Skip to content This library is free, offered as dana. If it helps your work, donate here →

This is one of 3,622 free prompts in the penguin tree ai library. Sign up once and the whole library opens, with new prompts by email.

open the library →

Client escalation email generator

Generates a ready-to-send client escalation email with a calibrated tone tier, an accountable issue summary, concrete next steps, named owners, and a specific timeline commitment — plus an internal handling note and a…

download .md

Purpose

Produce a professional, empathetic escalation email that acknowledges the client's problem, takes appropriate ownership, and commits to specific next steps and timelines — calibrated to the severity of the situation and the health of the relationship — so a CSM or support lead can respond quickly without under-apologizing, over-promising, or accidentally admitting liability.

When to Use

Invoke this skill when the user asks to:

  • Respond to an escalated support ticket or an executive escalation
  • Write to an angry, frustrated, or churn-risk enterprise client
  • Communicate an SLA breach, missed deadline, or broken commitment
  • Send an outage or incident notification to affected accounts
  • Address a repeated or unresolved issue the client has raised before
  • De-escalate a situation before a scheduled call or renewal conversation

Inputs Required

Before starting, collect:

  • What happened: The issue, incident, or complaint in the user's own words
  • Severity: Critical (outage/data/revenue impact) / High (blocking, SLA-breached) / Medium (degraded, workaround exists) / Low (annoyance, cosmetic)
  • Client relationship: Contract value tier, tenure, renewal proximity, and current sentiment (calm / frustrated / angry / considering churn)
  • Recipient: Name/role of who receives it (end user, admin, VP, C-suite) and whether it's a first contact or a follow-up
  • What's true right now: Root cause status (known / investigating), whether it's fixed, and any workaround available
  • Commitments the sender can actually make: Realistic timelines, owners, and next actions — only what's genuinely achievable
  • Sender identity: Name, title, and whether they're the direct owner or escalating on behalf of the team

If severity or sentiment is missing, default to High severity and frustrated sentiment, and note the assumption. Never invent a timeline or root cause the sender did not provide.

Instructions

Step 1: Select the Tone Tier

Match the emotional register to severity + sentiment. Do not over- or under-calibrate.

  • Tier 1 — Reassuring (Low/Medium severity, calm/frustrated client): Warm, confident, matter-of-fact. Acknowledge, then focus on the fix. Minimal apology.
  • Tier 2 — Accountable (High severity, frustrated client): Direct ownership, clear apology, emphasis on concrete corrective action. No defensiveness.
  • Tier 3 — Contrite & Executive (Critical severity, angry or churn-risk client): Sincere accountability without groveling, business-impact awareness, senior-level commitment, and an offer to speak live. Every sentence signals control and priority.

Step 2: Write the Subject Line

Produce 2 subject line options:

  • One clear and status-forward (e.g., "Update on [issue]: resolution in progress")
  • One relationship-forward for higher tiers Rules: no exclamation points, no vague "Following up," no blame language. Signal ownership and specificity in under 60 characters.

Step 3: Build the Email Body in Five Blocks

  1. Acknowledgment (1–2 sentences): Name the issue precisely and validate the impact on the client. Lead with empathy, not logistics. Never open with "Per your ticket."
  2. Ownership statement (1–2 sentences): Take the appropriate level of responsibility for the tone tier. Apologize where warranted — cleanly, once, without hedging ("I want to apologize for..." not "We're sorry if...").
  3. Issue summary (2–4 sentences): Restate what happened in plain language, including root cause status (known vs. under investigation) and current state (fixed / workaround / in progress). State only confirmed facts — flag anything unconfirmed as "we're verifying."
  4. Next steps + timeline (bulleted): 2–4 concrete actions, each with an owner and a specific time commitment. Use realistic windows ("by end of day Thursday, EST") over vague ones ("shortly"). Include the workaround here if one exists.
  5. Close + availability (1–2 sentences): Offer a live conversation for Tier 2/3, restate the next proactive update ("I'll send an update by [time] regardless of progress"), and sign off in the sender's voice.

Step 4: Apply Liability & Language Guardrails

  • Do not admit fault beyond what the sender confirmed; describe impact, not legal blame.
  • Never promise a fix time the sender didn't authorize; if none is confirmed, commit to a next-update time instead.
  • Replace passive corporate hedging ("efforts are underway") with a named human owner.
  • Strip filler ("just wanted to reach out," "at this time," "please be advised").

Step 5: Attach Internal Handling Note

Below the email, add a short internal-only note (not for the client) covering: recommended follow-up cadence, who else should be looped in, one risk to watch, and whether this account warrants a save/retention play.

Output Format

  • Structure: Two subject line options → full email body (ready to copy-paste) → separated internal handling note
  • Length: Email 150–280 words; internal note 60–100 words
  • Format: Email as formatted prose with a bulleted next-steps block; internal note as 4 labeled bullets
  • Sections: Tone tier used (one line) → Subject options → Email body → Internal handling note
  • Tone: Professional, human, and controlled — empathetic without groveling, specific without legalese; sounds like a competent senior operator, not a support macro

Quality Checklist

Before delivering output, verify:

  • [ ] The tone tier matches the stated severity and sentiment (and is labeled)
  • [ ] Every next step has a named owner and a specific time commitment — no "shortly" or "ASAP"
  • [ ] No timeline or root cause appears that the sender did not provide
  • [ ] Apology level fits the tier — no over-apologizing on Low severity, no under-apologizing on Critical
  • [ ] No liability-exposing language or blame; impact is described, fault is not overclaimed
  • [ ] A next proactive update time is committed to, even if the issue isn't yet resolved
  • [ ] Word counts are within range and the email is copy-paste ready with no placeholders

Notes

  • The single biggest failure in escalation emails is a vague timeline — "shortly" reads as evasion to an angry client. Always convert it to a date/time or a next-update commitment.
  • For Tier 3 (churn risk), the offer to speak live is the message; the email exists to earn the call.
  • When the root cause is unknown, silence is worse than "we don't know yet, and here's when I'll update you." Commit to a rhythm, not a fix.
  • If the client has escalated the same issue before, acknowledge that history explicitly in the acknowledgment block — pretending it's the first time destroys trust.

How to use this skill

This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.

Installing in Claude Code:

  • ~/.claude/skills/<folder-name>/SKILL.md — auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)
  • <your-project>/.claude/skills/<folder-name>/SKILL.md — auto-loads only when working in that project (recommended for project-specific skills)

Restart Claude Code or open a new session to register.

Using in Claude.ai, ChatGPT, Gemini, or other AI tools:

  1. Drag this file into a new chat, or copy and paste its contents
  2. Send your task or request as your next message — the AI will apply the skill where it fits

For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills