Adaptive Stage-Gate Protocol¶
Protocol — instantiates Malleability Window Governance
Allows progression only when evidence, reversibility, stakeholder legitimacy, and option-preservation criteria are satisfied.
The Adaptive Stage-Gate Protocol is a forward-progression rule: a deployment advances one rung at a time, and only when a pre-agreed conjunction of criteria is met — enough evidence has been gathered, enough reversibility still remains, stakeholders have signed off, and viable fallbacks are still alive. It borrows the ritual of the classic engineering stage-gate but rewrites what the gate is measuring. A conventional gate asks is the work finished, on budget, at quality? This one asks would advancing spend reversibility faster than it buys knowledge? An increment can be refused not because the work is unready but because taking it would slam the malleability window shut before the evidence to justify slamming it has arrived. The gate is the point where momentum has to convert itself into a defensible, evidence-linked case for the next rung.
Example¶
A city water utility plans to replace 180,000 residential meters with networked "smart" meters. Rather than one procurement, the protocol builds a ladder: (1) a 500-meter bench-and-lab rung; (2) a 2,000-meter neighborhood pilot on a month-to-month firmware licence; (3) a 20,000-meter district; (4) the full fleet on a ten-year vendor contract. Each rung carries its own evidence standard and an explicit lock-in budget. At the gate from rung 3 to rung 4, the technical evidence is strong — but the gate is refused, because advancing requires signing the ten-year proprietary head-end contract, which would collapse the utility's fallback to a rival system. The gatekeeper (an oversight committee, not the project sponsor) sends the rung back for redesign: renegotiate for an interoperable head-end first, then advance. The utility ends up scaling on evidence without spending its exit.
How it works¶
- A conjunctive gate. Passage requires all of: evidence sufficiency for the rung, reversibility remaining above a floor, stakeholder legitimacy, and preserved options. Any single failure holds or redesigns the advance — good outcomes cannot buy their way past a collapsed fallback.
- Pre-registered lock-in budgets. Each rung declares in advance how much dependency, contract length, and switching cost it is allowed to create, so a "pilot" cannot quietly sign a decade-long deal.
- A three-way verdict. A gate can pass, hold, or send back for redesign — not just go/no-go.
- Separated authority. The party that passes a gate is structurally distinct from the party pushing to advance.
Tuning parameters¶
- Rung granularity — few large steps or many small ones. Finer rungs buy more decision points and cheaper reversal, at the cost of process overhead and slower rollout.
- Gate stringency — how many criteria, and whether each is a veto or a weighted score. Veto-on-any is safest against lock-in but most prone to stalling.
- Lock-in budget per rung — how much dependency each step may create. Tighter budgets protect the window but can starve a rung of the commitment it needs to generate real evidence.
- Authority independence — how insulated the gatekeeper is from the sponsor. More independence resists momentum but adds friction.
When it helps, and when it misleads¶
Its strength is converting accumulating momentum into evidence-linked escalation: scale rises because learning justified it, not because a schedule demanded it, and reversibility is spent deliberately rather than by drift. Its central failure mode is gate theater — when the sponsor effectively controls the gate, criteria soften into a rubber stamp and each pass becomes another turn of the ratchet, the textbook pattern of escalation of commitment.[n1] A classic misuse is adding rungs without lock-in budgets, so the pilot rung signs the long contract and the later gates are deciding over an already-closed window. The guarding discipline is a genuinely independent gatekeeper judging against criteria pre-registered before the rung began.
How it implements the components¶
staged_commitment_ladder— defines the rungs, each with its own evidence standard, rollback path, and lock-in budget.governance_authority_gate— the gate itself: a named authority empowered to pass, hold, or redesign an advance against sponsor pressure.option_preservation_buffer— advance is conditioned on fallbacks and alternatives still being alive; a rung that would kill options is refused until options are restored.
It sets criteria for going forward, not for stopping what is already running — the emergency halt of a live deployment is moratorium_trigger, owned by its hazard-twin Pause or Moratorium Trigger Protocol; and it consumes rather than draws the information_malleability_curve_map produced by Collingridge Curve Workshop.
Related¶
- Instantiates: Malleability Window Governance — this protocol is the evidence-linked progression spine of the archetype.
- Consumes: Collingridge Curve Workshop supplies the curve map and horizon that set each rung's lock-in budget.
- Sibling mechanisms: Collingridge Curve Workshop · Deployment Impact Dashboard · Exit and Interoperability Rule · Pause or Moratorium Trigger Protocol · Post-Pilot Lock-In Audit · Regulatory or Operational Sandbox · Reversibility Horizon Review · Stakeholder Harm Reporting Channel · Sunset Clause with Renewal Hearing
Editorial Notes¶
Form Classification¶
Form family: Decision, Gate & Allocation
Rationale: The mechanism allows progression only when evidence, reversibility, stakeholder legitimacy, and option-preservation criteria are satisfied, so its operative form is a bounded choice, gate, routing, or allocation.
Independent corroboration: The frozen evidence defines Adaptive Stage-Gate Protocol as 'Allows progression only when evidence, reversibility, stakeholder legitimacy, and option-preservation criteria are satisfied', so its operative form is Decision, Gate & Allocation.
Nearest alternative: Protocol, Workflow & Routine — Its stage gates make bounded continuation and resource-release decisions; the staged protocol orders those gates.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Engineering & Design
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Engineering and product-development stage gates formalized evidence-based progression through increasingly committed design and deployment phases.
Related originating lineages:
- Innovation & Entrepreneurship — Lean experimentation and staged venture investment connect additional commitment to evidence, option preservation, and learning.
- Law & Governance — Stakeholder legitimacy, authorized advancement, and explicit criteria contribute procedural validity.
- Organizational & Management Science — Portfolio governance supplies independent gatekeepers, resource authority, documentation, and escalation paths between stages.
- Psychology — Research on escalation of commitment and sunk-cost bias motivates independent review and forward-looking criteria before the next irreversible step.
- Ethics of Technology & AI Governance — Reversibility and staged deployment under uncertainty are central contemporary governance requirements for risky technology.
Review resolution: Both reviewers identify engineering stage-gate practice as the primary operational lineage. Innovation management, legal accountability, organizational authority, commitment psychology, and technology governance materially reshape the gate around reversibility and legitimacy, so the cross-disciplinary synthesis is retained.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Escalation of commitment — the well-documented tendency to keep investing in a chosen course because of what has already been sunk into it, rather than on its forward merits. A gate controlled by the sponsor is where escalation hides, which is why gatekeeper independence is the load-bearing design choice here. ↩