Foundations Beginner 6 min read

Stakeholders and Comms: RACI, Audiences, Cadence

How to map stakeholders to a RACI matrix, decide who hears what at which cadence, and write the comms plan that prevents 'I didn't know' meetings.

Table of contents
  1. When does formal stakeholder mapping start paying back?
  2. What is the cost of not mapping stakeholders?
  3. What does the minimal RACI matrix look like?
  4. What does the comms plan look like?
  5. How does this scale to multi-team?
  6. What failure modes does the comms plan introduce?
  7. When is RACI overkill?
  8. Where should you go from here?

The reason most projects feel chaotic is not that the engineering work is hard - it is that nobody mapped who needs to know what. Some stakeholder gets surprised, drops in to "help", and an hour of explanation eats the team's afternoon. This chapter shows the two artifacts - RACI matrix and comms plan - that prevent that hour from happening.

When does formal stakeholder mapping start paying back?

Three signals.

More than 5 people care about the outcome. Below five, you know everyone by name and a Slack channel covers comms. Above five, you cannot keep mental track of who heard what when.

Cross-functional dependencies. Sales wants the launch date locked, Legal needs to review terms, Security needs to sign off, Marketing needs the brand assets. Each is a different audience with different cadence needs.

Visible delivery to a customer or executive. When a VP or client will judge the work, they need to feel informed - which means a comms plan, not ad-hoc updates.

If the project is internal, small team, and you can name every stakeholder in one breath, skip the formality and use a Slack channel.

What is the cost of not mapping stakeholders?

Three classic failure modes.

Surprise meetings. A VP reads about your project from a sales deck and pulls you in for an "alignment chat". The chat costs the team a half-day of context switching. With a comms plan, the VP was already on the monthly digest list and not surprised.

Decision rework. Engineering ships a feature; Legal reviews post-ship and demands changes. Two weeks of work redone. Mitigation: Legal was Consulted on scope, not Informed.

Loudest voice wins. A single stakeholder dominates every review because no other voices are formally invited. The PM ends up serving one person's preferences instead of the project's priorities.

What does the minimal RACI matrix look like?

For a 5-person team, one markdown table per project:

# RACI — {{ Project Name }}

R = Responsible (does the work)  -  many
A = Accountable (signs off)      -  exactly one
C = Consulted (asked before)     -  some
I = Informed (told after)        -  many

| Item / Decision           | Sponsor | TL | PM | EM | QA | Design | Legal | Sales |
|---------------------------|---------|----|----|----|----|--------|-------|-------|
| Project go / no-go        | A       | C  | R  | C  | -  | -      | C     | I     |
| Architecture (records DB) | I       | A  | I  | C  | -  | -      | -     | -     |
| Feature scope             | A       | C  | R  | C  | C  | C      | -     | C     |
| Launch date               | A       | C  | R  | C  | C  | -      | I     | C     |
| User-facing copy          | I       | -  | C  | -  | -  | A      | C     | C     |
| Data retention policy     | I       | C  | C  | -  | -  | -      | A     | -     |
| Hire decisions            | I       | C  | -  | A  | -  | -      | -     | -     |
| Vendor selection          | A       | C  | R  | -  | -  | -      | C     | -     |

The single biggest error is having two A's per row. The matrix is useless if "we both decide". When you cannot pick one, escalate to the sponsor and force the answer.

What does the comms plan look like?

Audiences mapped to cadences and channels:

# Comms Plan — {{ Project Name }}

## Audiences

| Audience          | Members              | Cadence    | Channel              | Format               |
|-------------------|----------------------|------------|----------------------|----------------------|
| Active team       | Engineers, TL, QA, Design | Daily | #project-{{name}} | Standup notes        |
| Project review    | Above + EM, PM      | Weekly     | 30-min sync + doc    | Status report (RAG) |
| Stakeholders      | Above + Sales, Legal, Marketing | Bi-weekly | Email digest | 1-page exec summary |
| Executive sponsor | VP / CTO            | Monthly    | 15-min sync          | Top 3 risks + ETA    |
| Customer / public | Customer success    | At launch  | Release notes        | What changed         |

## Escalation triggers (skip cadence)

- Schedule slip > 1 week: notify sponsor within 24h
- Critical risk realised: notify sponsor immediately
- Scope cut: notify all C and A before deciding
- Production incident: see [incident chapter](/tech-pm/incident-and-rollback)

The escalation triggers are the most-skipped part. They are what turns a comms plan from theatre into actual risk management. When something goes wrong, you have already pre-agreed who hears.

How does this scale to multi-team?

flowchart TB
    Active[Daily: active team<br/>standup] --> Project[Weekly: project review<br/>status report]
    Project --> Cross[Bi-weekly: cross-team<br/>roadmap sync]
    Cross --> Org[Monthly: org review<br/>top initiatives]
    Org --> Exec[Quarterly: exec review<br/>portfolio]

Each layer summarises the one below. The status report from one project becomes one paragraph in the cross-team sync; the cross- team sync becomes one slide in the org review. The skill is ruthless compression: an executive does not want your full RAID log; they want the three risks that need their action.

What failure modes does the comms plan introduce?

When is RACI overkill?

Two cases.

Two-person project. RACI is overhead. Both people decide everything together; the comms plan is a Slack DM.

Highly autonomous team with clear charter. A platform team that owns its area completely does not need RACI for routine work; only for cross-team decisions.

The matrix earns its cost when accountability is genuinely unclear

Where should you go from here?

Next chapter: scope and tradeoffs - how to respond when stakeholders ask for more, faster, with the same budget. After that, the methodology chapters (Scrum essentials) cover the cadence that runs inside the team.

Frequently asked questions

Why exactly one Accountable per item?
Because if two people are accountable, neither is. The classic failure mode: scope is 'co-owned' by Product and Engineering, neither makes the cut decision when the deadline approaches, the project ships late. RACI's design is one A per row; the C and I columns can have many. Forcing the choice surfaces ambiguity early.
How is RACI different from the role table from chapter 1?
RACI is per item (a feature, a deliverable, a decision), the role table from chapter 1 is per project. Same person can be R on one item and A on another. Use the role table for the project README; use RACI for big decisions, deliverables, or ambiguous areas.
What cadence should I use for each audience?
Three cadences cover most: daily for the active team (standup), weekly for Consulted stakeholders (status report), monthly or quarterly for Informed leadership (executive summary). When an issue lands, escalate immediately to Accountable - skipping cadence is the right move when severity demands it.
How do I avoid drowning in stakeholder meetings?
Default to async (written status, recorded demo) and reserve sync time for decisions. A weekly 30-min stakeholder review beats five 1-on-1s. The status report chapter covers the format that absorbs 80% of stakeholder questions before they need a meeting.