Economists are increasingly drawing on text data, including CEO interviews, congressional speeches, organisation mission statements, and scientific publications. Text analysis offers new ways to understand organisational structure, communication, and decision-making. This session introduces practical tools for turning unstructured text into data, from dictionary methods and topic models to modern semantic embeddings. We’ll look at how to represent language numerically, compare approaches, and assess their strengths and limitations. The session also offers guidance on choosing between methods and provides resources for future work. I’ll go into depth on two examples from my own research: team knowledge recombination in innovation, and the emergence of new scientific fields.