Alumni tracking surveys are critical for evaluating and improving youth employment training programmes. However, there is little evidence about how to design these surveys, especially in less developed countries. We conducted short message service and Web surveys of alumni of a South African employment training programme. Short message service surveys returned higher response rates and more representative data than Web surveys, albeit at three times the cost. Combining modes increased data quality but may not be warranted, given the additional cost and complexity. This evidence suggests that training programmes can feasibly collect quality, actionable data on their alumni, even with constrained budgets and staff. Copyright (C) 2018 John Wiley & Sons, Ltd.
