Surgical robotics have been used for many years in orthopaedic procedures in the hip and knee. Robots tend to offer high accuracy and repeatability but add increased cost, complexity, time, and workflow disruption. This work outlines the design and development of a surgical robot that navigates using force feedback. Flexible components tether the patient to the robot and reaction loads are measured allowing the robot to “feel” its way around the pre-operative plan. Differences calculated between measured and desired loads are converted to Cartesian corrections that the robot used to navigate. The robot was tested first using simple square paths to test accuracy, repeatability and functionality. A pre-operative plan was established for implantation of the surgical system and allowed the robot to be tested doing a complex glenoid implant path. Finally, a study was performed and compared the robot’s surgical method to current surgical techniques of a trained surgical fellow on shoulder analogs. Based on this study, the robot performed as well as or better than the surgeon in almost every measurement parameter with less than 1 mm of implant placement error in many measurement metrics and less than 2° of implant orientation error in each angular measurement.