Structured Analysis and System Specification (1978), published by yourdon-inc's yourdon-inc imprint under edward-yourdon, is DeMarco's solo contribution to the structured-methods-era and the book that established his reputation in software engineering before the humanistic turn of peopleware. It introduced the data flow diagram (DFD) as the primary notation for structured analysis and provided the first comprehensive methodology for applying structured techniques to requirements specification.
The structured analysis methodology
DeMarco's central contribution was a systematic approach to decomposing and documenting system requirements using graphical notation. The data flow diagram — showing processes, data flows, data stores, and external entities — became the canonical tool of structured analysis in the late 1970s and 1980s. The book provided not only the notation but a process for creating and validating DFDs, including rules for leveled decomposition (the ability to expand any process into a more detailed sub-diagram) and data dictionary conventions for documenting the contents of data flows.
The structured analysis approach addressed a real problem of the era: requirements documents were typically dense prose specifications that were difficult to validate with users, ambiguous in meaning, and nearly impossible to translate mechanically into design. DFDs were intended to be readable by non-technical stakeholders, precise enough to guide design, and structured enough to expose inconsistencies and gaps.
Context within the structured methods tradition
The book positioned DeMarco within the structured-methods-era alongside edward-yourdon and Larry Constantine (whose larry-constantine structured design work preceded it) as the leading advocates for systematic, notational approaches to software development. The Yourdon-DeMarco DFD convention became the dominant flavor of data flow diagramming, distinct from the Gane-Sarson convention that was also in circulation.
The intellectual ambition of the book was to bring engineering discipline to the front end of the software process — the part most notorious for chaos, miscommunication, and late specification changes. In this sense it was continuous with the broader software engineering movement's goal of applying engineering principles to programming. DeMarco's later software-engineering-an-idea-whose-time-has-come-and-gone essay can be read as a retrospective judgment on how well this ambition was realized.
Legacy and transition
Structured Analysis and System Specification remained in use well into the 1990s and influenced several generations of systems analysis training. CASE tools of the 1980s and early 1990s were largely organized around DFD-based methodologies. The Unified Modeling Language's use case and data flow notations are distant descendants of structured analysis conventions.
The book also marks the beginning of DeMarco's sustained interest in the economics and measurement of software work, which he developed further in controlling-software-projects and which ultimately fed into the empirical orientation of peopleware. His move from process methodology to organizational sociology was gradual — Structured Analysis treats software development as a technical and methodological problem; Peopleware argues that the deeper problem was always human.
The transition from the structured methods tradition to the humanistic tradition was not a repudiation — DeMarco continued to respect systematic analysis — but a reorientation of emphasis toward the factors that the methods community had largely ignored.