In the age of the Internet, where misinformation is spread with great ease, trustworthiness of news sources has become an important issue. The aim of this study is to assess how users perceive credibility of different online news interfaces by employing experimental methodology. An online experiment was conducted using the conjoint analysis approach. Participants were recruited to complete a series of surveys, which displayed different news interfaces, each with varying elements (e.g. menu type, presence of the author). Participants scored their perception of the credibility of each interface. Through conjoint analysis, the results were used to identify which element combinations have the strongest effects on credibility perception for different user groups. Findings from the logistic regression on the whole sample revealed that the presence of all elements is important, but has a detrimental impact on trust. Conversely, the k-means cluster analysis identified five distinct user segments based on demographic and conjoint data. Based on these clusters, the study offers recommendations for developing a trustworthy news interface. This study provides valuable insights into how users evaluate credibility of online news interfaces, and contribute to the development of strategies for improving trust in online news.