Gsd-skill-creator stakeholder-communication

Stakeholder identification, communication planning, and facilitation for project teams. Covers power/interest grid mapping, communication plans (audience, frequency, channel), status reporting (RAG, earned value, burn charts), RACI matrices, meeting facilitation, conflict resolution, change management (Kotter's 8 steps), psychological safety (Edmondson), managing up, and remote/distributed team communication. Integrates GSD's "docs are the story" principle as a communication strategy.

install
source · Clone the upstream repo
git clone https://github.com/Tibsfox/gsd-skill-creator
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/Tibsfox/gsd-skill-creator "$T" && mkdir -p ~/.claude/skills && cp -r "$T/examples/skills/project-management/stakeholder-communication" ~/.claude/skills/tibsfox-gsd-skill-creator-stakeholder-communication && rm -rf "$T"
manifest: examples/skills/project-management/stakeholder-communication/SKILL.md
source content

Stakeholder Communication

Projects do not fail because of technology. They fail because of people — miscommunication, misaligned expectations, unresolved conflicts, invisible stakeholders who emerge at the worst moment with the power to block everything. Communication is the connective tissue of project management. Without it, a technically perfect project delivers the wrong thing to the wrong people at the wrong time.

Simon Sinek's insight applies directly: people don't buy what you build, they buy why you build it. A status report that says "Sprint 7 complete, 34 points delivered" communicates nothing to an executive who needs to know "Are we going to make the Q3 launch?" The same information, framed differently, serves different audiences. This skill covers the frameworks for identifying who needs what, building communication plans, running effective meetings, resolving conflicts, and managing organizational change.

Agent affinity: sinek (pedagogy/leadership, "Start With Why"), brooks (organizational dynamics, communication overhead)

Concept IDs: bus-stakeholder-theory, bus-corporate-governance, bus-business-ethics, crit-argument-construction, crit-evidence-evaluation

Stakeholder Identification and Mapping

The Power/Interest Grid

Classify stakeholders along two dimensions:

  • Power: Ability to influence project decisions, resources, or outcomes
  • Interest: How much the stakeholder cares about the project's outcome
High InterestLow Interest
High PowerManage Closely — Key players. Regular engagement, involve in decisions.Keep Satisfied — Dangerous if surprised. Periodic high-level updates.
Low PowerKeep Informed — Enthusiastic supporters or worried users. Regular updates, feedback channels.Monitor — Minimal effort. Available information if they seek it.

Worked Example — E-Commerce Platform Rewrite

StakeholderPowerInterestStrategy
VP of EngineeringHighHighManage closely — weekly 1:1, involved in architecture decisions
CFOHighLowKeep satisfied — monthly cost summary, flag budget variances early
Customer support teamLowHighKeep informed — sprint demos, early access to staging
Legal/complianceHighMediumKeep satisfied — notify on data handling changes, PII flow reviews
End users (beta group)LowHighKeep informed — changelog emails, feedback survey per release
Marketing teamMediumMediumKeep informed — feature announcements 2 weeks before launch
Infrastructure team (DevOps)MediumHighManage closely — daily standups, shared Kanban board
Board of directorsHighLowKeep satisfied — quarterly executive summary

Hidden stakeholders. The most dangerous stakeholders are the ones you did not identify. Common hidden stakeholders: security/audit teams (veto power over production deployment), data protection officer (GDPR implications), downstream service owners (API contract changes), executive assistants (gatekeepers to executive time).

Salience Model (Mitchell, Agle, Wood)

For more nuanced classification, the salience model uses three attributes:

  • Power: Ability to impose will
  • Legitimacy: Perceived appropriateness of the stakeholder's involvement
  • Urgency: Time sensitivity of the stakeholder's claim

Stakeholders with all three attributes are definitive — they require immediate, full engagement. Stakeholders with two are expectant — they will escalate if ignored. Stakeholders with one are latent — they may emerge as expectant if conditions change.

Communication Planning

The Communication Matrix

For each stakeholder or stakeholder group, define:

ElementQuestion
WhoWhich stakeholder or group?
WhatWhat information do they need?
WhenHow frequently? Triggered by what events?
HowChannel: email, meeting, dashboard, Slack, document?
WhyWhat decision or action does this communication enable?

Worked Example — Communication Plan

WhoWhatWhenHowWhy
VP EngineeringSprint progress, blockers, risksWeekly (Friday)15-min 1:1 + written summaryResource allocation, escalation decisions
CFOBudget vs. actual, forecast to completeMonthlyEmail with 1-page dashboardBudget approval, investment decisions
Dev teamSprint goals, dependencies, blockersDaily15-min standupSynchronization, impediment removal
Customer supportUpcoming features, known issuesPer sprintDemo + release notesPrepare for customer inquiries
End users (beta)New features, bug fixesPer releaseChangelog email + in-app notificationFeedback collection, adoption
BoardProject status (RAG), strategic alignmentQuarterly2-slide executive summaryStrategic governance

Channel Selection

Rich channels (face-to-face, video call) for: ambiguous topics, conflict resolution, relationship building, bad news.

Lean channels (email, documents, dashboards) for: routine updates, reference information, audit trails, asynchronous consumption.

Brooks's communication overhead. In The Mythical Man-Month, Brooks showed that communication paths grow as n(n-1)/2 for n people. A team of 5 has 10 communication paths; a team of 10 has 45. This is why Scrum limits team size and why communication planning becomes critical as organizations scale. Every person added to a project increases communication cost quadratically.

Status Reporting

RAG Status (Red/Amber/Green)

The simplest and most widely understood status format:

  • Green: On track. No action needed.
  • Amber: At risk. Mitigation underway. Needs attention.
  • Red: Off track. Requires immediate intervention or escalation.

RAG rules:

  • Never report Green when you know Amber is coming next week. Report Amber now with "trending toward Green" or "trending toward Red."
  • A status that has been Amber for 4 consecutive weeks is effectively Red — the mitigation is not working.
  • Red is not a failure; it is an honest signal. Punishing Red reports guarantees dishonest Green reports.

Worked Example — Sprint Status Report

Project: Commerce Platform Rewrite
Sprint: 7 of 12
Date: 2026-04-10
Overall: AMBER

Schedule: GREEN — 34/36 points delivered (94%)
Scope: AMBER — Payment gateway integration blocked by vendor API delay.
         Vendor committed to API delivery by April 15. If missed, checkout
         feature slips to Sprint 9. Contingency: mock API for continued
         development.
Quality: GREEN — 0 P1 defects. 3 P3 defects in backlog. Test coverage 87%.
Budget: GREEN — $142K spent of $200K budget. Forecast: $185K at completion.

Key decisions needed:
1. Approve contingency plan for payment API delay (VP Engineering by April 12)
2. Confirm beta user group expansion to 500 users (Product Owner by April 14)

Risks surfaced this sprint:
- R07: Payment API documentation incomplete (NEW, Score: 12)
- R03: Staging environment — RESOLVED (provisioned April 8)

Earned Value Management (EVM)

EVM provides objective, quantitative project status by comparing planned work to completed work and actual cost.

Core metrics:

  • PV (Planned Value): Budgeted cost of work scheduled. What you planned to have done by now.
  • EV (Earned Value): Budgeted cost of work performed. What you actually have done, valued at planned rates.
  • AC (Actual Cost): Actual cost of work performed. What it actually cost to do what you did.

Derived metrics:

  • SPI (Schedule Performance Index) = EV / PV. SPI < 1 means behind schedule.
  • CPI (Cost Performance Index) = EV / AC. CPI < 1 means over budget.
  • EAC (Estimate at Completion) = BAC / CPI. Forecasts total project cost based on current efficiency.

Worked example.

MetricValueInterpretation
BAC (Budget at Completion)$200,000Total project budget
PV (end of Sprint 7)$117,000Should have completed 58.5% of scope
EV (end of Sprint 7)$108,000Actually completed 54% of scope
AC (end of Sprint 7)$112,000Spent $112K to achieve 54%
SPI0.928% behind schedule
CPI0.964% over budget
EAC$208,333Forecasted overrun of $8,333

Translation for the CFO: "We are slightly behind schedule and slightly over budget. At current rates, we'll overshoot by about $8K. The payment API delay is the primary driver. If resolved by April 15, we expect to recover the schedule variance by Sprint 9."

Burn Charts

Burn-down chart. Shows remaining work over time. The Y-axis is remaining story points (or tasks); the X-axis is time (days within a sprint, or sprints within a release). A healthy burn-down trends toward zero by the sprint end.

Burn-up chart. Shows completed work and total scope over time. Two lines: scope (total work) and completed (done). The gap between them is remaining work. Burn-up charts reveal scope changes — if the scope line keeps rising, the project is absorbing more work than planned.

Cumulative flow diagram (CFD). Shows the total number of items in each state (To Do, In Progress, Done) over time. The vertical distance between bands represents WIP; the horizontal distance represents lead time. A widening "In Progress" band signals a bottleneck.

RACI Matrix

Defines roles for each deliverable or decision:

  • Responsible — Does the work
  • Accountable — Ultimately answerable (one person per item)
  • Consulted — Provides input before the decision
  • Informed — Notified after the decision

Worked Example — Feature Delivery RACI

ActivityDev LeadProduct OwnerQA LeadArchitectVP Eng
User story definitionCA/RCII
Technical designA/RICCI
ImplementationRIICI
Test planningCIA/RII
Sprint demoRARII
Release decisionCRCCA

Common mistakes:

  • Multiple "A" for one item (nobody is accountable)
  • No "R" for an item (nobody does the work)
  • Everyone is "C" on everything (decision-by-committee; nothing gets decided)

Meeting Facilitation

The Meeting Tax

Every meeting costs: (number of attendees) x (duration) x (average hourly rate). A 1-hour meeting with 8 people at $75/hour costs $600. If the meeting could have been an email, that $600 was waste (Lean principle 1: eliminate waste).

Meeting Hygiene

  1. Purpose. Every meeting has a stated purpose. If you cannot articulate it in one sentence, cancel the meeting.
  2. Agenda. Distributed in advance. Timed sections. Clear outcomes for each section: decision, information, or brainstorm.
  3. Attendees. Only people who are R, A, or C on the agenda items. Everyone else gets notes afterward (I in RACI terms).
  4. Time-box. Start on time, end on time. If discussion extends, schedule a follow-up rather than holding the room hostage.
  5. Notes and actions. Who, what, by when. Distributed within 24 hours. Actions tracked in the project tool.

Facilitation Techniques

Round-robin. Go around the room; everyone speaks once. Prevents dominant voices from monopolizing.

Fist of Five. Quick consensus check. Everyone holds up 0-5 fingers: 5 = strong support, 3 = can live with it, 1 = serious concerns, 0 = veto. Anything below 3 triggers discussion.

Parking lot. Off-topic but valuable items go on a visible list. Addressed after the meeting or in a follow-up session. Prevents tangent spirals without losing ideas.

Silent writing. Before discussion, everyone writes their input on sticky notes or a shared doc for 3-5 minutes. Then share. This prevents anchoring to the first speaker's opinion and gives introverts equal voice.

Conflict Resolution

Thomas-Kilmann Conflict Modes

ModeAssertiveCooperativeWhen to use
CompetingHighLowEmergency decisions, non-negotiable standards
CollaboratingHighHighImportant issues where both sides have valid concerns
CompromisingMediumMediumTime-pressured decisions where a quick resolution matters more than an optimal one
AvoidingLowLowTrivial issues, or when emotions need to cool before productive discussion
AccommodatingLowHighWhen you are wrong, or when the relationship matters more than the issue

Default to collaborating for technical disagreements. "Let's understand both approaches, test both if needed, and decide based on evidence." Competing ("we're doing it my way") and avoiding ("let's not talk about it") both accumulate unresolved tension.

Worked Example — Architecture Disagreement

Situation. The backend lead wants microservices. The frontend lead wants a monolith. The deadline is 3 months.

Poor resolution (competing). The architect picks microservices because they outrank the frontend lead. The frontend lead is demoralized and disengaged.

Better resolution (collaborating).

  1. Both leads present their rationale with specific criteria: deployment frequency, team structure, operational complexity, time to market.
  2. The team scores each approach against the criteria.
  3. Data reveals: microservices win on deployment independence but lose on time-to-market (need service mesh, API gateway, etc.).
  4. Agreed approach: start with a well-structured monolith (modular boundaries, clean interfaces) with a documented migration path to microservices post-launch.
  5. Both leads contributed to the decision; neither was overruled.

Change Management — Kotter's 8 Steps

John Kotter's framework for leading organizational change:

  1. Create urgency. Why must we change? What happens if we don't? Data, competitor moves, customer pain.
  2. Form a guiding coalition. Assemble people with power, expertise, credibility, and leadership to drive change.
  3. Create a vision for change. A clear, compelling picture of the future state. Simple enough to explain in 5 minutes.
  4. Communicate the vision. Use every channel. Repeat 10x more than you think necessary. Walk the talk — behavior must match the message.
  5. Empower action. Remove obstacles: organizational structures, processes, systems, and people who block progress.
  6. Create short-term wins. Visible, unambiguous improvements within 6-12 months. Wins build momentum and silence critics.
  7. Build on the change. Don't declare victory too early. Use credibility from wins to tackle bigger changes.
  8. Anchor in culture. Connect the change to organizational success. "This is how we work now."

Worked Example — Adopting GSD in an Organization

StepAction
1. Urgency"Our last 3 projects missed deadlines by 30-60%. Current processes are not working."
2. CoalitionEngineering director + 2 respected tech leads + PM manager
3. Vision"Structured phases with autonomous execution. Ship every 2 weeks, not every quarter."
4. CommunicateDemo GSD on a small pilot. Share results weekly.
5. EmpowerGive pilot team authority to skip legacy approval gates.
6. Short-term winsPilot delivers first feature in 2 weeks vs. typical 8 weeks.
7. Build on changeExpand to 2 more teams. Publish playbook from pilot learnings.
8. AnchorGSD phases become the standard project template. Handoffs and retrospectives are mandatory.

Psychological Safety (Amy Edmondson)

Edmondson's research at Harvard demonstrates that team performance correlates with psychological safety — the belief that one can speak up without punishment. High-performing teams are not error-free; they surface errors faster.

Observable behaviors in psychologically safe teams:

  • People ask questions without fear of looking stupid
  • Mistakes are discussed openly, not hidden
  • Team members challenge ideas regardless of hierarchy
  • Bad news travels up quickly

Observable behaviors in psychologically unsafe teams:

  • People stay silent in meetings, then complain in hallways
  • Problems are discovered late because nobody raised the alarm
  • Status reports are always Green until the project explodes
  • Post-mortems find blame, not causes

Building psychological safety:

  • Frame work as learning problems, not execution problems
  • Acknowledge your own fallibility ("I might be wrong about this")
  • Model curiosity — ask questions, don't make statements disguised as questions
  • Thank people who surface problems early
  • Respond to failure with "What can we learn?" not "Whose fault is this?"

GSD connection. GSD's blameless session handoffs serve this function. HANDOFF-SESSION-*.md files record what happened honestly without assigning blame. The retrospective pattern (carry-forward from session to session) normalizes learning from mistakes as the default mode. (Cross-reference: retrospective-learning skill.)

Remote and Distributed Communication

Challenges

  • Time zones. A team spanning US Pacific to India has zero overlapping business hours without someone working outside normal hours.
  • Reduced bandwidth. No body language, no hallway conversations, no whiteboard sketches.
  • Isolation. Remote workers miss informal context that co-located teams absorb osmotically.
  • Tool fatigue. Slack + email + Zoom + Jira + Confluence + wiki = context scattered across 6 systems.

Practices

Default to asynchronous. Write it down. Record meetings. Put decisions in documents, not chat. Asynchronous communication respects time zones and creates an audit trail.

Overlap windows. Define 2-4 hours per day where all team members are available for synchronous discussion. Protect this window — do not schedule non-team meetings during it.

Documentation as communication. GSD's "docs are the story" principle applies directly. If a decision is not written down, it did not happen. STATE.md, HANDOFF documents, and PLAN.md files are communication artifacts — they make invisible work visible to anyone who reads them, regardless of time zone or attendance at the original meeting.

Camera-optional, presence-required. Do not mandate cameras (fatigue, bandwidth, privacy). Do require active presence — responses to questions, participation in polls, visible engagement.

Managing Up

What managing up means. Ensuring your manager and leadership have the information they need to support your project. It is not manipulation — it is professional communication directed upward.

The no-surprises rule. Bad news must travel up before it becomes a crisis. An executive who learns about a problem from a customer complaint will never forgive the team that hid it.

Framing for executives:

  • Lead with impact, not detail. "We may miss the Q3 date" not "The Kubernetes ingress controller configuration is incompatible with our TLS termination strategy."
  • Present options, not problems. "Option A: reduce scope (low risk, lower value). Option B: extend timeline 2 weeks (medium risk, full value). Option C: add 2 contractors (high cost, high risk). I recommend B."
  • Quantify. "80% probability of on-time delivery" not "we think we'll make it."

When to Use / When NOT to Use

When to use structured communication planning

  • Multiple stakeholders with different information needs
  • Distributed or remote teams
  • Projects with organizational change impact
  • Any situation where "I thought someone told them" has caused problems before

When NOT to use structured communication planning

  • A single developer working on a well-defined task (communicate with yourself via comments and commit messages)
  • When the plan becomes more elaborate than the project itself
  • When communication plans substitute for actual communication ("We have a RACI matrix, why is nobody talking to each other?")

Cross-References

  • sinek agent: Leadership communication, purpose-driven messaging, "Start With Why" framework.
  • brooks agent: Department chair. Communication overhead scaling (n(n-1)/2), organizational structure as communication constraint (Conway's Law).
  • deming agent: Quality communication — making processes visible, statistical process control as a communication mechanism.
  • lei agent: Lean communication — eliminating waste in meetings and handoffs, value stream mapping for information flow.
  • retrospective-learning skill: Retrospective formats for team reflection, psychological safety practices, organizational learning.
  • risk-management skill: Risk communication — reporting risks to stakeholders, escalation triggers, risk register as communication artifact.
  • agile-methods skill: Scrum ceremonies as structured communication (standup, review, retrospective), Kanban boards as visual communication.

References

  • Sinek, S. (2009). Start With Why. Portfolio/Penguin.
  • Brooks, F. P. (1975/1995). The Mythical Man-Month. Addison-Wesley.
  • Edmondson, A. C. (2018). The Fearless Organization. Wiley.
  • Kotter, J. P. (1996). Leading Change. Harvard Business Review Press.
  • Thomas, K. W. & Kilmann, R. H. (1974). Thomas-Kilmann Conflict Mode Instrument. Xicom.
  • Mitchell, R. K., Agle, B. R., & Wood, D. J. (1997). "Toward a Theory of Stakeholder Identification and Salience." Academy of Management Review, 22(4), 853-886.
  • PMI. (2021). A Guide to the Project Management Body of Knowledge (PMBOK Guide). 7th edition. Project Management Institute.
  • DeMarco, T. & Lister, T. (2013). Peopleware: Productive Projects and Teams. 3rd edition. Addison-Wesley.