The pre-Kanban era spans the period from Anderson's Feature-Driven Development work in Singapore (1997) through the microsoft-xit-kanban-2004 experiment — the years in which Anderson developed the intellectual and practical foundations that would eventually be synthesized into the Kanban Method, but before the kanban insight had taken shape.
Feature-Driven Development at United Overseas Bank (1997)
Anderson's work with Jeff De Luca on Feature-Driven Development (FDD) at United Overseas Bank in Singapore in 1997 was his first significant engagement with iterative, feature-oriented software delivery at enterprise scale. The UOB project was one of the early large-scale demonstrations that iterative methods could work in a commercial banking environment. FDD's emphasis on short feature teams, modeling, and feature-centric planning left Anderson with a practical understanding of what iterative delivery looked like at scale — distinct from the XP and Scrum frameworks that dominated early Agile discourse.
IBM, Sprint, Motorola
After the Singapore work, Anderson consulted with major corporations including IBM, Sprint, and Motorola on agile adoption and software engineering improvement. These engagements exposed him to the organizational challenges of introducing iterative methods in large, process-mature organizations with established project management cultures — the problem that would later motivate the Kanban Method's evolutionary change approach.
"Agile Management for Software Engineering" (2003)
The agile-management-for-software-engineering book (2003) is the primary intellectual artifact of this era. It applied eliyahu-goldratt's Theory of Constraints explicitly to software engineering — arguing that software projects should be managed for throughput rather than resource utilization, and that the constraint (the bottleneck) in software development should be identified and exploited rather than balanced. This was Anderson's first book and established the TOC-oriented analytical framework he would later augment with don-reinertsen's queueing theory.
TOC-oriented thinking
Throughout this period, Anderson's primary analytical lens was Theory of Constraints. Where many Agile practitioners were framing software improvement through XP practices (pair programming, TDD, continuous integration) or Scrum ceremonies, Anderson was framing it through throughput accounting, bottleneck management, and pull systems. This made the October 2004 request from dragos-dumitriu — to design a pull system for a constrained offshore team — a natural fit with Anderson's existing intellectual toolkit.