales, platform biases render them less reliable predictors than nuanced sentiment. Generalized linear modeling confirms review sentiment affects sales magnitude and direction. Findings advance literature on review valence (Chevalier and Mayzlin in J Market Res 43(3):345–354, 2006) and negativity bias (Baumeister et al. in Rev Gen Psychol 5(4):323–370, 2001), highlighting the need for granular sentiment analysis beyond aggregate ratings (Mudambi and Schuff in MIS Q 34(1):185–200, 2010a). Methodologically, it demonstrates NLP’s utility in linking sentiment to sales outcomes (Cui et al. in Int J Electron Commer 17(1):39–58, 2008). For practitioners, results advocate for sentiment monitoring to inform customer-centric strategies. For academics, it provides empirical grounding for consumer behavior and eWOM theories, considering platform effects.