The number of women studying science and engineering at undergraduate and postgraduate levels has increased markedly in recent decades. However females have lower retention rates than males in these fields, and perform worse on average than men in terms of promotion and common research metrics. Two key differences between men and women are the larger role that women play in childcare and house work in most families, and the narrower window for female fertility. Here we explore how these two factors affect research output by applying a common ecological model to research performance, incorporating part-time work and the duration of career prior to the onset of part-time work. The model parameterizes the positive feedback between historical research output (i.e. track record) and current output, and the minimum threshold below which research output declines. We use the model to provide insight into how women (and men) can pursue a career in academia while working part-time and devoting substantial time to their family. The model suggests that researchers entering a tenure track (teaching and research) role part-time without an established track record in research will spend longer in the early career phase compared to full-time academics, researchers without teaching commitments, and those who were beyond the early career phase prior to working part-time. The results explain some of the mechanisms behind the observed difference between male and female performance in common metrics and the higher participation of women in teaching-focussed roles. Based on this analysis, we provide strategies for researchers (particularly women) who want to devote substantial time to raising their families while still remaining engaged with their profession. We also identify how university leaders can enable part-time academics to flourish rather than flounder. In particular, we demonstrate that careless application of metrics is likely to further reduce female participation in research, and so reduce the pool of talent available.