Social media offer huge potential for accessing discourses about health and illness, but it can be challenging to develop good search strategies to generate ‘information rich data’. In this case study, Phillip Brooker and colleagues describe the challenge as one of ‘finding needles in haystacks’. They discuss an approach of working with software developers to capture data from twitter on cystic fibrosis through following users, rather than searching for key words, to generate useful datasets: https://methods.sagepub.com/Case/srmpromo/PDuI8U/user-driven-data-capture-twitter-conversation-cystic-fibrosis-keywords