Agricultural statistics on small farms are critical for informing sustainable development policies, but often suffer from selection bias and are time consuming and costly to collect. Less burdensome and reliable methods are needed. We report on a scalable method using a respondent's knowledge about their land, high resolution imagery, and tablet devices to draw spatially explicit plot boundaries. We find the method may work best with respondents that own their plots are farmers, and for smaller plots (<1 hectare). We also find incongruence between survey questions and spatially-derived data, indicating the importance of incorporating spatial data to verify responses about plot characteristics. © 2020 John Wiley & Sons, Ltd.