Aim While species distribution models (SDMs) are standard tools to predict species distributions, they can suffer from observation and sampling biases, particularly presence‐only SDMs, which often rely on species observations from non‐standardized sampling efforts. To address this issue, sampling background points with a target‐group strategy is commonly used, although more robust strategies and refinements could be implemente...