The paper discusses the use of internet pages as data for social research. Websites deliver process generated data which are shaped by the authors’ concentration on their potential readers. This type of data – not primarily designed for research purposes – offer valuable insights, but there are also shortcomings. Regarding longitudinal analysis the missing date of publication is considered to be problematic. This weakness, like others, can be addressed during data collection or by combining different types of data sources.

First we will generally discuss the characteristics, advantages and disadvantages of websites, compare websites to other types of process generated data and ask how specific problems of this type of data can be solved. Then we will present two examples of research questions that can be addressed by using websites as data. Both examples investigate communication strategies, but differ in their methodological approach, in their specific research objective and in the degree of professionalism of the research subjects.

The first case study belongs to the area of political communication and uses data from internet pages to analyse how unemployed organisations as comparatively weak actors try to influence the public discourse and the political decision making. The second case explores what is the use of websites to study the phenomena of trust in consumer markets. Websites can be seen as one means of firms to present themselves trustworthy and thus foster consumer’s trust in their products. As an example, websites of German wholefood supermarket chains are analysed.