In this thesis I explore the concept of serendipity by looking at if and how it is possible to foster surprising discoveries of books during online searching, similar to such discovery while browsing bookshelves in a bookshop or a library. Serendipity is a word originating in the Persian fairy tale The three princes of Serendip. In 1754 Horace Walpole uses the word for the first time, explaining that in the story the princes are “always making discoveries, by accidents and sagacity, of things which they were not in quest of”. In other words, serendipity is about finding interesting and valuable things not looked for. Looking at “finding” in the context of computer systems and Human-Computer Interaction (HCI), Carr (2015) describes retrieval in three different modes: linear, non-linear and serendipitous. Both linear and non-linear searches have a well-defined goal and the result of a search is in clear correlation with the search intention. Serendipity, on the other hand, has “misalignment of intention and outcome”. This is to say that if we look at the result of finding something as a process of searching for a pre-defined outcome only, serendipity fails as a retrieval process. However, inherent in the word serendipity is the fact that if something of interest is found during the search, it has some sort of value. The approach chosen to study serendipitous search is Research through Design (RtD). The core of this approach is that knowledge can be produced through making. The knowledge emerges while reflecting on making of concrete artefact, as well as how these artefacts support finding answers to the research questions. My aim was to create visual representations, or visual speculations as I call them, which demonstrate how serendipity can be a valuable approach in certain design situations. By doing so, I hope to open up a space for discourse about serendipity in design. I exemplify my approach by showing how I used sketching and visualisation to make visual speculations. Furthermore, I have used reflection, debate sessions and annotated portfolios to capture new insights. Finally I have evaluated the effectiveness of my artefacts in relation to the research questions.