In order to maximize efficiency, should conversation contracts include incentive payments and also be put up for tender? This work uses laboratory experiments to investigate this question. We find that there exists an optimal share of performance payment which yields maximum total stewardship effort and expected environmental outcome. While cost-effectiveness is maximized with the totality of payments linked to outcomes, it comes at the cost of reduced participation. Tendering such contracts yields additional benefits in terms of effort extraction and cost-effectiveness, but these benefits rapidly decline with the share of performance payment. Combining high shares of performance payments with tendering runs the risk of falling far short of the environmental target.