In the laser-based powder bed fusion of metals (PBF-LB/M) process, part shrinkage occurs as a result of the repeated melting and solidification of metal powder and solid material during manufacturing, leading to thermally induced distortions. To improve the accuracy of the parts, the dimensional deviations are compensated for by using scaling factors, finite element simulations, or data-driven methods based on measurements. PBF-LB/M users often rely on optical measurement systems, such as 3D scanners, to measure the complex structures that are common in additive manufacturing. However, uncertainties in the 3D scan data and local surface errors are reasons for a lack of reliability in the dimensional accuracy assessment. In this study, we measure the positional accuracy and step heights of appropriately designed pin specimens, considering the tolerance fields of a structured light 3D scanner that comply with the required capability indices of measurement system analyses. Surface roughness measurements determine the tolerance fields for the diameter and roundness of the pins. By adjusting the scaling factors, we achieve a 70 % reduction in reproducible, systematic positional deviations, bringing them below the capability threshold of the 3D scanner. The diameter deviations and roundness are also smaller than the tolerance fields. Some of the step height errors are outside the tolerance but are one order of magnitude smaller than the local errors. The results of this study show the potential for improving dimensional accuracy through scaling factor optimization. For users of 3D scanners, it is important to consider the measurement capabilities when evaluating dimensional accuracy to verify the required tolerances.