/
DE

⚡ Promptolis Original · Sales & Revenue

🛡️ Customer Service De-Escalation Scripts (Refunds, Reviews, Complaints)

Voss tactical empathy + LAST principle + NVC. Acknowledge BEFORE solving. 80% of rage de-escalates with 'I hear that this is frustrating.' Scripts for refunds, deadline misses, scope-creep, public reviews, late-payment chasing.

⏱️ 4 min 🤖 20-30 min full script + decision matrix 🗓️ Updated 2026-05-11
⚡ Quick Answer

Customer Service De-Escalation Scripts (Refunds, Reviews, Complaints) — Voss tactical empathy + LAST principle + NVC. Acknowledge BEFORE solving. 80% of rage de-escalates with 'I hear that this is frustrating.' Scripts for refunds, deadline misses, scope-creep, public reviews, late-payment chasing. Setup: 4 min · Best AI: Claude Sonnet 4.6 (fast iteration). Opus for high-stakes B2B disputes. · Cost: Free, MIT-licensed.

Why this is epic

Acknowledge BEFORE solving — Voss-derived. Most customer-service rage is about feeling unheard.

Refund decision matrix: full vs partial vs no, with criteria. Helps user decide if undecided.

Public review response template — for FUTURE customers reading, not the angry one.

📑 Page navigation + Key Takeaways Click to expand

📌 Key Takeaways

  • What it is: Voss tactical empathy + LAST principle + NVC. Acknowledge BEFORE solving. 80% of rage de-escalates with 'I hear that this is frustrating.' Scripts for refunds, deadline misses, scope-creep, public reviews, late-payment chasing.
  • Best for: Etsy / Shopify / Amazon refund disputes
  • Time investment: 4 min setup, 20-30 min full script + decision matrix output
  • Recommended AI model: Claude Sonnet 4.6 (fast iteration). Opus for high-stakes B2B disputes.
  • Cost: Free forever — MIT-licensed, no signup, no paywall

📑 On this page

  1. The prompt (copy-ready)
  2. How to use it (4 steps)
  3. Example input + output
  4. Common use cases
  5. Pro tips + variants
  6. FAQ

⚙️ At a glance

Category:
Sales & Revenue
Setup time:
4 min
Output time:
20-30 min full script + decision matrix
Best AI model:
Claude Sonnet 4.6 (fast iteration). Opus for high-stakes B2B disputes.
License:
MIT (free commercial use)
Last reviewed:
📊 Promptolis Original vs generic AI prompts Click to expand
Feature Promptolis Generic prompts
Structure: XML + chain-of-thought Role-play one-liner
Example output: Real full example Rare
Variants: 3-7 per prompt Single
Output quality: +30-50% accurate [Anthropic] Baseline

On the other hand, generic prompts work fine for simple lookups. Promptolis Originals shine for nuanced reasoning where precision matters.

The prompt

Promptolis Original · Copy-ready
<role> You are a Customer Service De-Escalation Script writer for solopreneurs, Etsy/Shopify sellers, freelancers, and small B2B teams handling angry customers, refund disputes, missed-deadline conflicts, and complaint escalations. You use Chris Voss tactical empathy framework (mirror + label + accusations audit), the LAST principle from hospitality (Listen, Acknowledge, Solve, Thank), and Marshall Rosenberg's Nonviolent Communication for separating observation from judgment. You know that 80% of customer-service rage is about feeling unheard, NOT the actual issue. Acknowledging the feeling first usually de-escalates within 2-3 messages. You produce specific scripts for: refund requests, deadline misses, product defects, scope-creep clients, contract disputes, late-payment chasing, public negative review responses. </role> <principles> 1. Acknowledge BEFORE solving. Most rage de-escalates with 'I hear that this is frustrating' BEFORE any solution discussion. 2. Mirror: repeat their last 3-5 words as a question. Voss's tactic. Buys time + signals understanding. 3. Label the emotion: 'It seems like you're feeling [X]' — names what they're feeling without diagnosing. 4. Accusations audit: pre-empt their worst objections. 'You're probably thinking I'm trying to avoid responsibility' — defuses before they say it. 5. Refund decision matrix: when YES (preserve relationship + brand), when PARTIAL (split-difference), when NO (with empathy + clear reasoning). 6. Public complaints (reviews, social) are read by future customers. Response is for THEM, not the angry person. Stay professional even when justified to defend. 7. Late-payment chasing: escalating tone over messages 1-2-3. Don't start hostile. 8. Scope-creep: clear language about what's in / out + what new scope costs. Not apologetic, not aggressive. 9. Avoid: defensive 'as our policy states,' over-apologizing, capitulating without context, threatening lawyers in early messages. 10. End with: specific scripts + decision matrix + escalation paths. </principles> <input> <situation-type>{refund-request / deadline-miss / product-defect / scope-creep / contract-dispute / late-payment / public-review / complaint-escalation}</situation-type> <context>{specifically what happened — your version of events}</context> <customer-position>{what they're saying, demanding, threatening}</customer-position> <your-position>{what you can / will / won't do — honestly stated}</your-position> <channel>{email / chat / public review / phone / social DM}</channel> <relationship>{first interaction / repeat customer / long-term client / B2B / B2C}</relationship> <key-question>{specific — should I refund, how to phrase, how to handle the rage, etc.}</key-question> </input> <output> ## Diagnose: What This Customer Actually Needs [Beyond the stated demand — what's the underlying complaint? Heard? Compensated? Boundary clarified?] ## Acknowledgment Script (FIRST message) [Specific words. Mirror + label. Don't solve yet.] ## Solution Script (SECOND message after acknowledgment) [The actual offer / position / refund decision. With reasoning.] ## If They Push Back (THIRD message) [Escalation path. When to hold firm, when to flex, when to offer alternative.] ## Decision Matrix [Refund yes/partial/no. With criteria. Help user decide if not already decided.] ## What NOT to Say (Common Mistakes) ['As our policy states,' over-apology, defensive 'we always,' threatening lawyers prematurely.] ## If This Becomes Public (Reviews, Social) [Response template. Remember: response is for FUTURE customers, not the angry one.] ## Escalation Path (when to disengage) [When to stop responding. When to engage lawyer/platform-support. When to ban customer.] ## On Your Key Question [Direct address.] ## What This Cannot Predict [How customer responds, whether they leave a review anyway, whether full resolution is possible.] </output>
🚀 Open in ChatGPT ✨ Open in Claude 💎 Open in Gemini

Example: input → output

Here's how this prompt actually performs. Real input below, real output from Claude Opus 4.

📝 Input
<situation-type>Refund request that's borderline + customer is escalating tone</situation-type>
<context>I sell handmade ceramic mugs on Etsy. A customer ordered 4 mugs, received them. Two arrived chipped (clear shipping damage, photos confirm). She's now demanding refund for ALL 4 mugs PLUS shipping both ways PLUS a 'compensation for stress.' She's sent 3 messages over 2 days, increasingly hostile, last message threatened to 'leave a 1-star review and tell everyone.'</context>
<customer-position>Wants: full refund for 4 mugs ($120), refund for original shipping ($12), her cost to ship back ($14), plus 'compensation for stress' ($50) — total $196 demanded. Original sale was $132 ($120 + $12 shipping).</customer-position>
<your-position>Reasonable: refund for the 2 broken mugs + return shipping for the 2 broken mugs. The other 2 mugs are fine and she'd been planning to keep them. Total fair refund: ~$67 (2 mugs $60 + return ship $7). I will NOT pay 'stress compensation.' I'm willing to lose $20-30 to make this go away clean.</your-position>
<channel>Etsy messaging system</channel>
<relationship>First-time customer. No prior interaction.</relationship>
<key-question>How do I de-escalate without giving in to the full demand? She's threatening a bad review which would hurt my shop. But $196 on a $132 sale is unreasonable.
🤖 Output

Diagnose: What This Customer Actually Needs

Underlying complaint: she received broken items + her time/energy was wasted (real). The 'stress compensation' demand is performative-rage, not genuine claim — she's escalating because she feels unheard + powerful (review threat).

Three things she actually needs:

1. To feel heard. Acknowledgment of the genuine problem (2 broken mugs).

2. Fair compensation for actual harm. 2 mugs broken = legitimate refund. Time + return shipping = legitimate.

3. A way out of the spiral. Right now she's escalated and feels she has to threaten to get response. Give her the off-ramp.

What she does NOT actually need: $196 (overcharge by ~$130 from fair). Her demand is the negotiation opener, not her actual floor.

Likely fair settlement landing point: ~$67-80. Most reasonable customers accept this when offered with empathy.

Acknowledgment Script (FIRST message)

Send within 2 hours of receiving her latest message. Don't wait.

'Hi [Name],

First, I'm so sorry. Two of your mugs arrived chipped — that should never have happened, and I completely understand why you're frustrated. You ordered them in good faith and what you got is not what I sent.

I looked at the photos you shared. The damage looks like shipping impact (the boxes get stacked at the carrier), but that's not your problem to solve. Your problem is that your mugs are broken and you're out money.

Let me make this right today. Can you give me until end of day to send you a clear resolution?

[Your name]'

Notice what this script does:

  • Acknowledges the FACT (mugs broken, frustrating)
  • Names the EMOTION (frustrated)
  • Validates her position (your problem to solve)
  • Doesn't solve YET — buys time for the second message
  • Doesn't argue about the 4-vs-2 issue yet

Does NOT do:

  • 'As per Etsy policy' — defensive
  • 'I always pack carefully' — defensive
  • Address the 'stress compensation' demand — premature
  • Apologize 5 times — over-apology weakens

Solution Script (SECOND message, end of same day)

'Hi [Name],

Thank you for your patience. Here's what I'm offering, and I want to be transparent about my reasoning:

For the 2 broken mugs: Full refund of $60 (item cost), plus I'll cover return shipping ($7) so you can send them back to me at no cost. That's $67 going back to you today.

For the 2 mugs that arrived intact: Those are yours to keep. They're not damaged, and I'm assuming you'd planned to use them.

On the broader request: I want to be honest with you — I'm a one-person shop. I can absorb the cost of replacing damaged items because that's part of doing business. I can't compensate beyond that, because there's not margin in handmade ceramics for stress fees, and being honest with customers about what I can do builds the trust I need to keep my shop running.

If the 2 intact mugs aren't actually what you wanted now (sometimes after a bad experience the rest doesn't feel right), I can also accept those as a return + refund the additional $60. Just let me know.

For return shipping: I'll send you a prepaid label as soon as you confirm.

I know this isn't the full $196 you asked for. I hope my reasoning makes sense — and I genuinely am sorry the experience went this way.

[Your name]'

Key moves in this script:

  • Specific number ($67) breaks down the 'I'm not getting full refund' fog
  • Offers alternative (return all 4) — gives her agency
  • Names YOUR constraints (one-person shop, no margin for stress fees) — humanizes you
  • Explicit reasoning for why you can't go higher
  • Stays warm, doesn't capitulate to performative rage

If She Pushes Back (THIRD message)

If she accepts the $67: Send the prepaid label, send the refund, send a brief 'thank you for understanding, I hope your remaining mugs serve you well' note. Move on.

If she insists on $196 + threatens review:

'Hi [Name],

I hear that you're not satisfied with what I've offered. I understand the threat of a bad review is significant for me, and I want to be honest with you — I'd rather have a 1-star review than build my business on capitulating to demands beyond fair compensation. That said, I'd love to find a resolution that works.

My maximum I can offer:

  • Full refund of $60 for the 2 broken mugs
  • Return shipping covered ($7)
  • Original shipping refunded ($12, partial — half of the order was damaged)
  • A $20 credit toward a future order if you ever want to give us another shot

Total: $99. That's where my limit is. If that doesn't work for you, I'll process the $79 refund for the broken mugs + return shipping today and consider this resolved on my end.

I'm sorry we couldn't land on something that worked for both of us.

[Your name]'

This third-message strategy:

  • Names the review threat directly (defangs it)
  • States that a bad review is acceptable cost
  • Offers ONE more concession ($99 vs $67) but firmly
  • Sets the boundary: $99 max, otherwise $79 + close
  • Doesn't capitulate to performative rage
  • Saves face for both parties

If she keeps escalating (4th-5th message demanding more): Stop responding. Process the $79 refund (2 mugs + return shipping). Send: 'I've processed your refund for the broken mugs and return shipping. The full amount is $79 and you'll receive it within 3-5 business days. I wish you well.' Then disengage.

Decision Matrix

Refund YES (full, no fight):

  • Item arrived broken + customer is reasonable
  • High-margin item (you can absorb)
  • Long-term customer (relationship value > short-term cost)
  • Public-platform with auto-refund-or-review-impact dynamics (Etsy, Amazon)

Refund PARTIAL (split-difference):

  • Genuine issue + customer over-asking
  • Borderline case (your error or theirs?)
  • Want to keep relationship intact
  • Most common sweet spot

Refund NO (with empathy + reasoning):

  • Customer is clearly trying to get free product (used + claiming defective)
  • Demand is structurally unreasonable ('stress fee', '10x compensation')
  • Customer pattern of past abuse (chargebacks, multiple disputes)
  • Public-review threat alone is not sufficient reason to capitulate

For your specific case: PARTIAL refund ($67-99) is correct. The 2 broken mugs are real harm; the 'stress compensation' is performative.

What NOT to Say

Avoid these phrases — they escalate, not de-escalate:

  • 'As per our shop policy...' — defensive, reads as legalistic
  • 'I always pack my mugs perfectly' — defensive, calls her version into question
  • 'Etsy will side with me on this' — threat, legalistic
  • 'I'm so sorry, I'm so sorry, I'm so sorry' — over-apology weakens position
  • 'You agreed to terms when you bought' — argumentative
  • 'I'll have to consult my lawyer' — premature lawyer-threat escalates
  • 'I think you're being unreasonable' — true but counterproductive
  • 'Other customers have been happy with my mugs' — irrelevant to her experience

If This Becomes Public (Etsy 1-Star Review)

IF she leaves the review anyway, public response template:

'I'm sorry your order didn't go well. Two of your four mugs arrived damaged in shipping, and I refunded $79 for those mugs + return shipping to make it right. I understand you'd hoped for additional compensation, but I'm a one-person shop and offered what I could honestly afford. I hope your remaining mugs serve you well. — [Your name]'

This response is for FUTURE customers reading reviews, not for her. Future customers will read:

  • Acknowledgment of the genuine issue
  • Specific number ($79 refunded)
  • Reasonable + transparent reasoning
  • Polite tone despite obvious provocation

Most future customers will side with you reading this.

Escalation Path

When to STOP responding:

  • After 4-5 messages where she's not engaging your offers
  • If she becomes abusive (slurs, personal attacks, threats beyond reviews)
  • If she begins making increasingly outlandish demands ($300, $500)

When to involve Etsy support:

  • If she opens a case (you can present your offers as record of attempting resolution)
  • If review violates Etsy policy (false claims, abuse) — request removal

When to consider banning customer:

  • After this transaction, you can choose not to sell to her again. Etsy allows blocking customers.

On Your Key Question

Direct:

1. Don't give in to $196. That's 49% over the original sale price for 2 broken items. Capitulating teaches both this customer and your future patterns: rage works. Don't reinforce.

2. $67-99 is the right landing zone. Specifically: open at $67, escalate to $99 if needed, close at $79 + return shipping if she rejects $99.

3. The review threat is real but not catastrophic. One 1-star review on a shop with otherwise-good reviews barely impacts conversion. New customers read reviews holistically. A measured public response (template above) shows future customers your professionalism.

4. Move within 24 hours. Slow response makes rage worse. Fast acknowledgment + measured solution within same day prevents the escalation spiral.

5. Document everything in Etsy's messaging. If she opens a case or files a chargeback, your record shows reasonable resolution attempts.

6. Lose-$20-to-make-it-go-away is fine. $99 vs $67 is the cost of preventing one bad review + one Etsy case. Worth it.

What This Cannot Predict

Whether she accepts any offer. Whether she leaves the review regardless. Whether she opens a case anyway. None of these are predictable.

What IS reliable: your acknowledgment + reasonable offer + measured public response = correct positioning regardless of her response.

📋 How to use this prompt (4 steps · under 60 seconds) Click to expand
  1. 1 Copy the prompt above. Click "Copy prompt". XML-structured prompt now on clipboard.
  2. 2 Open ChatGPT, Claude, or Gemini. One-click launch above. Recommended: Claude Sonnet 4.6 (fast iteration). Opus for high-stakes B2B disputes..
  3. 3 Paste + fill placeholders. Replace {curly braces} with your context. Specificity = quality.
  4. 4 Run + iterate. Setup: 4 min. Output: 20-30 min full script + decision matrix.

Common use cases

  • Etsy / Shopify / Amazon refund disputes
  • Service-business deadline-miss conflicts
  • Scope-creep client conversations
  • Late-payment chasing (3-message escalation)
  • Public negative review responses
  • B2B contract disputes
  • Solopreneur abusive-customer disengagement

Best AI model for this

Claude Sonnet 4.6 (fast iteration). Opus for high-stakes B2B disputes.

Pro tips

  • Acknowledge BEFORE solving
  • Mirror: repeat their last 3-5 words as question
  • Label the emotion: 'It seems like you're feeling X'
  • Specific number breaks the rage-fog
  • Public response is for FUTURE customers, not angry one
  • Don't capitulate to performative rage
  • Walk-away point pre-defined

Customization tips

  • For B2B contract disputes (vs B2C complaints): different stakes, different scripts. Specificity about contract terms, scope-of-work, change-orders matters more than emotional acknowledgment.
  • For platform-disputes (Amazon FBA, eBay, Etsy): platform's policies + auto-refund mechanisms shape the negotiation. Different from direct-business disputes.
  • For service-business late-payment chasing: 3-message escalation cadence — gentle reminder, formal request, final notice. Then collections agency if needed.
  • For public negative reviews on Yelp/Google Reviews: lower-stakes than Etsy reviews. Brief professional response template; don't engage in argument publicly.
  • For abusive customers (slurs, threats, personal attacks): different protocol. Stop responding. Document. Block. Refer to platform support or law enforcement if threats credible.
  • For refund-fraud patterns (multiple chargebacks, claiming items not received when they were): different protocol. Decline + document + report to platform.
  • For cultural/language differences in customer-service expectations: adjust tone. Some markets expect more apology + warmth; others expect direct + transactional.
  • Premium pack content: industry-specific script libraries (Etsy/Amazon/Shopify/freelance/B2B/services), public-review response template library, scope-creep prevention contract clauses, late-payment escalation cadence templates.

Variants

Refund Dispute (Etsy/Shopify/Amazon)

B2C platform-context

Deadline Miss (Service Business)

Service-recovery + repair

Scope-Creep Client

Specifying in/out + change-order

Late-Payment Chasing

3-message escalation cadence

Public Negative Review

Response for future customers

B2B Contract Dispute

Different stakes + legal-aware

Abusive Customer Disengagement

When to stop responding + ban

Frequently asked questions

Common questions about this prompt and how to get the best results from it.

How do I use the Customer Service De-Escalation Scripts (Refunds, Reviews, Complaints) prompt?

Open the prompt page, click 'Copy prompt', paste it into ChatGPT, Claude, or Gemini, and replace the placeholders in curly braces with your real input. The prompt is also launchable directly in each model with one click.

Which AI model works best with Customer Service De-Escalation Scripts (Refunds, Reviews, Complaints)?

Claude Sonnet 4.6 (fast iteration). Opus for high-stakes B2B disputes.

Can I customize the Customer Service De-Escalation Scripts (Refunds, Reviews, Complaints) prompt for my use case?

Yes — every Promptolis Original is designed to be customized. Key levers: Acknowledge BEFORE solving; Mirror: repeat their last 3-5 words as question

What does it cost to use this prompt?

The prompt itself is free, MIT-licensed, with no email signup required. You only pay for your AI model subscription (ChatGPT Plus $20/mo, Claude Pro $20/mo, Gemini Advanced $20/mo) — and even those have free tiers that work with most Promptolis Originals.

How is this different from PromptBase or PromptHero?

PromptBase sells prompts in a marketplace ($2-15 each). PromptHero focuses on image-generation prompts. Promptolis Originals are free, MIT-licensed text/reasoning prompts hand-crafted with full example outputs, multiple variants, and a recommended best AI model per prompt. We don't sell anything.

Explore more Originals

Hand-crafted 2026-grade prompts that actually change how you work.

← All Promptolis Originals
P

Curated by · Last reviewed

Editorial process + credentials ▼

Credentials: Independent prompt-engineering team since 2026. Sister projects: SeoScore.tools and 9bench.com. Meet the team →

Editorial process: Each prompt is built from primary sources (research papers, established frameworks, professional methodologies), structured with XML tags + chain-of-thought scaffolding for 2026-grade LLMs, tested across multiple models before publishing.