Workplace Communication
Build a Stakeholder Map Before Your Project Gets Conflicting Feedback
Map affected groups, experts, confirmed authority, and information recipients before requesting feedback. Follow a filled map and three tailored risk updates into a bounded prototype decision.
Build a stakeholder map around the change you are proposing: who will use it, whose expertise establishes constraints, who can decide, and who needs the result. For each person or group, record the question they must answer and the point when their input can still change the work. Agree how conflicting feedback will reach a decision before you circulate a design.
A useful map identifies missing involvement and distinguishes consultation from approval. This LandOffer workplace guide develops a fictional museum software example, including a filled map, three audience-specific risk updates, and the disposition of competing requests. The examples illustrate a working method; no museum, employee, consultation, or approval described here is real.

Start with a bounded change and a decision
“Stakeholders for the new system” is too broad to guide a review. State the particular change, the stage of work, and the decision being prepared. Someone may approve a prototype while having no authority to authorize production deployment, procurement, or a records-policy exception.
In this example, the team is changing an internal museum collection-location request screen. Staff currently read a location code and separate handling instructions. The proposed screen would bring those details together. This review will choose what context to show in a prototype using invented records. It will not change collection permissions, authorize object movement, or release software to staff.
That boundary determines whose involvement is necessary now. Collection assistants understand the request task. A registrar can explain handling constraints. A designer can examine presentation options. Engineering can establish what the prototype can demonstrate. A product owner can choose the prototype direction within this agreed scope.
The Association for Project Management's stakeholder guidance includes people with an interest or role in a project and people affected by it. Its analysis guidance connects understanding different perspectives to planning, delivery, and governance. Use that broad starting point to find affected groups before narrowing participation for a particular decision.
Separate impact, expertise, and authority
Start with the work people do. Who enters information, interprets it, acts on it, maintains it, or answers questions when it is wrong? Follow dependencies beyond the people already invited to the project channel. A quiet downstream team may experience the consequences of a decision without attending its planning meetings.
Record why each group belongs on the map and what it can contribute. An affected person supplies task evidence. An expert checks a constraint. A decision owner chooses among feasible alternatives. An information recipient needs an accurate account of what changed. One person can occupy several roles, but write those roles separately when they involve different authority.
Atlassian's DACI play distinguishes a driver, an approver, contributors, and informed people. It defines one approver for a decision within that method. This can help clarify a bounded decision, but it does not replace your organization's required approvals. A product decision cannot waive a security restriction simply because a product owner appears in the approver column.
Confirm decision rights directly rather than inferring them from seniority. Confirm the named person's remit and availability. If nobody can confirm who decides, record that as an open governance question rather than distributing a document that quietly assigns authority.
A filled stakeholder map for the prototype
The following map is fictional and prepared before the first prototype review. Mina coordinates the review and owns map revision 1. Each named participant has a different task; the entries are illustrative commitments inside the example, not evidence of real consent.
| Person or group | Connection to the change | Required input and timing | Decision boundary and communication |
|---|---|---|---|
| Collection assistants, represented by Hugo and Tess | Affected users who prepare requests | Walk through two invented requests by Tuesday noon; identify missing context and confusing terms | Their examples inform the design; two representatives do not establish agreement from all assistants. Mina sends the revised task summary for correction. |
| Farah, registrar | Subject expert for handling instructions | Check which warnings must be visible before a request proceeds, by Tuesday 15:00 | Establish the documented handling constraint; does not select the visual layout. Review the options note and flag misinterpretation. |
| Marco, designer | Contributor responsible for presentation alternatives | Prepare two layouts against the same task and constraint, by Wednesday 10:00 | Recommend a layout and explain trade-offs; does not approve deployment. Comment in the shared options note. |
| Lin, engineer | Contributor checking prototype feasibility | Confirm what invented-data interactions can be demonstrated, by Wednesday 10:00 | Own the feasibility statement; do not infer production readiness from a working prototype. Add limitations to the note. |
| Vera, product owner | Approver for prototype direction | Choose a feasible direction at Wednesday 14:00 review after inputs are available | Can select prototype scope within the agreed allocation; cannot authorize handling-policy exceptions or release. Record decision and unresolved checks. |
| Accessibility reviewer, not yet assigned | Needed expertise before a later usability session | Mina must obtain a named reviewer and review slot by Thursday noon | An unfilled role is a gap, not completed consultation. No accessibility clearance is claimed. |
| Service desk, coordinated by Arun | Information recipient at this prototype stage | Receive Friday summary of direction, limits, and next review | No support-process change or staff announcement is requested yet. Reclassify as contributor before planning rollout. |
This map includes an uncomfortable blank: the accessibility reviewer is not assigned. Keeping this gap visible gives Vera a reason to limit approval to the internal prototype rather than staff testing. If that review is required earlier under the team's actual process, move the gate earlier; the table is not permission to bypass it.
The map also records whose perspectives are missing. Hugo and Tess work the daytime request desk. They cannot speak for occasional evening staff. Mina lists evening coverage as a sampling gap for the later usability plan instead of naming a representative who has not agreed to participate.
Set a feedback route before sending the design
Tell reviewers what input you need and where it belongs. A short intake note can say: “For each comment, identify the task affected, the evidence or constraint behind it, and the consequence if we leave the design unchanged. Add it to the options note by your agreed review time.” Give people a way to submit an example privately when it contains information unsuitable for the shared document.
The driver groups comments by the question they address. “Show all instructions” and “keep the screen compact” may appear incompatible, but they may express different needs: avoiding missed handling conditions and reducing scanning effort. Preserve both underlying needs before debating a layout.
Distinguish a preference from a required constraint, and verify the basis of the constraint with its responsible owner. Do not downgrade a warning to a cosmetic preference because it is inconvenient. Conversely, an expert's layout preference should not acquire policy authority merely because the expert also owns a genuine constraint.
A late comment can still reveal a material issue. Agree that new safety, access, or feasibility evidence triggers review of the affected decision. An unrelated enhancement goes into a separate scope discussion. This rule describes a route for reconsideration; it does not promise that a deadline automatically outweighs new evidence.

Give three audiences the same risk with different asks
Atlassian's stakeholder communication play recommends identifying who needs involvement or information, then setting channels, cadence, and expectations. Apply that distinction to the work each audience can actually do. Tailoring a message should change its emphasis and request while keeping the underlying risk consistent.
Here is the completed communication set from Tuesday in the fictional case. The common risk is that a compact layout could hide handling context needed before someone proceeds. No actual object movement is part of the prototype.
To the affected-user representatives: “Hugo and Tess, your walkthrough showed that opening a separate instructions page interrupts request preparation. We are comparing two ways to keep context near the location code. By Tuesday noon, please mark the exact point in each invented request when you need the warning. Your examples inform the prototype; we have not selected a layout or confirmed evening-staff needs.” Mina posts this in the walkthrough thread with links to both invented requests.
To the subject expert: “Farah, the compact option could hide a handling warning until expansion. Please identify by Tuesday 15:00 which information must be visible before proceeding, and point to the applicable instruction in our example requirements. We need the constraint, rather than a general preference for more text. If neither layout satisfies it, we will revise the options before Vera's review.” Mina adds this request to the shared options note, keeping any sensitive reference material in its approved location.
To the prototype approver: “Vera, Wednesday's choice depends on Farah's visibility requirement and Lin's feasibility check. The current compact option may conceal necessary context. Marco will submit revised alternatives by Wednesday 10:00. Please choose a feasible prototype direction at 14:00 or defer that choice if the constraint remains unclear. This review does not authorize release, and an accessibility reviewer is still unassigned.” Mina places the decision request at the top of the review note.
These updates share a risk, sequence, and boundary. They ask for task evidence, a constraint, and a decision respectively. Each message names an input, its owner, and its timing; the approver also sees the missing review and release boundary.
Route competing comments to a bounded decision
In the completed fictional review, Hugo asks for instructions beside the location code; Marco initially recommends a compact screen with all detail collapsed. Farah checks the example requirements and confirms that the handling warning must be visible before proceeding, while supplementary background can be expanded. Lin confirms that both revised layouts can be represented with invented data.
Mina records the remaining trade-off: show the entire background by default, or show the warning and make supplementary context expandable. Marco prepares both versions, preserving Farah's warning requirement in each. Hugo prefers the full background; Tess prefers the shorter version after walking through the same invented task. That disagreement remains in the note rather than becoming “users want the shorter screen.”
At Wednesday 14:00, Vera chooses the second option for the internal prototype: visible warning, expandable supplementary context, and no production release. The rationale is to explore a shorter scanning path without hiding the required warning. This is a hypothesis to examine, not measured evidence that the design is faster or easier.
The disposition also retains three open items. Mina must secure the accessibility reviewer by Thursday noon. Hugo will help identify an evening-staff participant for the later research plan, without promising that person's participation. Lin will document the difference between the invented-data prototype and production behavior. No usability result or rollout approval exists yet.
Mina sends contributors the revised options note with Vera's decision, the warning requirement, and the three open items. Hugo confirms that his preference is recorded but does not endorse the chosen layout. That is an accurate receipt: the feedback was considered, a bounded decision was made, and agreement remains partial.
Maintain the map as the work changes
Revisit the map when scope, dependencies, staffing, or the decision stage changes. Before rollout planning, the service desk may need to contribute rather than receive a summary. Production access owners may become necessary approvers. Update the specific relationship and question instead of copying the prototype map into a release checklist unchanged.
For your own project, begin with one upcoming decision. List affected groups, the evidence each can provide, required constraints, and confirmed authority. Send each named participant their proposed role for correction, mark unconfirmed involvement openly, and establish the feedback route before requesting comments. Save the revised map and options note together, retaining unconfirmed roles, the bounded decision, dissent, and owners of open questions. Further workplace guides are available through LandOffer.
Sources and Further Reading
- APM: Stakeholder engagement — scope of stakeholder identification and the use of analysis in planning and governance.
- Atlassian: Stakeholder Project Communication Plan — audiences, channels, cadence, and involvement expectations.
- Atlassian: DACI Decision-Making Framework — a method for distinguishing decision roles; apply within confirmed local governance.


