Customer Support Prompt Examples

Generates professional, empathetic support responses that address the customer's issue, provide clear steps, and maintain brand voice. Adapts tone based on customer sentiment.

You are a customer support agent for [COMPANY]. Draft a response to this customer ticket.

Brand voice: [friendly and professional / casual and helpful / formal and precise]

Customer context:
- Name: [NAME]
- Plan: [PLAN LEVEL]
- Account age: [DURATION]
- Previous tickets: [NUMBER]

Customer message:
"""
[PASTE CUSTOMER MESSAGE]
"""

Guidelines:
1. Acknowledge their frustration if they're upset — don't jump straight to the solution
2. Provide a clear, step-by-step solution
3. If you need more information, ask specific questions (not vague "can you clarify?")
4. Include relevant help article links where applicable
5. End with a clear next step
6. Keep the response under 200 words unless the issue is complex
7. Use the customer's name naturally (once, not repeatedly)

If the issue cannot be resolved via email, explain what will happen next (call, screen share, escalation).

Classifies incoming support tickets by category, urgency, and required expertise, enabling automated routing and SLA assignment.

Classify this support ticket. Return a JSON object with the following fields:

{
  "category": "billing | technical | account | feature_request | bug_report | general",
  "subcategory": "specific subcategory",
  "urgency": "critical | high | medium | low",
  "sentiment": "positive | neutral | frustrated | angry",
  "expertise_needed": "tier1 | tier2 | tier3 | engineering",
  "estimated_complexity": "simple | moderate | complex",
  "key_issue": "one sentence summary",
  "suggested_response_type": "template | custom | escalation"
}

Classification rules:
- Critical urgency: service is completely down, data loss, security issue
- High: major feature broken, blocking their work
- Medium: feature not working as expected, workaround exists
- Low: question, minor issue, feature request

Ticket:
From: [FROM]
Subject: [SUBJECT]
Body: [BODY]

Generates a concise summary for escalated tickets, including the customer's issue, steps already taken, customer sentiment, and recommended next action for the senior agent.

Summarise this support conversation for escalation to a senior agent.

Conversation thread:
"""
[PASTE FULL CONVERSATION]
"""

Provide the escalation summary in this format:

**Customer**: [Name], [Plan], [Account age]
**Issue**: [1-2 sentence description]
**Impact**: [How this affects the customer's business]

**Timeline**:
- When the issue started
- When they first contacted us
- Number of interactions so far

**Steps Already Taken**:
1. [What has been tried]
2. [Result of each attempt]

**Customer Sentiment**: [calm / frustrated / angry / threatening to churn]
**Churn Risk**: [low / medium / high] — explain why

**Recommended Next Action**: [What the senior agent should do first]
**Relevant Documentation**: [Any KB articles or known issues that might apply]

Analyses a batch of customer feedback (reviews, surveys, support tickets) to identify themes, sentiment trends, and actionable insights.

Analyse the following batch of customer feedback and provide insights.

Feedback entries:
"""
[PASTE 10-20 FEEDBACK ENTRIES, SEPARATED BY ---]
"""

Provide:

1. **Theme Analysis**: Identify the top 5 recurring themes. For each:
   - Theme name
   - Frequency (how many entries mention it)
   - Overall sentiment for this theme
   - Representative quote
   - Suggested action

2. **Sentiment Distribution**:
   - Positive: X%
   - Neutral: X%
   - Negative: X%
   - Trend vs previous batch (if provided)

3. **Urgent Issues**: Any feedback that requires immediate attention (safety, legal, major bugs)

4. **Product Improvement Opportunities**: Top 3 product changes that would address the most feedback

5. **Positive Highlights**: What customers love — use these for marketing and team morale

Present findings in a format suitable for a weekly product team review.

Creates customer-facing knowledge base articles from internal support notes, including clear step-by-step instructions, troubleshooting tips, and related articles.

Create a customer-facing knowledge base article based on these internal support notes.

Topic: [TOPIC]
Internal notes:
"""
[PASTE INTERNAL NOTES, RESOLUTION STEPS, OR SUPPORT TICKET THREAD]
"""

Article requirements:
1. **Title**: Clear, searchable (how a customer would phrase the question)
2. **Summary**: 1-2 sentences describing what this article covers
3. **Applies to**: Which plans/products/versions this applies to
4. **Steps**: Clear, numbered instructions with expected outcomes for each step
5. **Troubleshooting**: "If this doesn't work..." section covering common variations
6. **Screenshots needed**: [Describe where screenshots would be helpful — we'll add them manually]
7. **Related articles**: Suggest 2-3 related topics

Writing rules:
- Use simple, non-technical language (explain jargon if unavoidable)
- Write at a Year 8 reading level
- Use active voice and direct instructions ("Click Settings" not "Settings should be clicked")
- Include the exact names of buttons, menus, and options as they appear in the product

Drafts apology communications for service incidents, adapting the tone and content based on incident severity, customer impact, and relationship value.

Draft an apology communication for the following service incident.

Incident details:
- What happened: [DESCRIPTION]
- Duration: [HOW LONG]
- Customers affected: [NUMBER/SEGMENT]
- Business impact: [WHAT CUSTOMERS COULDN'T DO]
- Root cause: [IF KNOWN]
- Resolution: [WHAT WE DID TO FIX IT]
- Prevention: [WHAT WE'RE DOING TO PREVENT RECURRENCE]

Communication type: [EMAIL TO AFFECTED CUSTOMERS / STATUS PAGE UPDATE / SOCIAL MEDIA POST]
Severity: [MINOR / MODERATE / MAJOR / CRITICAL]

The communication should:
1. Acknowledge the specific disruption (not vague "issues")
2. Take clear responsibility (no passive voice, no blaming)
3. Explain what happened in plain language
4. State what we've done to resolve it
5. Explain what we're doing to prevent it happening again
6. Offer compensation if appropriate for the severity
7. Provide a direct contact for follow-up questions

Tone: sincere, transparent, and action-oriented. Do not minimise the impact or use corporate jargon.