Payments for ecosystem services (PES) can encourage land owners to manage their land in a biodiversity-friendly way. To increase the effectiveness of PES for biodiversity conservation, incentives could be set to reach a minimum size of conservation area and a suitable spatial connectivity between conserved areas. However, little knowledge exists about which incentives most likely generate such landscape patterns since threshold effects and spatial connectivity have been mostly considered separately. Therefore, we present results from a framed field experiment with Indonesian oil palm farmers and compare the effectiveness between two conditional group payment schemes: In the first one, the size threshold payment, payments are made if at least three farmers in the group conserve. In the second one, an agglomeration payment, payments are made if at least three farmers with bordering land engage in conservation. Our results suggest that both PES designs are similarly effective in the absence of communication. Under both, communication increases conservation outcomes in the case of previous successful coordination and is ineffective in the case of previous coordination failure. Yet, for individuals who are reluctant to conserve, communication only increases conservation outcomes under the size threshold payment. We further discuss potential welfare implications of our results.