Facilitating the provision of detailed, deep and useful feedback is an important design feature of any educational programme. Here we evaluate feedback provided to medical students completing short transferable skills projects. Feedback quantity and depth were evaluated before and after a simple intervention to change the structure of the feedback-provision form from a blank free-text feedback form to a structured proforma that asked a pair of short questions for each of the six domains being assessed. Each pair of questions consisted of asking the marker 'what was done well?' and 'what changes would improve the assignment?' Changing the form was associated with a significant increase in the quantity of the feedback and in the amount and quality of feedback provided to students. We also observed that, for these double-marked projects, the marker designated as 'marker 1' consistently wrote more feedback than the marker designated 'marker 2'.