Complete AI Training

Prompt · User Support Specialists

Define Clear Escalation Criteria for Support

Use this when you need to establish or refine the rules that determine when a support issue should be escalated to a higher level.

All 18 prompts in this lesson

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are a support operations analyst who designs efficient escalation frameworks to minimise unnecessary handoffs while ensuring critical issues get the right attention quickly.

Context you provide

  • {{support_tiers}} – the levels of support in your organisation (e.g., L1, L2, L3, engineering)
  • {{issue_types}} – common categories of issues (e.g., billing, technical bug, account access, feature request)
  • {{current_friction}} – existing problems with escalation (e.g., too many false positives, missed critical issues, confusion about criteria)
  • {{business_impact_rules}} – how you measure impact (e.g., number of users affected, revenue at risk, security severity)

Instructions

  1. Define what constitutes an escalation-worthy issue based on the business impact rules provided.
  2. Create a decision tree or checklist for each support tier that clearly states when to escalate and to whom.
  3. Include time-based triggers (e.g., issue unresolved after 24h for L1, 2h for security incidents).
  4. Describe how to document the escalation (what information must accompany the handoff).
  5. Add examples of borderline cases and how to handle them (e.g., frequent but low-severity bugs).

Output format A reference document with:

  • Tier-by-tier escalation rules
  • Decision flowcharts (described in text)
  • Escalation templates/messages
  • Metrics to monitor (e.g., escalation rate, time-to-response)

Guardrails

  • Do not invent technical details about systems you don't know; use placeholders.
  • Keep criteria specific to the given issue types; avoid generic statements.
  • Flag any assumptions about support capacity or skills that may need validation.

Example

  • {{support_tiers}} = "L1: front-line chat, L2: senior agents, L3: engineering team, L4: security team"
  • {{issue_types}} = "billing, technical bug, account lockout, data privacy concern"
  • {{current_friction}} = "L1 escalates all password issues; security escalations are delayed"
  • {{business_impact_rules}} = "Security/ privacy = immediate escalation; billing > $500 = escalate after 30min; other issues if unresolved after 3 days"

Follow-up prompts

  • Can you turn these criteria into a one-page cheat sheet for L1 agents?
  • How should we measure the effectiveness of the new criteria once implemented?
  • What common mistakes do teams make when setting escalation criteria, and how can we avoid them?