Conservation of biodiversity and ecosystem services in natural environments requires careful management choices. However, common methods of evaluating the impact of conservation interventions can have contextual shortcomings. Here, we make a call for counterfactual thinking—asking the question “what would have happened in the absence of an intervention?”—with the support of rigorous evaluation approaches and more thoughtful consideration of human dimensions and behavior. We review and contrast different evaluation approaches and highlight the advantages of counterfactual approaches over alternative methods. We also illustrate how even robust estimates of ecological impact can fail to estimate the impact of specific policy interventions. The latter depend importantly on human preferences and responses to regulations and incentives that cannot be captured by studies of ecological impact. We propose specific and practical steps that all evaluations can implement now to immediately improve their credibility and accountability.