As information technology has rapidly developed, the online-to-offline (O2O) model has become increasingly popular, and third-party review platforms have continued to play a significant role in the real economy. To mitigate consumer loss aversion due to uncertainty about product quality, new start-up companies rely on third-party review platforms to conduct product trials and collect ratings from trialists. However, some companies also pay higher costs to conduct fake trials by hiring the Internet “water army” and other means. Although government departments, market regulators, and third-party review platforms have attempted to curb fraudulent ratings, rating fraud persists because it is undetectable and hugely benefits fraudulent companies. Some third-party review platforms have adopted a rating display threshold mechanism by which the platform discloses a rating only when the number of ratings accumulated by the company reaches the threshold. This mechanism has been considered to inhibit rating fraud. In this study, a model for new start-up companies to conduct product trials was developed, the characteristics of their optimal trial strategies were explored, whether third-party review platforms can develop rating disclosure mechanisms to curb firm fake ratings was investigated, and the effects of such mechanisms were considered. First, based on consumer loss aversion, a model for the quality disclosure of companies through free trials was developed, and then trial and pricing strategies were analyzed under different consumer perception scenarios. The results showed that the optimal number of trials conducted by honest companies increased as consumer loss aversion and initial consumer cognitive ambiguity increased but decreased as unit trial costs decreased. Next, based on this model, the possibility that companies would conduct fake trials was considered, a model was developed according to which companies could conduct both real and fake trials, and then trial strategies were analyzed. The results showed that the optimal strategy for enterprises was to purchase either all of the true trial ratings or all of the false trial ratings, but not both ratings simultaneously.Second, a review platform rating display threshold mechanism that has been applied in practice was examined. The results showed that in a monopoly scenario of a single company, the rating display threshold mechanism was simple and effective if the market size was small. Moreover, it did not negatively affect honest companies. If the market size was large, simply using rating display thresholds to inhibit rating fraud forced honest companies to provide an excessive number of trials, thus negatively affecting their interests. Based on the current movie industry, a numerical experiment was designed to apply the results derived from the above theoretical scenario. Finally, the previous model was extended based on the Hotelling model to consider the decision-making problem encountered regarding the purchase of trial ratings by two enterprises in competition. The effects of the rating display threshold mechanism of the review platform were also examined. The results of the theoretical analysis and numerical experiments showed that the use of rating display thresholds alone to inhibit rating fraud tended to negatively affect the interests of honest companies. To avoid these effects and improve social welfare, government regulators, industry associations, and third-party review platforms should cooperate to jointly inhibit rating fraud by reducing its benefits, increasing its costs, and developing an appropriate rating disclosure mechanism.In short, product trials and potential rating fraud are important features of review systems. In practice, third-party review platforms have applied rating display threshold mechanisms, but research on their effects is insufficient. The main contributions of this paper include exploring an optimal strategy for firms to conduct real product trials, providing a theoretical justification for third-party review platforms to inhibit firm rating fraud through rating display thresholds, and indicating the mechanism′s potential damage to honest companies and corresponding coping strategies. The present study could be extended by considering that rating information will continue to accumulate as the number of customers increases after a firm opens and by investigating how rating platforms can continually manage fraudulent ratings.