– Process is perfect when nothing can be taken away.
Naked Process (formally: Naked Decision Process / NDP) is a subtractive methodology for software development. Instead of prescribing what to do, it provides a decision tree that strips process down to what your specific team actually needs — and removes everything else.
Most frameworks add. Naked Process removes.
The software industry spends billions annually on certifications, frameworks, and tooling that optimize the least important variable in the equation. Decades of research — from Sackman (1968) to DeMarco & Lister (1987) to the Standish Group (2020) — consistently show that developer quality is the dominant predictor of project success, not process adherence.
Scrum, Kanban, SAFe, and their derivatives all assume that better rules produce better outcomes. The evidence says otherwise: better people produce better outcomes, and the right amount of process depends entirely on context.
Naked Process takes every known process element — from Scrum, Kanban, XP, classical project management, and beyond — and subjects each to a single question:
Under which conditions does this actually help?
What survives becomes a decision tree. Given your team's specific context — capability, deployment maturity, stakeholder access, decision speed, problem clarity — the tree outputs the minimal viable process: the smallest set of practices that addresses your actual constraints.
Nothing more. Nothing less.
- Subtractive, not additive. Start with everything, remove what doesn't earn its place.
- Context-dependent. There is no universal best process. There are only fitness conditions.
- People over ceremony. Process exists to serve capable people, not to compensate for incapable ones.
- Decision speed over ritual. The co-creator of Scrum's own data shows that decision latency < 1 hour predicts success at 68% — independent of methodology.
- Transparent accountability. Process that distributes accountability to the point of invisibility is camouflage, not management.
The core of Naked Process is a decision tree built on six dimensions:
| Dimension | What It Captures |
|---|---|
| Team Capability | Uniformly strong / mixed with clear lead / junior-heavy |
| Deployment Capability | Continuous delivery possible vs. batched releases |
| Stakeholder Access | Direct and frequent vs. mediated through layers |
| Decision Authority | Where decisions are made and how fast |
| Problem Clarity | Genuinely emergent vs. understood upfront vs. unexamined |
| Coordination Need | Team size, distribution, integration complexity |
The tree doesn't output "use Scrum" or "use Kanban." It outputs a minimal set of context-dependent practices — communication channel, release cadence, feedback mechanism, decision authority — assembled from whatever source survives the removal test.
→ See decision-tree/ for the current working model.
| Document | Description |
|---|---|
| The Manifesto | Why Naked Process exists — the philosophical foundation |
| Points Don't Matter | The research-backed argument that developer quality beats process |
| The Decision Tree | The practical tool — how to find your minimal viable process |
| Methodology | How the decision tree was built: dimensions, decomposition, and validation |
Naked Process grew out of two observations across 30 years of software delivery:
- Every successful project shared one trait: the right people making fast decisions with minimal overhead.
- Every failed project shared one trait: process was present, but judgment was absent.
The manifesto article, "Points Don't Matter: Why Developer Quality Beats Process Every Time," provides the research foundation. The decision tree is the constructive counterpart — not just tearing down, but building what replaces ceremony.
This is a living framework. If you have data — empirical research, case studies, or concrete experience — that challenges or refines any dimension, open an issue or PR. Opinions without evidence will be politely ignored.
CC BY-SA 4.0 — Share and adapt freely, with attribution, under the same terms.
"Economists already know this. Psychologists already know this. DeMarco and Lister wrote it down in 1987. Sackman measured it in 1968. Points don't matter. People do."