The influence of the online community, professional critics, and location similarity on review ratings for niche and mainstream brands

OWOM (online word of mouth) is sometimes referred to as electronic WOM (eWOM).

An alternative explanation is that as reviewers gain more expertise in the area, their opinions start to mirror or track those of critics (experts), as highly experienced reviewers become de facto experts in their own right. This may not be the result of critics influence but due to the evolution in taste, or the refining of the palate that allows an experienced reviewer to detect product attributes that less experienced reviewers miss. To address this possibility, in our econometric model, we control for the experience of the reviewer which should at least partially capture the evolving taste of the reviewer. In addition, we also include time effects, which controls for overall changes in taste and preferences over time. However, it is important to also show exposure to critics ratings. From www.​beerdvocate.​com, we were able to obtain some evidence that suggests that the reviewers are exposed to critics ratings for the products. First, the website posts the winners of the critics ratings competitions. For example, the following link showcases the winners for 2019: https://​www.​beeradvocate.​com/​community/​threads/​2019-great-american-beer-festival-gabf-winners.​624310/​. Second, the beer reviewers on the online forum discuss the critics ratings. Please see the following example of a reviewer discussing the influence of the rating received by critics during the Great American Beer Festival (GABF): “Setting aside the fact that “best” is subjective, I’m not sure how that’s the same logic. GABF is judged by a panel of judges, and I would hope that they are people with refined palates and extensive beer knowledge. In other words, what they think should carry some weight.”

The Brewer’s Association serves as a leading authority on the beer industry in the U.S. and organizes beer festivals across the world (e.g., The Great American Beer Festival).

The results are qualitatively similar if we use only the overall score.

We, nevertheless, still tested an ordered Probit model as an alternative analysis approach and found no substantive changes in the results. This consistency in results has been found by prior researchers as well (e.g., Goldfarb & Tucker, 2011; Koçaş & Akkan, 2016).

We provide additional details on the Copulas approach in Web Appendix A.

Web Appendix B shows the results from different nested models. The coefficients of the key measures are qualitatively similar. This provides additional robustness to our findings.