This document discusses change vector tracking (CVT), a technique for modeling change as a weighted vector to track how changes impact software design and architecture over time. CVT involves periodically reviewing change requests and documenting them as a change vector. This allows teams to identify any accumulating design debt and make informed decisions about refactoring architecture. The document provides an example of how CVT could be applied from the system level down to classes, and also in the context of microservices. Regular CVT meetings are proposed to analyze past changes from source control and future changes from the product backlog to maintain the change vector.