The Corbis Experiment Era is the period in which the Kanban Method was invented, tested, documented, and published. It runs from the October 2004 microsoft-xit-kanban-2004 experiment through the May 2010 kanban-book-publication — six years in which Anderson moved from a promising pull system implementation at Microsoft to a fully articulated change management method supported by empirical case studies, metrics, and a theoretical framework.
The Microsoft experiment (2004)
The era begins with dragos-dumitriu's October 2004 request for help designing a pull system for Microsoft's XIT Sustaining Engineering team. Anderson designed a system inspired by eliyahu-goldratt's Drum-Buffer-Rope mechanism, implemented virtually on Microsoft Product Studio. The system worked — the team's performance improved — but Anderson's theoretical framework for explaining why was still primarily TOC-oriented.
The Reinertsen reframing (winter 2005)
The decisive intellectual event of the era was Anderson's winter 2005 conversations with don-reinertsen, who persuaded Anderson to explore virtual kanban systems through the lens of queueing theory. This reframing provided a more rigorous analytical foundation: where TOC explained WIP limits in terms of bottleneck management, queueing theory explained them in terms of the mathematical relationship between utilization, queue length, and cycle time. The two frameworks were compatible but queueing theory was more precise and more directly applicable to the variability-heavy environment of knowledge work.
Corbis (2006-2008)
The substantive development of the Kanban Method occurred at corbis, where Anderson arrived in September 2006 as Senior Director of Software Engineering. Working with rick-garber, corey-ladas (from winter 2007), and dan-vacanti (from May 2007), Anderson designed and evolved a kanban system that, by March 2007, exhibited all the core features of the mature method:
The Corbis system was a live laboratory: problems with the system's performance revealed systemic issues that motivated further refinement, and the metrics generated by the system provided empirical evidence for the method's claims.
Method codification and the Kanban book
The 2007 corbis-kanban-presentation at Reinertsen's LSSC conference was the first public presentation of the Corbis work. It established the connection between Anderson's empirical case study and Reinertsen's queueing theory framework in front of a product development audience. The subsequent years (2008-2010) were devoted to codifying the method — developing the six core practices, the change management principles, and the evolutionary change framework that would be published as the kanban-book in May 2010.
Significance
This era is the intellectual core of Anderson's biography. Every subsequent development — the conference series, kanban-university, the kanban-maturity-model, fit-for-purpose — builds on the foundation established here. The Corbis experiment transformed a promising technique into a method, and the Kanban book transformed a method into a movement.