p — Predictability
Stability and legibility of commitments, outputs, and procedures. How consistently the organization does what it says it will do.
How the system reads organizational structure.
The MARTRO method operates through two diagnostic instruments — NORN and EOS — that observe the same underlying structural layer from different angles: one internal, one external. Their outputs feed a shared dynamic model that maps optionality and irreversibility across decision paths.
Five structural invariants
Both instruments converge on the same five quantities — structural conditions that determine whether any organizational strategy remains executable over time. These are not performance metrics. They are the underlying layer that performance metrics rest on.
Each invariant is scored on a normalized 0–1 scale. Together they form the capability vector x = (p, c, τ, r, i) that seeds the dynamic model.
c — Cognitive margin
Headroom for non-routine decisions. The inverse of organizational overload — how much capacity exists for judgment beyond execution.
τ — Transferability
Degree to which know-how is codified, portable, and replicable. How much of the organization's capability survives a personnel change.
r — Resilience
Capacity to absorb shocks without collapse of core viable options. Redundancy, substitutability, and structural buffer.
i — Inspectability
Degree to which organizational state and performance signals are visible and interpretable in time. The condition for any corrective action.
NORN
NORN collects structured evidence from inside the organization. Three roles respond separately — Owner/Leadership, Staff, Operators — across ten operational areas, with ten items each. The scale is unified: higher always means better. No reverse scoring.
The critical signal is not the absolute score. It is the gap between roles. When leadership perceives a process as functioning and operations experience it as broken, that divergence is precisely where organizational debt accumulates invisibly. NORN measures both maturity and fractures — and treats the fracture as the more diagnostic of the two.
A second channel — a structured interview method — captures what surveys cannot: real behavior under pressure, informal workarounds, and where the nominal process diverges from the actual one. Together the two channels cover the gap between declared structure and lived structure.
NORN outputs: maturity indices per area and role (0–100), fracture coefficients per area, a Governance–Execution gap index (GMI), and the initial state vector s₀ = (x₀, z₀, G₀, m₀) for the dynamic model.
Methodological posture: procedurally aligned to ISO 10667-1 and ISO 20252 as delivery references. Quality controls include completeness checks, straightlining flags, Cronbach's alpha per area, and Harman common method bias diagnostic. These are instrument-level controls, not certification claims.
EOS
EOS estimates the same five invariants from publicly observable signals — without internal access. It is designed for contexts where a full internal survey is not feasible: competitive screening, portfolio assessment, due diligence, or out-of-sample validation of the model itself.
The instrument comprises 50 items across the five invariants (10 each), scored 1–5 with a mandatory evidence rule: every item requires at least one external reference. In the absence of evidence, the score defaults to 1 or 2 and confidence is marked low. The instrument does not fill gaps with inference.
EOS outputs: invariant estimates with per-item confidence flags, an Optionality index O (weighted toward resilience and inspectability, which are the most regime-sensitive), a Non-Optionality index N, structural bottleneck identification, and a trigger list for re-scoring under changed conditions.
When both instruments are available, EOS serves as a cross-check on the NORN-derived state. Divergences above a defined threshold flag structural inconsistency rather than being averaged away — because divergence is itself diagnostic.
From observation to decision intelligence
The integration step — contained within the EOS pipeline — takes the estimated structural state and runs it through the MARTRO dynamic model. The organization is represented as a constrained stochastic system evolving under regime-switching environmental conditions. The model simulates multiple scenario paths and computes how different decision sequences affect the range of structurally viable futures.
The central output is the door-closing delta: for any decision under evaluation, how much does it contract or expand the set of viable future states? Decisions that close options irreversibly — even if they appear rational in the short term — are flagged before commitment. Decisions that preserve or expand structural room for maneuver are ranked above those that do not, all else equal.
Hysteresis tracks the cumulative cost of past irreversible commitments. It degrades future invariants even after the originating decision is no longer active — making the model sensitive to decision sequences, not only individual choices. This is what organizational debt means in formal terms.
The integration layer does not produce a recommended action. It produces a ranked, comparative map of structural consequences — with sensitivity bounds, not point estimates. Ranking stability across parameter variation is treated as the primary validity criterion.
What the model claims — and what it does not
- The model distinguishes structural constraints from temporary noise. A low score on a single invariant is not a diagnosis; a pattern of low scores with high fracture coefficients is.
- All assumptions are inspectable and contestable. The weight matrix mapping operational areas to invariants is documented, versioned, and auditable — not a black box. Parameter choices are stipulations with stated rationale, not hidden calibrations.
- Outputs are comparative and model-conditional. Rankings are preferred over point estimates. Instability in rankings under scenario variation is reported as structural information, not averaged away.
- The model does not predict financial performance, claim causal inference, certify maturity, or guarantee outcomes. Its claim is narrower and stronger: under explicit structural assumptions, it provides a formal, testable way to detect door-closing decisions and quantify preserved optionality.
For instrument access in this controlled phase, open NORN directly. Open NORN.