The Peopleware breakthrough era spans 1987 to 1999 — from the publication of peopleware through the second edition of that book, which confirmed its status as a durable classic. This is the central period of DeMarco and timothy-lister's intellectual achievement: the articulation and empirical grounding of the peopleware-thesis, the dissemination of the coding-war-games findings, and the establishment of the human-factors argument as a permanent fixture of software management thinking.
Peopleware and the central argument
peopleware (1987) is the defining document of this era. DeMarco and Lister argued, on the basis of the coding-war-games-study data, that software project failures are predominantly sociological rather than technical: the problems that actually kill projects — team dysfunction, managerial interference, poor work environments, high turnover — are human and organizational, not algorithmic. This argument reversed the dominant assumption of software engineering, which attributed project failure to inadequate methodology or technology.
The peopleware-thesis has three empirical legs, all grounded in the Coding War Games data: the office-environment-effect (workspace quality predicts programmer performance), flow-and-interruption-cost (interruptions destroy the flow states that generate productive work), and the finding that the best performers cluster in organizations that provide the conditions for effective work. The spanish-theory-of-management and furniture-police concepts are the book's satirical language for the organizational pathologies that prevent those conditions from existing.
Team jell and organizational learning
team-jell — the phenomenon by which stable teams develop a shared identity and performance multiplier — is the positive corollary of the peopleware argument. If bad environments destroy performance, good environments enable a form of collective excellence that exceeds what individual talent alone predicts. The conditions for team jell are the same conditions that management by spreadsheet destroys: stability, autonomy, shared purpose, and freedom from arbitrary reorganization.
organizational-learning-disability — the pattern by which organizations systematically suppress the feedback that would allow them to improve — is the concept that explains why the peopleware argument was necessary in the first place. Organizations that cannot learn from their own failures will continue to make the same human-factors mistakes even after those mistakes have been documented and named.
DeMarco's solo work in this period
why-does-software-cost-so-much (1993) collects DeMarco's essays from this period, extending the peopleware arguments into adjacent territory. the-deadline (1997), DeMarco's business novel, dramatizes the peopleware arguments in narrative form — an unusual but effective vehicle for reaching managers who would not read a business argument book. The novel format allowed DeMarco to show organizational dysfunction in action rather than merely describing it, and to give the peopleware arguments emotional as well as intellectual force.
The second edition and consolidation
The 1999 second edition of peopleware confirmed the book's enduring relevance and updated the arguments for the decade since the first edition. By 1999, the Agile movement was forming, and the second edition arrived as Agile's organizational arguments — co-located teams, reduced bureaucracy, human-centered development — were being assembled from the same intellectual materials that Peopleware had provided. In this same year, DeMarco was elected an IEEE Fellow, marking formal institutional recognition at the peak of his influence. The column human-capital (1998), co-authored with timothy-lister in IEEE Software, extended the era's central argument by framing people as a software organization's primary capital asset rather than a cost to be minimized. The atlantic-systems-guild provided the institutional continuity through this period, and the atlantic-systems-guild-founding in 1983 is the organizational precondition for the era's sustained output.