The first public, co-authored presentation of scrum as a software development methodology, delivered by ken-schwaber and jeff-sutherland at the OOPSLA (Object-Oriented Programming, Systems, Languages, and Applications) conference in Austin, Texas in 1995, in the paper "The Scrum Development Process."
Context
jeff-sutherland had been developing Scrum at Easel Corporation since 1993, drawing on new-new-product-development-game by hirotaka-takeuchi and ikujiro-nonaka, as well as on ideas from complexity theory and empirical-process-control. ken-schwaber had been developing similar ideas independently at Advanced Development Methods. OOPSLA 1995 was the first time they presented the methodology together publicly.
The paper
"The Scrum Development Process" described an iterative, incremental approach to software development organized around short cycles with daily coordination meetings, a backlog of work, and a small cross-functional team. The paper articulated the empirical-process-control foundation: transparency, inspection, and adaptation as the mechanism for managing complex work.
The presentation was at OOPSLA's workshop on business object design and implementation, not a main conference track — Scrum was not yet mainstream enough for a plenary presentation.
Significance
This event marks the public emergence of Scrum as a named methodology. Prior to 1995, the practices existed in development at individual companies but had not been synthesized into a named, transferable framework. The OOPSLA presentation established Schwaber and Sutherland as the co-inventors of Scrum and began the process of spreading the methodology through the object-oriented and software engineering communities.
Six years later, ken-schwaber and jeff-sutherland both attended snowbird-meeting-2001 and signed the agile-manifesto, representing Scrum in the broader convergence of lightweight methods.
The paper was later published as agile-software-development-with-scrum in 2001, providing the first book-length treatment.
Gap
The exact date within October 1995 is approximate. The full text of the original OOPSLA paper and its precise relationship to the subsequent book are not fully documented here.