The paper discusses the concept of an automated service that aggregates offers on the cost and delivery time of goods from different transport companies. An important part of the service work is the ranking of search results in order to give out offers from the most reliable companies in the first place. At the same time, the ranking of results cannot be based solely on user ratings due to the limited sample of customers and the possible subjectivity of the reviews taken into account. The study presents an algorithm for ranking transport companies based on the analysis of statistical metrics of the company and user reviews. The algorithm allows you to obtain a consistent rating of companies using a posteriori estimate of the probability of truth of hypotheses about positions in the rating.