Modern society is increasingly relying on the accumulated opinions of its peers that they can find on the Internet. The contribution of consumers on technology platforms enables easier interaction between like-minded people with common interests, and thus facilitates the decision-making process. Within this technological context, service sector organizations such as tourism and hospitality have to face the challenge of consumer-driven content management. Marketing experts have found a way to take advantage of such interactions, which emphasizes the importance of implementing new knowledge in organizations that will help collect, analyze, interpret and manage online social influences. The subject of the doctoral dissertation research is the qualitative analysis of online reviews of consumers of catering services in Serbia. Compared to numerical ratings of users, text reviews reflect customer satisfaction or dissatisfaction but in a much more detailed way because they contatin more information, and thus gain a realistic insight into real consumer experiences. Identifying the type and importance of determinants of satisfaction and dissatisfaction in consumer reviews according to hotel type (city, mountain or spa hotel) is one of the main tasks of the dissertation. For the puroposes of the research, reviews of hotels and restaurants in Serbia were collected. A combination of qualitative and quantitative methods was used in order to prove the set hypotheses. Qualitative analyzes that were applied are word frequency analysis, review length analysis, sentiment analysis, readability analysis and Latent Dirichlet Allocation (LDA). Among the quantitative methods, multiple regression was used to determine the mutual influence of variables. By analyzing the frequency of words, the words that appeared most often in reviews of hotels and restaurants were singled out. When it comes to hotels, positive reviews featured words that referred to the characteristics services provided in a certain type of hotel and contained more positive descriptive adjectives related to the experience of consumption. In negative hotel reviews, regardless of the hotel type, negative descriptive adjectives and words that indicated the material (tangible) elements of the hotel products appeared more often. In the positive reviews of restaurants, there are also a lot of positive descriptive adjectives, and in negative reviews, the negative aspect of the price of restaurant’s services is emphasized. Although the reviews are negative, there are a lot of positive descriptive adjectives in them, indicating that there were aspects of the services that they were satisfied with. The analysis of the length of reviews showed that in the reviews of both hotels and restaurants, many more words and sentences are used to describe a negative expericence than a positive one. A readability analysis was conducted to determine the average number of years of formal education necessary to understand reviews on first reading. The results of analysis showed that the values of the readability index vary form very low (reviews that are understandable to everyone) to very high (extremly difficult to understand). The average value of the readability index indicates that readers must be in their senior years of high school to understand the text on the first reading. Sentiment analysis analyzed the feelings in the reviews. The range of sentiment values varies from extremely negative to extremly positive sentiments, but the largest number of reviews, both positive and negative, contained neutral and positive sentiments. By analyzing sentiment in restaurant reviews, similar resutls were obtained as in hotel reviews. The range of sentiment values vaires from extremely negative to extremely positive sentiments, and as the rating increases, so does the value of the sentiment. Such results may indicate that, although they were dissatisfied, the user experience was not accompanied by negative feeling, which are often responsible for the spread of negative electronic recommendation. Using LDA, the determinants of satisfaction and dissatisfaction with services in hotels (depending on the type of hotel and category, as well as the type of traveler) and restaurants were isolated. Based on the assumption that the determinants of satisfaction and dissatisfaction differ depending on the type of hotel, category and type of travelers, obtained results partially confirm these assumptions. It was assumed that different determinants influence satisfaction and dissatisfaction with restaurant services, which was partially confirmed. By using multiple regression, the effects of the technical characteristics of reviews (polarity, readability and length) on the ratings and helpfulness of the reviews were tested. The obtained results confirmed the positive impact of sentiment and the negative impact of the length of reviews on user rating of hotel reviews. In the case of restaurants, the assumed impacts were not confirmed. In the case of the influence of tecnical characteristics of hotel reviews on reviews helpfulness, no significant influence was found, while in the case of restaurant reviews, a positive influence of length and a negative influence of sentiment on review helpfulness were found. The results obtained in this dissertation have numerous theoretical and practical implications for the hospitality industry. Since customer satisfaction is an integral part of the hospitality business, the identified determinants of customer satisfaction and dissatisfaction can help hoteliers and restauraters improve their business. Based on the established impact of technical characteristics of review on rating and helpfulness, hoteliers and restauraters can strive to improve the performance of reviews they receive from customers by reducing negative and long reviews by providing superior service.