Résumé:
Online reviews have become a widespread instrument in the research and adoption phases of a product in many industries. Product evaluations can be interpreted as signals that produce a variety of effects. They can inform the public about the properties and the quality of a product, thus improving its information set. Additionally, they can produce behavioural effects on consumer decisions and economic effects on a firm’s pricing strategies and sales. The goal of the paper is to examine the informational content of quality reviews formulated by experts as well as by other users. In our study, we will focus on the wine industry. Like many other industries, the wine industry is characterised by online reviews providing feedback on wine tourism, wine-related activities, and wine quality. Review websites allow consumers to access aggregate customer opinions, which are easily understandable and communicated through aggregation into stars, bubbles or scores. This benefits consumers in their decision-making process before making purchases. This is especially useful in the purchasing process of an experience good, such as wine, where the product's true quality can only be determined after experiencing it (Nelson, 1970; Shapiro, 1983). In this case, the more information gathered and disseminated, the better the consumer's purchasing decision should be. Thus, the different review websites have revolutionised how consumers make decisions, how corporations respond to feedback, and how they interact with clients. The emergence of review websites has also directly impacted a specific group of market participants: experts. Although the power of experts is still significant, it has diminished and is now more prone to debate compared with crowd-sourced reviews. The interactions between expert and crowd-sourced opinions have become increasingly intricate and are not yet fully understood. For instance, some experts have criticised crowd-sourced reviews, calling them propaganda and refusing to take their informational content seriously. Conversely, consumers increasingly place more trust in fellow consumers' opinions than in a single expert's taste. In various contexts, academia has also demonstrated that group decision-making may be superior or, at the very least, not worse. Malkiel (2015), for example, suggests that groups tend to make better decisions than individuals. Informed discussion within the group can enhance the decision-making process if more information is shared and differing perspectives are considered. This is particularly relevant in the wine market because wine is not a static product but evolves qualitatively over time. Additionally, subjective evaluations may arise due to behavioural or environmental factors affecting the overall quality impression at any given moment. However, this does not imply that experts are no longer necessary. They still have a significant role in providing information and structure to the wine market. Nevertheless, the crowd can provide valuable quality assessments for fine wine and self-regulate who it deems to have the most helpful expert opinions. Additionally, it may improve the functioning of the wine market by tasting wines that expert tasters may not consider worthwhile. This paper analyses these aspects related to crowd-sourced reviews on the wine market using data from CellarTracker for the top 250 producers in the Piedmont region of Italy for all vintages between 1947 and 2013. The Piedmont region presents an interesting case study due to its complexity, with many small producers working on multiple terroirs and relatively lower coverage by experts than Bordeaux. As a result, informational content may be more important and valuable to consumers in decision-making. The first objective is to infer crowd-sourced evaluations for wines from the Piedmont region and understand how these tasters evaluate different vintages, producers, and terroirs in this wine region. The results of a multivariate regression indicate that online ratings are higher for older vintages (1947, 1958, 1961, and 1964). Regarding wine producers, Giacomo Conterno, Bruno Giacosa, Angelo Gaja, Roberto Voerzio, and Bartolo Mascarello receive higher scores. The preferred terroirs include Rocche dell' Annunziata, Bussia, and Martinenga. The second step is to investigate the information content and, thus, the impact of reviewer characteristics, such as
experience and expertise, on score quality and precision. It is hypothesised that reviewers with more experience and expertise may provide more informative content than novices, making their reviews more valuable to the general public. Experience and expertise are measured by the existence and/or length of the tasting note, the number of wines owned, producers or vintages tasted thus far, and the length of time since a taster signed up on CellarTracker or provided the first rating. Three additional characteristics that may influence how reviewers rate wines are also controlled for. First, it is checked if a given reviewer has a more generous or severe approach to wine tasting. For instance, a reviewer may only award scores between 90 and 100 points, never going below 70 points for poor wines or never going beyond 90 points, even if a wine is exceptional. Additionally, we account for the variability of wine ratings by using a multiplicative heteroscedastic model (Harvey, 1976) to better understand the taster-induced variability of a wine rating. Last, we examine whether a reviewer aligns with consensus estimates or has more extreme tastings and ratings. The third step is to examine the relationship between crowd-sourced and expert reviewers. Using data from the Piedmont region has the advantage that, unlike Bordeaux, only one expert, Antonio Galloni of Vinous, can be considered influential and followed by wine enthusiasts seeking information on wines from this region. This study aims first to determine whether there is a consensus between the ratings of Antonio Galloni and the crowd and, then, whether the release of information from Antonio Galloni affects how the crowd perceives and rates different wines. Overall, findings suggest that viewing the difference between an expert and an amateur in a "black or white" way is inappropriate. Wine expertise should be seen as a continuum ranging from pure amateur to recognized expert with an interconnection between all actors. Overall, we show that crowd-sourced reviews and tasting notes may interest consumers in a market where experience goods are traded. Increasing the informational content beyond expert opinions in this opaque and complex region is beneficial and helps consumers make well-informed decisions before purchasing a good. It is also argued that expert opinions do not become obsolete but rather that both experts and the crowd add substance in a complementary way.
SWITCH edu-ID
Administration
