Leavitt Diamond Model Questionnaire

To run a Leavitt diamond check, use this questionnaire. It is a fast second-order effects check for a specific proposed change, scoring the disruption to each of the four variables from 1 (no effect) to 5 (major disruption).

Note the inverted scale. Unlike the other questionnaires here, a high score is a warning, not a good result. You are looking for large unplanned effects.

Budget thirty minutes. This is a checklist, not a diagnostic.

See leavitt-diamond-model.

Evidence: Weak. A conceptual heuristic from 1965, never validated. Use it to surface consequences nobody has thought about, then use a real diagnostic.

Step 0: State the change

Write one sentence describing the specific change being proposed, and circle which variable it is primarily aimed at: Task, People, Structure, or Technology.

Change: ________________________________________________

Primary target: [ ] Task [ ] People [ ] Structure [ ] Technology

Leavitt's own observation: change agents choose a favorite point of entry — technologists change technology, HR changes people, executives change structure — and are then surprised by effects at the other three. The remaining steps examine those three.

Step 1: Task

1. Work content

• Question: How much does this change what the work actually is, as opposed to how it is done?
  • Rating (1–5): [ ]

2. Judgment

• Question: How much does it change what people must decide, or transfer judgment from a person to a system?
  • Rating (1–5): [ ]

3. Difficulty

• Question: How much does it change how hard the work is, or shift the balance between interesting work and exception handling?
  • Rating (1–5): [ ]

Step 2: People

1. Skills

• Question: How much does it change which skills matter, and whose hard-won expertise becomes less valuable?
  • Rating (1–5): [ ]

2. Roles and identity

• Question: How much does it change who someone is at work — their professional identity, status, or sense of competence?
  • Rating (1–5): [ ]

3. Numbers

• Question: How much does it change how many people are needed, or which people?
  • Rating (1–5): [ ]

Step 3: Structure

1. Authority

• Question: How much does it change who decides, who approves, and who can act without asking?
  • Rating (1–5): [ ]

2. Information

• Question: How much does it change who holds information, and therefore who holds influence?
  • Rating (1–5): [ ]

3. Boundaries

• Question: How much does it change team boundaries, handoffs, or who depends on whom?
  • Rating (1–5): [ ]

Step 4: Technology

1. Tools

• Question: How much does it change the tools, systems, or techniques people use daily?
  • Rating (1–5): [ ]

2. Dependency

• Question: How much does it create a new dependency on a system, a vendor, or a team that did not exist before?
  • Rating (1–5): [ ]

3. Failure modes

• Question: How much does it change what happens when things go wrong, and whether people can still work?
  • Rating (1–5): [ ]

Step 5: Analysis

• Any variable scoring 4 or 5 that is not your primary target is an unplanned consequence. Name an owner for it now, before the change proceeds. This is the entire point of the exercise.

• A change with a high score on only its target variable is either genuinely contained or has not been thought through. In practice the second is far more common. Ask someone who does the work daily to score it independently; their scores are usually higher than the sponsor's.

• High People-identity scores predict resistance that will be misread. Loss of status, competence, or professional identity produces objections that get described as "resistance to change" and are actually accurate reports of a real cost. See bridges-transition-model.

• High Structure-authority scores mean this is a power change, whatever it is being called. Power changes announced as tooling changes are detected quickly and resented durably.

• This model has no environment term. It will tell you nothing about whether the change is right, only about what else it will move. Pair it with a real diagnostic: nadler-tushman-congruence-model-questionnaire for fit, or galbraith-star-model-questionnaire for structural decisions.