摘要：Online platforms operate reputation systems to transmit quality information of products and alleviate asymmetric information. We explicitly model the process of quality inference and a ?rm’s incentive of building reputation through accumulating reviews. A high-quality ?rm has an incentive to make introductory o?ers, that is, setting a low initial price to quickly accumulate sales and favorable reviews. However, noise in reviews and intense competition can entail di?-culty in quality inference and discourage high-quality ?rm from building reputation. Using the data from Zaihang, a consulting service platform, we identify high-quality and low-quality ?rms by historical prices and sales. We provide empirical evidence that intense competition slows down reputation building. Hence, platforms (or regulators) face a trade-o? between promoting competition and the e?ectiveness of reputation system.