Customer Support & Ops

Verified against ChatGPT · 2026-08-14

Write an escalation note that gives a manager everything needed to decide in one read

Produces an escalation note for a manager or specialized team that states the ask up front, what authority is needed and why, and what happens if no one acts in time — instead of a narrative the manager has to interpret.

ChatGPT (GPT-5.1)5 fillable variables

The prompt

Ready to copy — highlighted parts are example details you can swap.

You are writing an escalation note to hand a case up to a manager or specialist team. The note needs to let them make a decision on one read, without needing to ask clarifying questions first.

CASE SUMMARY
Enterprise customer is requesting a refund of $4,200 for six months of an add-on they say they never knowingly enabled.

WHY THIS NEEDS ESCALATION (not just difficulty, but why it's above the front-line agent's authority)
Refunds above $500 require manager approval per policy; this is also a potential billing-consent issue that may need a look at whether the add-on's opt-in flow was clear.

WHAT DECISION OR ACTION IS BEING REQUESTED
Approve or deny the $4,200 refund, and flag to product whether the add-on opt-in flow needs review.

TIME SENSITIVITY
Customer's contract renewal decision is due in 5 business days and they've said this will factor into whether they renew.

WHAT HAPPENS IF NOBODY ACTS
Customer likely doesn't renew a $30k/year contract; this is their largest recorded frustration point in three years as a customer.

HOW TO STRUCTURE THIS
Open with the specific decision or authority being requested, not a narrative lead-in — a manager reading an escalation queue needs to know in the first sentence whether this is a refund-approval ask, a policy-exception ask, or something else, so they can route it or act on it immediately. State exactly why this is above the front-line agent's own authority — a genuine policy limit, a dollar threshold, a legal-sounding claim — not just "this seemed complicated," since escalations that don't state a real authority gap train managers to distrust the escalation queue as a whole. Give the time sensitivity as a concrete deadline or trigger, not a vague "soon," and state the actual consequence if no one acts by then, so the manager can weigh it against their other priorities honestly. Include only the case detail that's relevant to the specific decision being requested — this is not a full case history, it's a decision brief.

WHAT NOT TO DO
Do not escalate a case that a front-line agent actually has authority to resolve just because it's emotionally difficult — reserve escalation notes for genuine authority or expertise gaps, and say so if you determine this case doesn't actually need escalation despite being requested. Do not bury the ask in a paragraph of context before stating it.

OUTPUT FORMAT
1. The ask, one sentence, first line.
2. Why this needs escalation specifically (authority gap, not just difficulty).
3. Relevant case detail (only what's needed for this decision).
4. Deadline/trigger and consequence of inaction.
5. If, on review, this doesn't actually require escalation, say so explicitly instead of writing the note as requested.

Customize

Optional — swap in your own details for the highlighted parts above.

Why this works

GPT-5.1 defaults to narrative structure for case write-ups — background, then complication, then request — because that's the more common shape for explaining a situation in general prose, but a manager triaging an escalation queue is optimizing for speed of decision, not narrative completeness, and a request buried after two paragraphs of context costs real time multiplied across every escalation in the queue that day. Requiring the specific authority gap rather than accepting "this seemed complicated" as sufficient justification addresses a structural weakness in how these models handle escalation reasoning: without a hard constraint, the model will happily generate a plausible-sounding reason for escalating almost anything, because from a pure language-generation standpoint, justifying an escalation is easy regardless of whether a genuine authority gap exists — forcing a concrete dollar threshold, policy limit, or legal-sounding element as the stated reason filters out the emotionally-difficult-but-technically-resolvable cases that shouldn't be clogging a manager's queue in the first place. Demanding a concrete deadline and stated consequence rather than accepting "soon" or "this is important" matters because vague urgency is unfalsifiable and every escalation note tends to claim it, which means a manager has no real signal to prioritize between competing asks — a stated trigger ("renewal decision in 5 business days") and a stated consequence ("likely non-renewal of a $30k contract") give a manager an actual basis for triage math instead of trusting a claimed urgency they can't verify. Instructing the model to say explicitly when a case doesn't actually warrant escalation, even though it was asked to write the note, is the single most valuable constraint here — it's the difference between a tool that rubber-stamps whatever request it's given and one that pushes back when the underlying judgment call doesn't hold up, which is exactly the check a front-line agent drafting their own escalation note is unlikely to apply to themselves.

What you get back

Ask: Approve or deny a $4,200 refund for an Enterprise customer, and flag a possible billing-consent issue to product. Why this needs escalation: Refund exceeds the $500 front-line approval threshold; also involves a potential consent/opt-in issue on the add-on itself, which is a product policy question, not just a refund amount question. Relevant detail: Customer states they never knowingly enabled the add-on; six months of charges at $700/month. No prior refund history on this account in three years. Deadline: Customer's renewal decision is due in 5 business days; they've stated this outcome will factor into renewal. Consequence of inaction: Likely non-renewal of a $30k/year contract — their largest stated frustration point as a customer to date.

Verified against

ChatGPT GPT-5.1 · 2026-08-14

Changelog

  • 2026-08-14 Initial publish, verified against ChatGPT GPT-5.1.

Need this built into your business?

If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.

EXPLORE WHAT SCULT BUILDS
All Customer Support & Ops prompts

Check your AI visibility

One URL in, a 0–100 score and the exact fixes out.

RUN THE CHECK

Browse all the tools

15 tools across six categories
13 of them never send your data anywhere

Free · No signup · No trial clock

SEE THE DIRECTORY