Experimental studies on positioning are often analysed using measurement scripts created by copying and editing from previous scripts. This means that such scripts often contain dead code as well as variables and parameters that are no longer needed, which makes maintenance and troubleshooting confusing. Whilst this is already true for the use of the same system in different areas of application, this is exacerbated when modifying scripts for a new system. The problem is the diversity of the data for Radio Frequency (RF) based localization, which is often made manageable by applying a layer model. For example, signal properties are typically used to determine the distance or angle, which in turn is the basis for positioning. The aim should therefore be to keep the measurement scripts as small as possible to focus on the current relevant layer. We achieve this by using an exemplary positioning framework, as well as the measurement database strategy proposed in this paper. Our approach is so generic that it contains the characteristics of the measurement campaign, the underlying measurement data as well as the derived estimates. Thus, it enables to start calculations from a defined point to only concentrate on the significant aspects. This facilitates easy benchmarking of new concepts and enables other research groups to compare their hardware and software.