Grasping in unstructured environments requires highly adaptable and versatile hands together with strategies to exploit their features to get robust grasps. This letter presents a method to grasp objects using a novel reconfigurable soft gripper with embodied constraints, the Soft ScoopGripper (SSG). The considered grasp strategy, called scoop grasp, exploits the SSG features to perform robust grasps. The embodied constraint, i.e., a scoop, is used to slide between the object and a flat surface (e.g., table, wall) in contact with it. The fingers are first configured according to the object geometry and then used to establish reliable contacts with it. This work introduces an algorithm that, given the object point cloud, computes the best pre-grasp gripper configuration from which to start the scoop grasp strategy. Several experimental trials in different scenarios confirmed the effectiveness of the proposed method.