Skip to content
Selected system · 02

Where is your operating model weakest for AI adoption?

This diagnostic examines the operating model underneath adoption: how work enters, how decisions get made, how capacity is set, and how the system learns. It identifies the lowest-scoring dimension in the current self-assessment.

Type
Illustrative instrument
Format
24 statements · eight dimensions · 0–3 scale · immediate read
Time
About ten minutes. No email required.
Data
Answers are scored in your browser and are not stored or transmitted.
Caveat
Illustrative readiness read. Not a validated measurement or a benchmark.
The instrument at a glance
  1. Demand Intake Discipline
  2. Problem Framing and Triage
  3. Capacity and Estimation Realism
  4. Cross-Functional Risk and Dependency Visibility
  5. Option Formulation and Tradeoff Clarity
  6. Decision Packet and Governance
  7. Decision Rights and the Executive Moment
  8. Execution Feedback and Learning Cadence

Dimension 8 feeds what it learns back into dimension 1

01The operating problem

The question underneath most AI programs is whether the spend turns into outcomes. That is an operating-model question. When intake, decision rights, capacity, and feedback are weak, AI amplifies the weakness, and adoption produces activity without results. When they are sound, AI compounds them.

Readiness is an operating question before it is a technology question. The read covers whether the organization can absorb demand, make tradeoffs, commit capacity, hold decision rights, and learn from execution: the same disciplines AI is about to accelerate, for better or worse.

02How weakness appears in each dimension

Demand Intake Discipline

Requests arrive as solutions in inconsistent shapes, so leaders see volume without being able to compare value, risk, effort, or urgency.

Problem Framing and Triage

Solution-asks are never converted into problems, so teams optimize locally and later discover they solved the wrong thing.

Capacity and Estimation Realism

Single-point estimates create false precision, capacity buffers vanish, teams overcommit, and roadmaps become fiction.

Cross-Functional Risk and Dependency Visibility

Dependencies and material risks are visible before work starts but are not surfaced, producing late escalations, surprise compliance or security reviews, and duplicated engineering work.

Option Formulation and Tradeoff Clarity

A single option is presented, so there is no real choice and the decision becomes a rubber stamp.

Decision Packet and Governance

Leaders receive status summaries instead of decision-quality information, so meetings become performative updates rather than decisions.

Decision Rights and the Executive Moment

It is unclear who holds the decision right, rationale is not recorded, and everything becomes a priority, so teams absorb hidden overload.

Execution Feedback and Learning Cadence

Execution data never returns to planning, no one owns sunsetting, and the organization repeats the same planning errors every quarter.

03How the system was designed

The diagnostic is derived from a fuller eight-stage intake-to-decision operating system: each dimension maps to one stage of that system, and each statement probes a discipline that stage depends on. The questions are grounded in the specific failure modes of the underlying operating system.

Every AI placement in the underlying design passes one test: it must be bounded in scope, observable in effect, and reversible if wrong. The executive decision moment deliberately has no AI in the loop.

04What the output means
0.000.99

Tool-led, model unset

The operating model cannot yet absorb AI. Adoption will produce activity without outcome. The first moves are upstream: intake and triage.

1.001.74

Fragmented and inconsistent

Pockets of discipline exist, and demand and decisions are not yet governed end to end. AI amplifies the weakest link. Fix the lowest dimension first.

1.752.49

Governed but uneven

Decisions are governed, and one or two capabilities lag, usually capacity realism or feedback cadence. AI helps where the model is solid and stalls where it is not.

2.503.00

Decision-ready operating system

The model can absorb AI as an amplifier. The risk shifts to holding discipline as scale and agent adoption increase.

Alongside the profile, the result names the lowest-scoring dimension in the current self-assessment, the associated failure mode, the operating cost that weakness can create, and the next operating move. When every dimension is strong, the result shows where the model holds and what discipline protects it at scale.

05The AI / judgment boundary
AI assists

Prepares, detects, routes, summarizes, and drafts across intake, triage, estimation checks, risk mapping, option drafting, and packet assembly.

Humans decide

Approve, fund, defer, accept risk, override, and commit capacity. The executive decision has no AI in the loop, and the rationale is recorded.

The diagnostic takes about ten minutes and returns an immediate read.