Vroom-Yetton-Jago Decision Model

Match participation to the decision rather than using the same autocratic, consultative or group process every time.

In one minute

The Vroom-Yetton-Jago model helps a decision owner choose how others should participate. Its classic process styles range from leader-led choice to group decision:

• Decide: the leader decides using available information.
• Consult individually: selected people provide information or views; the leader decides.
• Consult the group: the issue is discussed collectively; the leader decides.
• Facilitate a group decision: the group develops and owns the choice within agreed boundaries.
• Delegate: responsibility is assigned within clear constraints when the later model permits it.

The appropriate style depends on decision-quality requirements, information location, need for commitment, goal alignment, likely conflict and time.

Best for: consequential team decisions where participation design affects quality or implementation.
Avoid when: law or governance already fixes the accountable decision maker, or the issue is an emergency with a predetermined command protocol.

The problem it addresses

Leaders often confuse participation with either courtesy or loss of control. They consult when the decision is already made, or they schedule consensus meetings for choices that require a clear accountable owner.

The model treats participation as a design choice. More participation can improve information and commitment, but it also uses time and can amplify conflict. Less participation can be appropriate when quality requirements are clear and the owner has sufficient evidence.

When to use it

Use the model when:

• decision quality depends on knowledge held by others;
• implementation requires genuine team commitment;
• AI decision rights or human oversight are being designed;
• stakeholder goals may conflict;
• the leader’s information is incomplete;
• time pressure must be balanced against participation.

Apply it to one real decision, not to a leader’s general personality.

When not to use it

Do not use it:

• to avoid statutory or fiduciary accountability;
• as a test of whether a leader is democratic;
• after promising group authority that does not exist;
• when affected people are excluded from a high-impact decision without justification;
• to let an AI system become the de facto decision owner;
• without communicating the selected participation mode.

Use Kepner-Tregoe Decision Analysis to compare alternatives after the participation process is designed.

Inputs required

• a specific decision and accountable owner;
• required quality, safety and legal conditions;
• information held by the owner and others;
• stakeholder goals and likely conflict;
• implementation dependency and commitment needs;
• time available and cost of delay;
• non-delegable authority boundaries;
• a communication and decision-record plan.

Step-by-step process

1. State the decision and authority boundary

Name what will be decided, by when and who remains accountable. Record any authority that cannot be delegated.

2. Assess the quality requirement

Ask whether one choice could materially outperform another on customer outcome, safety, legality or strategy. If quality differences are small, a faster process may be appropriate.

3. Assess information distribution

Determine whether the owner has sufficient evidence to make a high-quality choice and whether others hold unique operational, technical or affected-user knowledge.

4. Assess problem structure

Ask whether requirements and evaluation rules are clear enough to support a defensible choice. Unstructured decisions often benefit from consultation or facilitation.

5. Assess commitment

Determine whether successful implementation depends on people accepting the decision, and whether they would support a leader-made choice without participation.

6. Assess goal alignment and conflict

Identify whether participants share the organisation’s objectives for the decision and whether preferred solutions are likely to produce significant conflict.

7. Consider time and participation cost

Record the real deadline and consequence of delay. Time pressure narrows the feasible process but should not become an excuse to ignore essential expertise.

8. Select the participation mode

Choose the least burdensome mode that can still meet quality and commitment needs. Specify who provides information, who is consulted, who recommends and who decides.

9. Communicate the process before engagement

Tell participants whether they are supplying evidence, advising the owner or sharing decision authority. False consultation damages future commitment.

10. Review the result

After implementation, assess decision quality, commitment, time used and whether the participation design excluded necessary information.

AI automation lens

The model is useful for defining human decision rights around AI, but the AI system is not another accountable participant. Separate roles such as:

• model prepares evidence or options;
• domain expert validates material claims;
• affected-user representative supplies experience evidence;
• risk owner sets non-negotiable controls;
• accountable manager approves deployment;
• operator can pause or reverse action.

High-impact or hard-to-contest decisions may require broader participation even when automation is technically feasible. Record who may override the model, who reviews exceptions and who communicates the outcome to affected people.

Visual model

Text alternative: decision-quality and commitment questions lead through information, goal-alignment, conflict and time checks. The result selects a participation mode from leader decision through individual or group consultation to facilitated group choice, while accountability remains named.

Interactive example

Scenario

A bank considers allowing an AI system to draft and approve low-value fee refunds. The operations director owns the deployment decision. Compliance understands contestability, agents know exception patterns, technology owns system controls and customers experience the outcome.

The director has little exception evidence. Implementation requires agents to identify unsafe cases, while compliance and operations disagree about acceptable autonomy.

Your move

Select a participation mode and explain the authority boundary.

Worked answer

A leader-only decision is inappropriate because essential information is distributed and implementation depends on agent behaviour. Pure delegation is also inappropriate because deployment accountability and regulatory obligations remain with the director.

The director should facilitate a structured group recommendation with operations, compliance, technology and customer-experience representation. The group defines eligible cases, controls and stop conditions. The director and designated risk owner retain formal approval. Agents receive explicit pause authority during the pilot.

Facilitation notes

• Apply the questions to a decision, not a personality.
• Distinguish advice from shared authority.
• Include people who hold unique evidence, not only senior roles.
• State non-delegable accountability at the start.
• Do not promise consensus if the owner will decide.
• Treat affected-user knowledge as decision evidence.
• Review whether urgency is real or manufactured.

Expected output

A sound application produces:

• a defined decision and accountable owner;
• quality and information requirements;
• commitment, alignment and conflict assessment;
• a justified participation mode;
• named evidence, advisory and decision roles;
• a communicated authority boundary;
• a post-decision review measure.

Common mistakes

1. Using one preferred leadership style. The situation is never analysed.
2. Calling information gathering consultation. Participants expect influence that they do not have.
3. Seeking consensus for non-delegable authority. Accountability becomes ambiguous.
4. Excluding operational knowledge. Decision quality falls despite senior participation.
5. Letting urgency erase safeguards. A preventable governance gap is created.
6. Treating AI as the decision owner. Accountability cannot be delegated to a model.
7. Ignoring implementation commitment. A technically sound choice fails in use.

Quality checklist

• One decision and deadline are defined.
• The accountable owner and non-delegable authority are clear.
• Decision-quality requirements are explicit.
• Information gaps and holders are identified.
• Commitment needs are assessed.
• Goal alignment and conflict are considered.
• Time pressure is evidenced.
• Participation and decision rights are communicated honestly.
• AI supports but does not own the accountable decision.

Template

Selected mode: decide / consult individually / consult group / facilitate / delegate

Roles and communication:

Knowledge check

A manager has already chosen a vendor but invites staff to a “group decision” workshop. What is the central failure?

A. The workshop is too short.
B. The stated participation mode does not match the real authority boundary.
C. Vendors should never be selected by managers.
D. Every decision must be delegated.

Answer: B. False participation damages the quality and legitimacy of the process.

Related tools

• Decision content with: Kepner-Tregoe Decision Analysis
• Distributed expert input through: Delphi Method
• Execution with: Objectives and Key Results, Process Decision Program Chart
• Not to be confused with: a personality test, responsibility matrix or voting rule

References

1. Vroom, V. H., and Yetton, P. W. Leadership and Decision-Making. University of Pittsburgh Press, 1973.
2. Vroom, V. H., and Jago, A. G. The New Leadership: Managing Participation in Organizations. Prentice Hall, 1988.
3. Vroom, V. H. “A New Look at Managerial Decision Making.” Organizational Dynamics, 1(4), 1973, pp. 66-80. DOI (opens in a new tab).
4. National Institute of Standards and Technology. Artificial Intelligence Risk Management Framework (AI RMF 1.0). NIST AI 100-1, 2023. Official publication (opens in a new tab).

Sources reviewed 12 August 2026.