Today, online communities in the World Wide Web become increasingly interactive and networked. Web 2.0 technologies provide a multitude of platforms, such as blogs, wikis, and forums where for example consumers can disseminate data about products and manufacturers. This data provides an abundance of information on personal experiences and opinions which are extremely relevant for companies and sales organizations. Subjects of postings can be partly retrieved by state of the art text mining techniques. A much more challenging task is to detect factors influencing the evolvement of opinions within the social network. For such a kind of trend scouting you have to take into account the relationships among the community members. Social network analysis helps to explain social behavior of linked persons by providing quantitative measures of social interactions. A new approach based on social network analysis is presented, which allows detecting opinion leaders and opinion trends. This leads a better understanding of opinion formation. The overall concept based on text mining and social network analysis is introduced. An example is given which illustrates the analysis process.