Soybean has been the world’s fastest growing crop over the last 15 years. Yet, as an untraditional and unfamiliar crop, soybean requires small farmers to move beyond their traditional production practices and marketing arrangements in order to produce a successful crop. We employ complexity theory to frame soybean’s production and market differences as significant and non-incremental for smallholder farmers, thus making soybean a ‘long-jump’ agricultural technology problem. We consider three estimation strategies using a primary dataset of smallholder women soybean producers in the Upper West region of Ghana. We first employ probit and ordinary least squares (OLS) regression models to understand adoption and performance.