Data collection and the technologies that facilitate it are incredibly popular within sport and fitness. However, there is a void between data collection and action. The problem lies in the fact that users currently collect data, but do not use it effectively to change their behaviours. Existing research on personal informatics, behavioural change and data visualisation theory focuses on data collection and analysis for reflective visualisations. However, the design of visualisations and the impact of predictive visualisations on behaviour change tends to be overlooked. Through an iterative design process using feedback from two user studies, this project has designed and developed a predictive visualisation to examine the usefulness of such a concept in supporting behaviour change, bridging the gap from data collection to action. This project resulted in the positive reception of a predictive bubble plot visualisation that recommends an exercise plan based on their historic data. The resulting design recommendations for data visualisations were produced as a result: gain user trust by producing a transparent, easily understandable visualisation, allow high levels of flexibility, focus on aesthetics, and utilise interactivity. These guidelines can be used to create meaningful data visualisations that users can engage with to make the most of their data and bridge the gap from data collection to action.