Stakeholder Engagement Assessment Matrix
The stakeholder engagement assessment matrix compares, for each stakeholder in a project, their current level of engagement with the level the project team considers necessary for success. Both levels sit on one scale: unaware, resistant, neutral, supportive, leading. A letter C marks the current level and a letter D the desired level. The gap between the two letters shows which stakeholders need to move and how far. The PMBOK Guide lists the matrix among the tools of two processes: planning stakeholder engagement and monitoring it.
Purpose
The stakeholder engagement assessment matrix is a grid: one row per stakeholder, one column per engagement level.
It answers a question the other tools leave open: which stakeholders must change position and how far? The stakeholder list says who is involved in or affected by the project. The stakeholder matrix ranks each one by power and interest and derives the attention they deserve. The engagement assessment matrix measures each stakeholder's attitude towards the project and sets them a target.
The matrix helps the team decide where to direct the communication effort and how hard. The deliverable is the completed, dated matrix. It feeds the stakeholder engagement plan, often called the stakeholder management plan, which turns each gap into actions, owners and deadlines.
For the business analyst, the matrix also flags an elicitation risk. A stakeholder who is a source of requirements and stays "unaware" or "resistant" gives incomplete answers in sessions or disputes the framing of the sessions. The analyst schedules interviews and workshops with that stakeholder once the team has acted on the gap.
Usage
When to use it
- Drawing up an engagement plan: set a target level for each stakeholder before writing the actions.
- A project that changes working practices: name the expected resistance before rollout.
- A limited communication budget: focus the effort on the gaps of stakeholders who can block the project.
- A phase review or steering committee: check that current levels are moving towards target levels.
- A change of sponsor, scope or organisation: reassess current levels, which may have moved.
When not to use it
- Stakeholders not yet identified: the matrix has no rows to fill. Draw up the stakeholder list first.
- A question of priority or attention: the matrix measures neither power nor interest. Use the stakeholder matrix.
Description
Two marks per row
The rows come from the stakeholder register: one person or group per row. The columns carry the five engagement levels, always ordered from least to most engaged. In each row, the team writes a C in the column of the current level and a D in the column of the desired level. The PMBOK Guide defines D as the level the team considers essential to project success.
A row can carry C and D in the same cell: the stakeholder is already where the project needs them, and the effort is limited to keeping them there. Otherwise, the distance between the two letters measures the ground to cover. To keep a stakeholder at their level, the team keeps informing them at the communication plan's usual pace and reviews their row again when the sponsor changes or the scope moves.
Five levels on one scale
| Level | PMBOK Guide definition | Observable signs |
|---|---|---|
| Unaware | Does not know about the project and its possible impacts. | Never mentions the project; first hears of it when someone presents it to them. |
| Resistant | Knows the project and its impacts, opposes the changes it brings. | Repeated objections, delayed approvals, missed sessions, requests for postponement. |
| Neutral | Knows the project, neither supports nor opposes it. | Attends without speaking up; answers requests without anticipating them. |
| Supportive | Knows the project and its impacts, supports the work and its outcomes. | Releases staff for the project, relays the project's messages, approves on time. |
| Leading | Knows the project and its impacts, works actively for its success. | Defends the project where the team has no access, settles conflicts, removes obstacles on their own initiative. |
The five levels form an ordered scale. The first differs from the other four in what the stakeholder knows: an unaware stakeholder has no attitude towards the project yet. The next four assume that the stakeholder knows the project and rank them by degree of support, from opposition to active commitment. The guide does not state this order, but figure 5-23 arranges the columns from "unaware" to "leading".
A gap is counted in columns: a C at "unaware" and a D at "neutral" make a gap of two levels. From "unaware", the first task is always to inform, whatever the target. Once informed, the stakeholder takes a position at one of the other four levels, and the team reviews their C.
Establishing the current level
The C is an observation. The analyst or project manager derives it from what the stakeholder has done: attendance at sessions, approval lead times, decisions taken or deferred, statements made in front of their teams. What stakeholders say in a session weighs less than what they do afterwards. Stakeholder analysis, which the PMBOK Guide places upstream, provides part of this information: it records each stakeholder's expectations and attitude, meaning their degree of support for the project. An individual interview fills in the picture when the evidence is missing.
Two team members assess each row separately, then compare their choices. A disagreement of one level is settled by discussion. A disagreement of two levels means that one of them is missing a fact, which has to be found before scoring. Next to each C, a sentence records the fact that justifies it. Without it, the next review cannot tell whether the person changed or the assessor did.
Setting the desired level
The D is a decision. The team sets it per stakeholder, according to what the project expects of them: an approval, resources, the project's messages passed on to a team, an absence of opposition. The desired level is the lowest level that covers this expectation. Each extra level costs team time and demands on the stakeholder, and most stakeholders do not need to drive the project. A neutral staff committee is often enough; a sponsor who is only supportive rarely is.
The stakeholder's power partly determines how high the D sits. A stakeholder able to stop the project justifies a high D and the effort that goes with it. A peripheral stakeholder can stay neutral without harm. The stakeholder matrix provides this ranking, and the team reads the two matrices together. Next to each D, a sentence records what the project expects of the person. That sentence becomes the objective of the engagement action.
Reading the gap
The PMBOK Guide says only that the gap guides the level of communication needed. The size of the gap gives a rough first ranking. The levels it crosses decide the effort, because each step calls for different work. The stakeholder's power decides the priority. This step-by-step reading of the gap comes from practitioners, not from the guide.
| Step | Work | Typical action |
|---|---|---|
| Unaware to any other level | Inform | Present the project, its impacts on the stakeholder and its timeline; record the position they take. |
| Resistant to neutral | Address an objection | Hear the objection, record it, change the solution or explain why it stays. |
| Neutral to supportive | Show a benefit | Link the project to one of the stakeholder's objectives; involve them in a decision that affects them. |
| Supportive to leading | Ask for a role | Hand over a responsibility: settle conflicts, present the project, carry a message. |
The step from "resistant" to "neutral" is the most expensive on the scale. As long as the objection goes unanswered, every message that only informs confirms to the resistant stakeholder that the team is not listening. For stakeholders who do not support the project, the PMBOK Guide suggests following the engagement assessment with a communication styles assessment. That assessment specifies the form, channel and content of the messages to send them.
Each gap becomes a row in the engagement plan: the stakeholder, the target step, the action, the owner and the deadline. The PMBOK Guide places the matrix in two processes: Plan Stakeholder Engagement produces this plan; Monitor Stakeholder Engagement tracks how the gaps shrink.
Repeating the assessment
The matrix is filled in several times over a project. At each review, the team reassesses the Cs in light of new facts and keeps the Ds unless the project has changed. A C that has not moved after two reviews points to an ineffective action. The team changes the action before the next review.
A D moves when what the project expects of the person moves: a change of scope, a new phase, a replaced sponsor. A row is added when the register grows. The team keeps the dated versions. By comparing two versions, it sees which gaps close, stall or open.
What distorts the matrix
A default "neutral". Lacking a fact, the assessor writes "neutral". The row can then hide an unaware stakeholder as easily as a silently resistant one, and neither receives an action. Without a fact, the assessor writes "?" in place of the C and schedules the interview that will supply the fact.
Every D at "leading". The matrix no longer ranks anything: the gaps that threaten the project get the same attention as those with nothing at stake.
A group that hides its divisions. A "team leaders" row merges two supportive team leaders and three resistant ones into a "neutral" that describes nobody. When a group splits, the team splits its row.
A circulated matrix. A person who finds themselves marked "resistant" in a shared document digs in. Only the team and the sponsor consult the matrix. An assessment of a named person's attitude is also personal data: under the Federal Act on Data Protection (FADP) in a private organisation, under cantonal law in a cantonal or municipal public body. The person can ask the organisation for the data it holds about them (art. 25 FADP; cantonal laws provide the same right of access). Restricted access therefore does not guarantee that the matrix stays internal. Each C describes a fact the person could read: "postponed two approvals", not "blocks everything".
AI considerations
A language model can propose a first C from meeting minutes and written exchanges. Appendix X3 of the PMBOK Guide describes this use under the name stakeholder sentiment analysis: language-processing tools read emails, messaging, minutes and social media to determine how stakeholders feel. The model spots repeated objections, late approvals and the subjects on which a person stays silent. For each gap, it drafts a first version of the messages and actions, adapted to the target step. Between two reviews, it compares the two versions of the matrix and lists the gaps that close or open.
The model reads what is written. Resistance voiced in a corridor escapes it, as does support shown in a meeting and then contradicted by actions. Its C remains a hypothesis that the team checks before recording it. The D remains a decision of the team and the sponsor, since it sets what the project expects of a person.
That leaves the data. In an internal project, most stakeholders are employees of the organisation. For a private employer, art. 328b of the Code of Obligations (CO) limits the data it processes about them to data concerning their suitability for the job or needed to perform the employment contract; in a public body, the cantonal law on staff or on data protection sets this framework. Art. 26 of Ordinance 3 to the Employment Act (ArGV 3) prohibits systems designed to monitor their behaviour at the workplace. A model that reads their emails and messaging to score their attitude leaves the first framework and comes close to the second. The team therefore limits the model's input to minutes and documents written for the project. Before any analysis, it consults the data protection advisor and uses only a tool the organisation has approved.
Examples
The matrix below shows that the same gap can call for two different kinds of work. A Spitex home care organisation is replacing its software for planning care rounds. The new software calculates the rounds that the nursing team leaders used to build by hand.
Director of nursing (sponsor)
secured the budget; leaves the team to settle scheduling conflicts alone
Nursing team leaders
fear losing control of the rounds
IT department
follows the project without assigning anyone to it
Cantonal data protection commissioner
not yet consulted; the software processes health data
Staff committee
informed, no position
C current level · D desired level
The team leaders know the project and oppose it, because it takes away a task that was their expertise. The cantonal data protection commissioner, at the same two-level gap, does not yet know about the project.
| Stakeholder | Target step | Action | Owner | Deadline |
|---|---|---|---|---|
| Nursing team leaders | Resistant to supportive | Hear the objection in a session; put the team leaders in charge of configuring the rounds rules. | Project manager | Before the software is configured |
| Cantonal data protection commissioner | Unaware to neutral | Present the project in a session; hand over the file describing how health data are processed. | Business analyst | Before the contract with the supplier is signed |
Visualisations
The columns keep the same order from one version to the next, from "unaware" to "leading", so that two versions can be compared. C and D are written as letters, even when a colour distinguishes them: the matrix is often printed in black and white for the committee. The header carries the review date. When a drawing shows two versions, the old C appears faded and the new one solid, linked to the old one by an arrow.
Three months later, the team repeats the assessment. Three gaps have closed. The team leaders' gap has narrowed, from "resistant" to "neutral". One gap has opened: the staff committee has moved to "resistant" since the new rounds changed working hours. Its D stays at "neutral", and an action enters the plan.
Review 1 · Review 2, three months later
Director of nursing (sponsor)
secured the budget; leaves the team to settle scheduling conflicts alone
Nursing team leaders
fear losing control of the rounds
IT department
follows the project without assigning anyone to it
Cantonal data protection commissioner
not yet consulted; the software processes health data
Staff committee
working hours changed by the new rounds
C1 before (faded) · C2 after (solid) · D desired level
Cost
| Dimension | Level | Rationale |
|---|---|---|
| Preparation | Low | The matrix takes its rows from the stakeholder register. |
| Execution | Medium | Each C needs facts, sometimes an interview; the Ds need a session with the sponsor. |
| Documentation | Low | One page, one sentence per C and per D, one dated version per review. |
Tooling
A spreadsheet is enough: one row per stakeholder, five level columns, one column for the fact that justifies the C, another for the expectation that justifies the D and a last one for the date. Conditional formatting colours the cells between C and D. One tab per review keeps the history.
In a session, a whiteboard, physical or online such as Miro or Mural, lets two assessors place their letters separately before comparing them.
Project management or stakeholder management software is suitable if it carries two attributes per stakeholder, "current level" and "desired level", with access rights narrower than those of the rest of the register. It suits an organisation that already keeps its register in such a tool.
Sources
- Project Management Institute (2025), A Guide to the Project Management Body of Knowledge (PMBOK Guide), 8th edition: §2.5.2.2 Plan Stakeholder Engagement and §2.5.2.6 Monitor Stakeholder Engagement for the place of the matrix among the data representation tools; Section 5 Tools and Techniques, entry "Stakeholder engagement assessment matrix" and figure 5-23 for the definition, the five levels, the C and D marks and the role of the gap; entry "Stakeholder analysis" for the attitude recorded per stakeholder; entry "Communication styles assessment" for the assessment that can follow the engagement assessment; Appendix X3, table X3-1, for stakeholder sentiment analysis.