Recently I have had a close encounter with Aggregating Cost of Poor Quality (COPQ) which was simply incorrect.
The case being discussed here is a scenario where COPQ is simply Rework effort / Total Effort. COPQ at project level, Portfolio level, Line of Business, & Business Unit was arrived at by Sum of All Rework (@ that level) / Total Effort (@ that level).  All was well until it was time to infer and make decisions.

Comparisons were made between groups, and teams with too low & Too high values had to explain the reasons. Over a period of time this led to the wrong behavior of managing the Effort Data entries rather then looking for ways to minimize rework & Poor Quality costs. Variations were mostly attributed to incorrect Data collection (on Rework) rather than real contributing factors to Poor Quality.

A more appropriate & meaningful representation is to calculate the COPQ only at the project level. All aggregations should be about % of Projects that meet / do not meet certain criteria. Hence the Reporting of COPQ at different levels would simply state % of projects with high Rework / Moderate Rework or low Rework. The Management systems should be tuned to monitor this and take corrective & Preventive actions.

This is not the solution, it is only a way forward. Since projects differ by complexity, team Size, Team Competency and other such factors. having a common goal may not suffice. E.g. a goal of 5% COPQ for both the low complexity / Small team  & High Complexity /Big team is not fair.

We will need to work on an approriate metric to that can help reinforce right behaviors and at the same time help identify the right contributing factors of Poor Quality Costs