Customer Escalation

Package a support escalation with full context so the receiving team can act without back-and-forth. Includes customer impact, reproduction steps, urgency classification, and recommended action.

Workflow

Step 1: Gather Context

Collect all relevant information about the issue:

• Ticket details: Read the ticket thread, including customer messages and internal notes.
• Customer profile: Account tier, ARR, contract renewal date, open opportunities.
• Issue history: Previous tickets for the same issue, related bugs, known workarounds.
• Scope assessment: Check whether other customers have reported the same or similar problems.
• Timeline: When the issue was first reported, how long it has been open, SLA status.

Step 2: Assess Impact

Evaluate the business and technical impact:

• Customer impact: Number of affected users, revenue at risk, workflow disruption severity.
• Breadth: Single customer vs. multiple accounts vs. platform-wide.
• Workaround availability: Is there a viable temporary fix or is the customer fully blocked?
• Churn risk: Explicit threats, sentiment trajectory, renewal proximity.
• Contractual obligations: SLA commitments, uptime guarantees, escalation clauses.

Step 3: Classify Urgency

Assign urgency based on a standard framework:

Step 4: Build Reproduction Steps

Document clear, testable reproduction steps:

• Environment details (browser, OS, API version, tenant ID).
• Step-by-step instructions to trigger the issue.
• Expected vs. actual behavior.
• Attach screenshots, logs, HAR files, or error traces when available.
• Note any conditions that make the issue intermittent.

Step 5: Draft the Escalation

Compose the escalation document for the target team.

Output Format

## Escalation: <Issue Title>

**Urgency**: P1 / P2 / P3 / P4
**Target Team**: Engineering / Product / Leadership
**Ticket**: <ID or URL>
**Escalated By**: <agent name>
**Date**: <today>

### Customer Context
- **Account**: <company name> | <tier> | $<ARR>
- **Contact**: <name, role>
- **Renewal**: <date> | **Churn Risk**: Low / Medium / High
- **Open Since**: <date> | **SLA Status**: Within / Breached

### Issue Summary
<2-3 sentence description of the problem and its business impact>

### Reproduction Steps
1. <step>
2. <step>
3. <step>
- **Expected**: <what should happen>
- **Actual**: <what happens instead>
- **Environment**: <browser, OS, version, tenant>

### Impact Assessment
- **Affected customers**: <count or "single account">
- **Revenue at risk**: $<amount>
- **Workaround**: <available / partial / none>

### Recommended Action
<specific ask: fix, investigate, prioritize, executive outreach>

### Supporting Evidence
- <links to logs, screenshots, related tickets>

Examples

User: "Escalate ticket #4521 to engineering -- the customer's data export has been broken for a week" Action: Pull ticket details, assess impact, classify as P2, build reproduction steps from the thread, draft escalation to engineering.

User: "Three customers reported the same SSO login failure today, loop in engineering" Action: Cross-reference all three tickets, identify the common pattern, classify as P1 due to breadth, package a consolidated escalation.

User: "Acme Corp is threatening to cancel, prepare an escalation for leadership" Action: Pull account profile, compile issue history, calculate revenue at risk, draft executive-level escalation with churn risk framing.

Best Practices

• Lead with impact, not symptoms. Engineering and leadership prioritize based on business consequence.
• Include reproduction steps even if the issue seems obvious. Teams in different contexts need them.
• Link to related tickets rather than summarizing them at length to keep the escalation scannable.
• State the specific ask clearly. "Please investigate" is weaker than "Please reproduce and provide an ETA for a fix by Friday."
• If multiple customers are affected, include the count and name the largest accounts.
• Update the escalation when new information arrives rather than sending separate follow-ups.
• Track escalation outcomes to refine urgency calibration over time.