"Stretching Agile to Fit CMMI Level 3" (2005) is a paper published through IEEE/ACM documenting Anderson's work at microsoft creating MSF for CMMI Process Improvement — the first agile method with a comprehensive mapping to CMMI (Capability Maturity Model Integration) Level 3 requirements. The paper belongs to the pre-kanban-era of Anderson's career and represents his last significant work before the corbis-kanban-experiment.
The CMMI mapping problem
In 2005, the dominant framing in large software organizations positioned agile methods and CMM/CMMI compliance as incompatible. CMMI Level 3 required documented processes, organizational standards, and formal quality management practices that appeared to conflict with the Agile Manifesto's emphasis on individuals over processes and responding to change over following a plan. Anderson's paper challenged this perceived incompatibility by constructing a formal mapping between agile practices (specifically Microsoft's MSF process framework) and CMMI Level 3 requirements.
The work was significant because it made the case that agile methods could satisfy institutional compliance requirements — a necessary argument for Agile adoption in regulated industries, defense contracting, and large enterprises with governance obligations. This was not a purely academic exercise: Anderson was working at microsoft on the MSF process framework and the CMMI mapping was a practical deliverable for enterprise customers.
Intellectual context
The paper belongs to the same intellectual context as agile-management-for-software-engineering: both are concerned with making agile methods viable in large organizations that have process governance requirements. The CMMI paper addresses the compliance dimension; the 2003 book addresses the economic dimension. Together they represent Anderson's pre-Kanban contribution to enterprise agile adoption.
The work with dragos-dumitriu at Microsoft (referenced in the microsoft-xit-kanban-2004 event) was part of the same period, during which Anderson was developing the process discipline and organizational change thinking that would later inform the kanban-method's change management framing.
Historical significance
The MSF for CMMI Process Improvement work demonstrates Anderson's consistent concern — evident from this 2005 paper through the kanban-book to the kanban-maturity-model — with the problem of organizational adoption: how do agile and lean methods get adopted in organizations that have existing process commitments, governance structures, and compliance requirements? The evolutionary change thesis of the Kanban Method can be read as a generalization of the insight embedded in the CMMI mapping work: that the path to better methods runs through existing institutional structures, not around them.