Customer Escalation Playbook

Build a structured escalation playbook for handling angry, high-value, or legally sensitive customer situations with de-escalation scripts and decision trees.

Prompt Template

You are a customer experience director. Create an escalation playbook for my support team:

Company/product: [name and what you offer]
Team size: [number of support agents]
Support channels: [email, live chat, phone, social media]
Current escalation issues: [describe 2-3 recurring difficult situations]
SLA targets: [response time, resolution time]
Escalation tiers: [e.g., L1 agent → L2 senior → L3 manager → VP]

Provide:
1. **Escalation criteria matrix** — what triggers each escalation level (severity × impact)
2. **De-escalation scripts** — word-for-word responses for the 5 most common angry customer scenarios
3. **Decision tree** — step-by-step flow for agents when they receive a hostile message
4. **Legal/compliance red flags** — phrases or situations that require immediate escalation
5. **Compensation authority table** — what each tier can offer without approval
6. **Post-escalation process** — how to document, learn from, and prevent repeat incidents
7. **Self-care protocol** — how agents should decompress after difficult interactions

Example Output

Escalation Playbook — Flowdesk Support Team

Escalation Matrix

| Trigger | Level | Owner | SLA |

|---------|-------|-------|-----|

| Standard complaint | L1 — Agent | Any agent | 4 hours |

| Repeated complaint (3+ tickets) | L2 — Senior | Senior agent | 2 hours |

| Customer threatening to cancel (Pro/Enterprise) | L2 — Senior + CS Manager notified | Senior agent | 1 hour |

| Data loss or security incident | L3 — Manager + Engineering | Support Manager | 30 min |

| Legal threat, regulatory mention, press threat | L3 — Manager → Legal | Support Manager | 15 min |

De-Escalation Script: Angry About Service Outage

**Don't say:** "I understand your frustration, but..."

**Do say:** "You're right to be upset — [specific thing] shouldn't have happened, and I'm sorry it affected your work. Here's exactly what I'm doing about it right now: [action]. I'll update you by [specific time] even if there's no change yet, so you're not left wondering."

**Key technique:** Name the emotion, own the problem, state the action, commit to a timeline.

Decision Tree: Hostile Message Received

1. Pause — don't reply for 60 seconds

2. Is the customer using threats or abusive language toward the agent personally?

→ **Yes:** "I want to help resolve this. I do ask that we keep the conversation respectful so I can focus on fixing this for you." If it continues → escalate to L2

→ **No:** Continue with de-escalation script

3. Is this a known issue with an existing fix?

→ **Yes:** Acknowledge + apply fix + offer compensation per authority table

→ **No:** Acknowledge + create internal ticket + commit to follow-up time

Compensation Authority

| Tier | Can Offer Without Approval |

|------|----------------------------|

| L1 Agent | 1 month free, extended trial |

| L2 Senior | Up to 3 months free, priority support |

| L3 Manager | Full refund, custom resolution |

Tips for Best Results

  • 💡Role-play escalation scenarios monthly with your team — scripts only work if agents have practiced saying them
  • 💡The 60-second pause before replying to hostile messages prevents reactive responses that escalate further
  • 💡Track escalation reasons in your ticketing system — patterns reveal systemic product or process problems

Frequently Asked Questions

What is the Customer Escalation Playbook prompt?

Build a structured escalation playbook for handling angry, high-value, or legally sensitive customer situations with de-escalation scripts and decision trees. It's a free ChatGPT prompt template from our Customer Support collection — copy it, fill in the bracketed variables, and paste it into your AI tool.

Which AI tools work with this prompt?

It's written and tested for ChatGPT, Claude and Gemini. Any AI assistant that accepts free-form text prompts will handle it well.

How do I customize this ChatGPT prompt?

Replace the bracketed variables — such as [name and what you offer], [number of support agents], [email, live chat, phone, social media] — with your own details before running it. Role-play escalation scenarios monthly with your team — scripts only work if agents have practiced saying them

Is this prompt free to use?

Yes. Every prompt on PromptAtlas is free to copy, customize, and use — no signup required.