The Pre-Agile Lightweight Methods era covers the period from approximately 1986 to 2000, during which multiple independent practitioners and teams developed iterative, incremental, collaborative approaches to software development without coordinating with each other. The convergence of these streams at the snowbird-meeting-2001 was possible precisely because the streams had already been flowing separately for years — sometimes decades.
Long Precursors (1957-1985)
The era's formal starting date of 1986 reflects the new-new-product-development-game paper by hirotaka-takeuchi and ikujiro-nonaka, but the intellectual roots extend further back. IBM practiced incremental development approaches in the late 1950s (approximate). Project Mercury — the NASA human spaceflight program — used half-day iteration cycles in 1958 (approximate). Harlan Mills at IBM developed staged delivery approaches in the late 1960s (approximate). These precursors establish that iterative, incremental delivery was not invented by the Agile movement but repeatedly rediscovered by practitioners working against waterfall's constraints.
The 1991 publication of Rapid Application Development (RAD) by James Martin gave the RAD approach a commercial and methodological identity, setting the context in which dsdm would emerge.
The 1986 Landmark: Takeuchi and Nonaka
The new-new-product-development-game paper by hirotaka-takeuchi and ikujiro-nonaka, published in the Harvard Business Review in 1986, is the intellectual origin point that ken-schwaber and jeff-sutherland both cite as an influence on scrum. Takeuchi and Nonaka studied how high-performing Japanese manufacturing teams — in industries outside software — organized their product development. They found self-organizing, cross-functional teams with significant autonomy, moving through overlapping development phases rather than sequential handoffs. They used "Scrum" as a metaphor. The manufacturing insight crossed into software through Schwaber and Sutherland's work in the early 1990s.
The Independent Developments (1993-1999)
Multiple lightweight methods emerged in this period with little coordination:
Scrum (1993-1995): jeff-sutherland developed the first Scrum process at Easel Corporation in 1993, drawing on new-new-product-development-game and other influences. ken-schwaber developed parallel approaches at Advanced Development Methods. They came together to present Scrum at OOPSLA 1995 — the oopsla-scrum-presentation-1995 — which established Scrum as a named, documented methodology in the practitioner literature.
DSDM (1994): The dsdm-consortium was founded in the UK in 1994, creating Dynamic Systems Development Method from the Rapid Application Development tradition. The DSDM approach — timeboxing, business collaboration, timeboxing — would be represented at Snowbird by arie-van-bennekum.
Extreme Programming (1996-1999): kent-beck developed XP while working on the Chrysler Comprehensive Compensation System (C3) project in 1996-1997. xp-explained-first-edition appeared in 1999, giving XP its first comprehensive book treatment. XP's distinctive contribution was pairing a project management philosophy (sustainable-pace, user-stories, planning game, short iterations) with a set of rigorous technical practices: test-driven-development, continuous-integration, pair-programming, refactoring, collective-code-ownership.
Crystal (mid-1990s): alistair-cockburn developed the Crystal family of methods at IBM in the mid-1990s. Crystal's contribution was the insight that different project contexts require different methods — Crystal Clear for small collocated teams, Crystal Orange for larger projects — based on team size, criticality, and other factors.
Feature-Driven Development (1997): Jeff De Luca and Peter Coad developed FDD at the United Overseas Bank in Singapore in 1997. FDD contributed the concept of features as the unit of planning and delivery, and a five-step iterative process.
Adaptive Software Development (1997): jim-highsmith developed ASD, drawing on complexity theory and the notion that adaptive management was required for the inherently complex, unpredictable work of software development.
The Rogue River Meeting (2000)
The rogue-river-lodge-2000 meeting in Oregon in 2000 brought together several of these practitioners — including kent-beck, martin-fowler, ken-schwaber, jeff-sutherland, alistair-cockburn, and others — and established that the independent streams shared common values. This meeting was the direct precursor to Snowbird; it created the social network and established the shared vocabulary that made the snowbird-meeting-2001 possible.
What the Era Established
By 2000, the lightweight methods movement had demonstrated:
These insights were about to be named, codified, and made portable at Snowbird. The snowbird-and-early-adoption era began with these foundations already laid.