Escalation Router

Step 2: Generate Escalation Brief

Invoke cc-internal-brief-writer with escalation context. Every brief follows this structure:

Headline (one line): What is happening, to which customer, and what is at stake. Example: "API latency issue affecting Acme Corp -- P1 open 12 days, EUR 85k ARR, renewal in 67 days"

Customer impact (2-3 sentences): How this affects the customer's business. Not just "they are unhappy" but "their analytics team cannot run reports, costing them an estimated 4 hours per week in manual workarounds."

Timeline (chronological):

• When the issue started
• Each escalation step taken, with date and outcome
• Current status

Resolution attempts (what has been tried):

• Every prior attempt to resolve, with the outcome of each
• This prevents the receiving team from re-trying approaches that have already failed

Specific ask (what you need):

• Not "please help" but "we need engineering to investigate the API latency on this account's instance and provide a fix or timeline by [date]"
• Include the definition of "resolved" from the customer's perspective

Customer communication status:

• What the customer has been told
• What they expect (specific timeline, specific outcome)
• Their current sentiment (patient, frustrated, threatening churn)

Commercial context (when relevant):

• ARR, renewal date, expansion pipeline
• Financial impact of losing this account

Step 3: Route and Confirm Receipt

1. Send the brief to the identified team via the configured channel (Slack for urgent, email for standard, ticketing system for trackable)
2. Log the escalation with: ID, timestamp, category, severity, routing target, brief content
3. Start the acknowledgement clock (4 hours for critical, 24 hours for standard)
4. If no acknowledgement within the SLA: automatically re-route with escalation to the team's manager and a note that the initial routing was not acknowledged

Step 4: Track Resolution

Monitor the escalation lifecycle:

Status update cadence:

• Critical escalations: daily update to CSM, automatic stall detection at 48 hours
• Standard escalations: update every 3 business days, stall detection at 5 business days

Step 5: Generate Escalation Report

Weekly summary for CS leadership:

{
  "report_date": "2026-03-10",
  "active_escalations": 7,
  "new_this_week": 3,
  "resolved_this_week": 2,
  "stalled": 1,
  "average_resolution_time_days": 4.2,
  "escalations": [
    {
      "id": "ESC-2026-0342",
      "account": "Acme Corp",
      "category": "product_defect",
      "severity": "critical",
      "status": "in_progress",
      "days_open": 5,
      "owner": "Mike Ross (Eng Lead)",
      "arr_at_risk": 85000,
      "customer_sentiment": "frustrated_but_patient",
      "next_update_due": "2026-03-11"
    }
  ],
  "trends": {
    "most_common_category": "product_defect",
    "avg_time_to_acknowledge": "3.1 hours",
    "stall_rate": "14%"
  }
}

Handoff to Human

Confidence and Limitations

• High confidence for routing logic and brief generation -- structured classification and documentation with defined rules
• High confidence for status tracking and stall detection -- timestamp-based monitoring with defined thresholds
• Medium confidence for severity classification -- the skill classifies based on defined criteria (ticket severity, ARR, renewal proximity), but the CSM may assess severity differently based on relationship context or cumulative frustration
• Low confidence for resolution timeline estimation -- how quickly an internal team resolves an issue depends on their capacity, competing priorities, and technical complexity. The SLAs are targets, not guarantees
• Cannot guarantee that the receiving team will act within the expected timeframe
• Cannot assess whether escalation is the right approach vs. a direct conversation, a workaround, or accepting the limitation
• Cannot navigate internal politics -- if the receiving team is resistant or under-resourced, the CSM must advocate through relationship and influence, not routing mechanics

Dependencies

Required:

• CRM API (escalation tracking, account context)
• Internal messaging (Slack or email for routing)
• cc-internal-brief-writer (brief generation)

Strongly recommended:

• bi-account-brief (full account context for briefs)
• lo-sla-monitor (for SLA breach-triggered escalations)
• Support platform (for ticket-linked escalations)

Downstream consumers:

• ic-cross-func-prep (escalation committee materials)
• ic-internal-notifier (escalation status notifications)
• cc-report-generator (weekly escalation summary)

References

• references/routing-rules.md -- Complete routing table, acknowledgement SLAs, and re-escalation protocols