This paper presents a method for using geometric algorithms to characterize CAD models for the purpose of automated design for supply chain. Improvements in computing allow for fast manufacturability analysis of the 3D geometry found in CAD files. For example, designers can determine the percentage of a 3D model that can be machined, or how many cores would be required to produce a sand casting of the model. Traditionally, this kind of information has been used for process planning or reducing cost via design for manufacture. However, market pressures and product complexity cause firms to outsource fabrication to external suppliers. It is therefore necessary to understand how early design decisions will impact the sourceability of a design, which encompasses cost, quality, and lead time in the supply chain. The goal of this research is to use geometric characterizations and production requirements of a conceptual design to automatically predict sourceability, and provide feedback that enables proactive design changes. This paper works toward this goal by providing a correlation analysis of geometry-based metrics of models classified by manufacturing process.