Who Does What, Who Needs to Know, and Who Has the Power to Stop You: The RACI Matrix and the Wider World of Stakeholder Engagement

Of the many tools available to delivery professionals, few are as widely used and as frequently misunderstood as the RACI matrix. In most project and programme environments it is completed early, filed in the programme management plan, and consulted rarely thereafter. That pattern of use is revealing, because it suggests that RACI is often treated as a documentation requirement rather than a living instrument of governance and engagement.

Used well, the RACI matrix is one of the most effective tools for preventing the ambiguity that causes delivery programmes to slow, drift, and generate unnecessary conflict. Used poorly, or used in isolation, it creates a false sense of clarity while leaving the deeper questions of stakeholder influence, sentiment, and engagement strategy entirely unaddressed.

Understanding what RACI actually does, what it cannot do, and how it relates to the broader suite of stakeholder engagement models is not a theoretical exercise. It has direct consequences for how well a programme manages its most complex relationships, and therefore for how reliably it reaches its intended outcomes.

What the RACI Matrix Is and What It Is Actually For

RACI is an acronym that defines four types of relationship between individuals or roles and specific tasks, decisions, or deliverables:

• Responsible: the person or people who do the work. There may be more than one, though excessive distribution of responsibility is itself a governance risk.
• Accountable: the single individual who owns the outcome and answers for it. There must be exactly one accountable owner per item; having more than one is the functional equivalent of having none.
• Consulted: those whose input is sought before a decision is made or work is completed. The communication here is two-way.
• Informed: those who are notified of progress or outcomes but who are not involved in the work itself. The communication is one-way.

At its core, RACI is a role clarity tool. Its primary contribution is to eliminate the ambiguity that arises when multiple people believe they are responsible for the same thing, when accountability is diffuse and therefore effectively absent, and when people who should be consulted are informed after the fact or not at all. These are not minor administrative inconveniences; they are among the most reliable predictors of delivery failure.

The RACI matrix earns its place precisely because it forces an explicit conversation about these questions at the outset. Who owns this? Who does the work? Who needs to be asked? Who needs to be kept informed? These questions sound straightforward, but in complex organisations with matrixed structures, multiple suppliers, and overlapping mandates, the answers are rarely self-evident. The discipline of constructing the matrix surfaces the ambiguity and provides a forum in which it can be resolved.

Where RACI Performs at Its Best

The RACI matrix is most powerful when applied to governance structures, decision rights, and the management of cross-functional work. In these contexts, it provides the clarity that prevents the two most common failure modes in delivery: diffusion of accountability (in which everyone is vaguely responsible and nobody is truly accountable) and over-escalation (in which decisions that could be made at the working level are pushed upwards because nobody is clear on who has the authority to make them).

In a well-constructed RACI, every material task and decision has one accountable owner, a clearly identified set of responsible parties, and a defined list of stakeholders who must be consulted before and informed after. This structure supports effective governance because it makes the escalation path visible and because it provides a shared reference point when disputes about authority or process arise.

RACI also performs well as an onboarding tool. When a new team member, a new supplier, or a new executive sponsor joins a programme, the RACI provides an immediate orientation to how the programme is structured, where their role sits within it, and what they are and are not expected to own. This reduces the ramp-up time and the risk of early missteps that come from misunderstood expectations.

Common RACI Errors That Undermine Its Effectiveness

• Multiple accountable owners. Shared accountability is not accountability. It is a mechanism for avoiding it.
• Confusing Consulted and Informed. Stakeholders who should be contributing to decisions but who are only notified of them after the fact will disengage and, in some cases, actively resist.
• Over-populating the Responsible column. When everyone is responsible, the individual incentive to own the outcome diminishes proportionally.
• Treating RACI as a one-time exercise. Programmes evolve. Roles change. New stakeholders emerge. A RACI that is never updated rapidly ceases to reflect reality.
• Constructing RACI in isolation. A matrix that is produced by the programme team without the involvement of the stakeholders it describes will contain errors, generate resentment, and be treated with scepticism by those it purports to govern.

What RACI Cannot Do: The Limits of a Role Clarity Tool

For all its utility, the RACI matrix has significant limitations that become visible when it is asked to carry more weight than it is designed to bear. The most important of these is that RACI tells you nothing about power, sentiment, or influence.

A stakeholder who appears in the Informed column may nevertheless have the organisational standing, the political relationships, or the informal authority to derail the programme entirely. A stakeholder designated as Consulted may be deeply hostile to the change and likely to use that consultation to obstruct rather than to improve. A stakeholder who does not appear in the matrix at all may be the most consequential voice in the room when the programme encounters difficulty.

RACI also tells you nothing about the quality of the engagement it prescribes. Consulting a stakeholder means very little if the consultation is perfunctory, if the input provided is not genuinely considered, or if the stakeholder has no confidence that their views will influence anything. The matrix describes the structure of engagement; it says nothing about the substance or the quality of the conversations that take place within it.

Finally, RACI is a static instrument in an inherently dynamic environment. Stakeholder landscapes shift as programmes progress. Priorities change. Personnel change. Political conditions change. A RACI matrix constructed at programme initiation may be materially misleading by the time the programme reaches its most critical phase.

Complementary Models: Building a Complete Stakeholder Picture

Effective stakeholder engagement requires a suite of tools, each chosen for what it adds to the overall picture. The following models each address dimensions of the stakeholder landscape that RACI leaves unexamined.

The Power/Interest Grid (Mendelow Matrix)

Developed by Aubrey Mendelow, this model maps stakeholders across two axes: their level of interest in the programme and their level of power to influence it. The resulting quadrants generate four strategic stances: manage closely (high power, high interest); keep satisfied (high power, low interest); keep informed (low power, high interest); and monitor (low power, low interest). The model is valuable precisely because it directs engagement effort to where it will have the greatest effect and prevents the common error of distributing attention uniformly across a stakeholder population of very unequal consequence.

The Salience Model (Mitchell, Agle, and Wood)

The salience model takes a more sophisticated approach, assessing stakeholders across three attributes: power (the ability to impose one's will), legitimacy (the social and moral basis of a claim on the programme), and urgency (the time-sensitivity of that claim). Stakeholders who score highly on all three attributes are definitive stakeholders who demand immediate and sustained attention. Those who score on only one or two are latent stakeholders whose salience may grow as the programme develops. The model is particularly useful in politically complex or public-facing environments where the legitimacy of stakeholder claims is itself contested.

The Stakeholder Engagement Assessment Matrix

Used within the Managing Successful Programmes and PMI frameworks, this model maps stakeholders across a spectrum of engagement states: unaware, resistant, neutral, supportive, and leading. For each stakeholder group, the matrix captures both the current state and the desired state. The gap between the two defines the engagement work required. Crucially, this model is designed to be updated at regular intervals, making it one of the few tools that tracks movement in the stakeholder landscape over time rather than capturing a single point-in-time view.

DACI and RAPID

These models focus specifically on decision rights rather than on the broader stakeholder landscape. DACI defines who Drives a decision, who Approves it, who Contributes to it, and who is Informed of it. RAPID (Recommend, Agree, Perform, Input, Decide) serves a similar function with slightly different categorisation. Both models are more granular than RACI on the question of decision authority and are particularly well suited to governance design in organisations with complex hierarchies or multiple layers of approval. They complement RACI by adding precision to the decision-making architecture that RACI establishes in outline.

The Stakeholder Onion Diagram

This model represents the programme at its centre, surrounded by concentric rings of stakeholders ordered by their proximity and directness of relationship to the work. The innermost ring contains the core delivery team; subsequent rings contain the immediate sponsor group, the wider organisational constituency, and external stakeholders respectively. The diagram is not an analytical tool in the way that the salience model or the power/interest grid are, but it is an effective scoping and communication device that helps teams see the full range of their stakeholder relationships and identify where engagement effort is concentrated relative to where it may need to reach.

Comparing the Models: A Summary

The table below summarises the primary purpose, strengths, and limitations of each model discussed, to support practitioners in selecting the right combination for their context.

Using the Models Together: An Integrated Engagement Architecture

The most effective delivery teams do not choose between these models; they use them in combination, each serving a distinct function within an overall engagement architecture. A practical and proven approach is to sequence their use across the programme lifecycle:

• At initiation: use the Power/Interest Grid and the Stakeholder Onion Diagram to map the landscape and scope the engagement effort; use RACI to establish accountability and role clarity across the programme governance structure.
• During planning: use the Salience Model to identify which stakeholders demand immediate and sustained attention; use DACI or RAPID to design the decision-making architecture with precision.
• Through delivery: use the Engagement Assessment Matrix to track movement in stakeholder sentiment at regular intervals and to identify where the engagement plan needs to be adjusted.
• At key transitions: revisit all models as the programme moves between phases, as personnel change, and as the external environment shifts. A stakeholder map that was accurate at initiation will rarely remain so through to benefits realisation.

This integrated approach ensures that role clarity, power dynamics, legitimacy, sentiment, and decision rights are all held in view simultaneously, and that the engagement strategy is responsive to a changing reality rather than anchored to a set of assumptions made at the beginning of the programme.

Embedding Engagement in Governance by Design

The most important thing that can be said about stakeholder engagement models is that they are only as valuable as the governance structures that embed them in the programme's operating rhythm. A RACI that lives in a SharePoint folder and a power/interest grid that was produced in a workshop and never revisited are not engagement tools; they are documentation artefacts.

Governance by design requires that stakeholder engagement is a standing item in programme governance: that the engagement assessment matrix is reviewed at every programme board, that changes in stakeholder sentiment trigger a deliberate response rather than a passive note, and that the accountable owner in the RACI for each material stakeholder relationship is visible and active. It requires that the communications and engagement function is connected to the decision-making process rather than operating as a separate workstream that delivers newsletters and town halls in parallel with the real programme.

When this connection is made, stakeholder engagement ceases to be a soft activity that delivery teams reluctantly fund and begins to be treated as a strategic asset. The intelligence gathered through structured engagement informs risk management. The trust built through honest and targeted communication accelerates decision-making. The clarity produced by a well-maintained RACI and a live decision rights framework reduces conflict and rework. Together, they represent the difference between a programme that merely completes and one that delivers.

The Hervey Dickens Perspective

The RACI matrix is a genuinely important tool, and its importance is not diminished by acknowledging its limits. What it does, it does well: it creates role clarity, surfaces accountability, and provides a shared reference point for governance. What it does not do is describe the full complexity of the stakeholder landscape that every significant programme must navigate.

At Hervey Dickens Consulting, we work with leadership teams to build stakeholder engagement architectures that are fit for the scale and complexity of the work: combining RACI with the analytical depth of the salience model, the dynamic tracking of the engagement assessment matrix, and the decision-rights precision of DACI. Our digital services are designed to make this architecture operationally real, connecting engagement intelligence to programme governance so that the conversations a programme holds actually shape the decisions it makes.

The question for any organisation preparing for significant change is not which model to use. It is whether the models in use are genuinely informing the programme, or simply being produced to satisfy an audit trail. The answer to that question determines more about delivery success than almost any other single factor.