Search engines provide a window into the changing association between websites and keywords across cultures and countries and over time. As such, they offer journalism and news researchers an opportunity to study how search engines, in this case Google, mediate news events and stories online. However, search results are not straightforward to study. Since search results are made in the act of searching and will have to be retrieved from Google Search in real-time, there is a range of different ontological and methodological issues related to this data source. This paper addresses these issues by discussing how factors in the search algorithm can be used proactively to study variations across searchers and in time. The paper identifies various endogenous and exogenous factors in the search algorithm one has to pay attention to and discusses ways to archive search results accordingly. Through a small case study, ways to work with the influence of endogenous factors (keywords, language settings, geo-location, Web history and clicking behaviour) and mitigate the effects of the exogenous factors (experimentation and randomisation) are suggested. Then, a new approach to studying search results is put forward, which builds on purposeful sampling of real-world participants or constructed research profiles. Finally, perspectives for news and journalism scholars in studying algorithmically generated content in a broader context are offered.