Student retention in higher education might be prioritised by funding authorities and universities but robust measurement of non-completion is elusive. This investigation explores untapped data sources to enrich understanding of non-completion. The analysis features the main undergraduate course in a part of a large UK university with retention concerns. Three data sources were explored; Registry, course administration and electronic information systems. Discrepancies emerged in data from them, with students' electronic engagement data revealing withdrawal patterns not evident in the other sources. The departures profile revealed by the information systems data is used to conjecture association between type of departure and stage of departure drawing on time-based retention models. The analysis also draws on material from focus groups of course administrators and tutors. The results demonstrate a new way of pinpointing and quantifying non-completion over time, electronic footprint analysis, which provides an additional basis for formulating and monitoring retention strategies. While not explaining why individuals withdraw, they reveal more about when they appear more likely to do so. Using data already captured they constitute feasible means of improving retention. (Contains 2 figures and 1 table.)