Administering a survey online has a number of practical benefits including lower costs, reductions in paper-use and data entry time, and a faster data collection time period. A significant challenge, however, entails creating and recruiting a relevant sample of respondents with specific characteristics when email lists are unavailable. In this case, we seek to encourage researcher conversation about relevant and effective recruitment strategies by describing the extensive pretesting process we undertook to develop our survey sampling and recruitment strategy for a study of wildfire mitigation behaviors in Western Colorado. Pretesting response rates varied from 3% to 30%, with the inclusion of a US$2 billion in the recruitment envelope yielding the highest response rate. The actual study subsequently used this approach and resulted in a 35% response rate.