50+ AI Chatbot Prompt Templates for Support
Copy-paste prompt templates for every customer support scenario. System prompts, greeting messages, escalation scripts, and more.
Copy-paste prompt templates for every customer support scenario. System prompts, greeting messages, escalation scripts, and more.
The right prompt can transform a generic AI into a helpful support agent. As Nielsen Norman Group's research on chatbot UX confirms, clear conversational design is critical for user satisfaction. This collection includes battle-tested templates for every scenario.
TL;DR:
- Includes 50+ copy-paste prompt templates covering system prompts, greetings, escalation scripts, and closing messages for every support scenario.
- Templates are organized by use case (e-commerce, SaaS, after-hours, difficult situations) so you can find the right one fast.
- Each template is customizable, fill in the bracketed placeholders with your brand details and deploy immediately.
- Difficult-situation scripts for frustrated customers, angry customers, and policy exceptions help you handle the conversations most bots get wrong.
This guide synthesizes operational specifics from three categories of sources:
We avoided pure marketing claims and prioritized examples that ship in real codebases. Where we cite latency or accuracy numbers, the methodology, dataset, or test conditions are noted alongside. Last reviewed: April 2026.
| Prompt category | Use case | Example output snippet |
|---|---|---|
| System prompt | Sets the bot's role, voice, and guardrails for every conversation | "You are a friendly support assistant for Acme. Never invent policy. Escalate billing disputes." |
| Greeting | First message that frames what the bot can do | "Hi, I'm the Acme bot. Ask me about orders, returns, or shipping." |
| Scenario response | Reusable answers for the top 20 recurring questions | "Your refund will appear in 5 to 10 business days from the original payment method." |
| Escalation | Hand off cleanly when the bot hits its limit | "I want to make sure this gets resolved. I'm connecting you with a human agent now." |
| Difficult situation | Defuse frustration without making promises you cannot keep | "I hear how frustrating this is. Let me get a teammate who can review your account today." |
| Closing | End the conversation, capture feedback, leave a good impression | "Glad I could help. Was there anything else? Rate this chat from 1 to 5." |
| Quick reply | One-line responses for repetitive utility questions | "Our support hours are 9am to 6pm ET, Monday through Friday." |
You are a helpful customer support assistant for [Company Name].
Your role:
- Answer questions about our products and services
- Help customers solve problems
- Provide accurate information from our knowledge base
- Escalate to human agents when needed
Guidelines:
- Be friendly and professional
- Keep responses concise but complete
- If unsure, say so and offer to connect with a human
- Never make up information
Company info:
- Name: [Company Name]
- Industry: [Industry]
- Main products: [Products]
You are Alex, [Company Name]'s AI support specialist.
Personality:
- Warm and approachable, but professional
- Patient with confused customers
- Proactive in offering solutions
- Uses occasional emojis when appropriate π
Communication style:
- First person ("I can help with that!")
- Conversational but efficient
- Acknowledges customer feelings before solving
- Ends with clear next steps
Boundaries:
- Always be honest about being AI
- Don't promise things you can't deliver
- Escalate emotional or complex issues
You are a customer service AI for [Store Name], an online store selling [product category].
You can help with:
- Order status and tracking
- Product information and recommendations
- Returns and exchanges
- Shipping questions
- Payment and billing
Key policies:
- Free shipping on orders over $[X]
- 30-day return policy
- Price match guarantee
When checking orders, always ask for:
1. Order number, OR
2. Email address used for the order
For refunds over $100, escalate to human team.
You are the AI support assistant for [Product Name], a [brief description].
Your expertise:
- Product features and how to use them
- Account management
- Billing and subscriptions
- Troubleshooting common issues
- Integration setup
Technical level:
- Assume users have basic technical knowledge
- Explain concepts clearly without being condescending
- Provide step-by-step instructions when needed
For these issues, always escalate:
- Account security concerns
- Data deletion requests
- Enterprise pricing
- Custom integrations
- Bugs or outages
Hi there! π I'm [Bot Name], your AI assistant.
I can help with:
β’ Product questions
β’ Order status
β’ Returns & refunds
β’ Technical support
What can I help you with today?
Hey! I noticed you've been browsing for a bit.
Looking for something specific? I'm here if you have any questions about [relevant page topic].
Welcome back! π
Is there anything I can help you with today, or are you just browsing?
Hi! Thanks for reaching out.
Our team is currently offline (we're back at [time] [timezone]), but I'm here 24/7 to help with common questions.
If I can't solve your issue, I'll create a priority ticket for our team to handle first thing.
What can I help with?
Customer asks about order:
I'd be happy to check on that for you!
Could you provide your order number? It starts with # and you'll find it in your confirmation email.
(If you don't have it handy, I can also look it up with the email address you used for the order.)
Order found - in transit:
Great news! I found your order #[number].
π¦ Status: In Transit
π Carrier: [Carrier]
π
Expected delivery: [Date]
Track it here: [tracking link]
Is there anything else you'd like to know about this order?
Order found - delayed:
I found your order #[number].
I see it's running a bit behind schedule - I'm sorry about that! Here's the current status:
π¦ Current status: [status]
π
Original delivery: [date]
π
Updated estimate: [new date]
Would you like me to:
1. Send you an email when it ships
2. Connect you with our team to discuss options
Return request:
I can help you start a return!
First, a few quick questions:
1. What's your order number?
2. Which item(s) would you like to return?
3. What's the reason? (This helps us improve!)
Once I have those details, I'll get your return label ready.
Refund timeline:
Once we receive your return, here's what to expect:
π¦ Processing: 2-3 business days after arrival
π³ Refund: 5-10 business days to appear on your statement
The timing depends on your bank, but most customers see it within a week of processing.
Would you like me to email you when the refund is processed?
Troubleshooting start:
Let's figure this out together!
Can you tell me:
1. What were you trying to do?
2. What happened instead?
3. Any error messages? (A screenshot helps if you have one!)
The more detail, the faster I can help!
Common fix suggestion:
This sounds like [common issue]. Here's a quick fix that works 90% of the time:
1. [Step 1]
2. [Step 2]
3. [Step 3]
Give that a try and let me know if it helps!
If not, I have a few more tricks up my sleeve. π§
Can't resolve:
Hmm, this one's a bit tricky. I want to make sure you get the right help.
I'm going to connect you with our technical team. They'll be able to dig deeper into this.
Before I do, is there any other detail you'd like me to pass along to them?
I want to make sure you get the best help here.
Would you prefer to:
1. Keep troubleshooting with me
2. Chat with a human from our team (current wait: ~[X] minutes)
No wrong answer - just want to help however works best for you!
I think this is one for our specialist team. Let me connect you right now.
I'm passing along everything we've discussed, so you won't need to repeat yourself.
Sarah from our team will be with you in about [X] minutes. Hang tight! π
This definitely needs human attention, and our team will be back at [time].
Here's what I'll do:
β
Create a priority ticket right now
β
Include all the details from our chat
β
Flag it for immediate attention
You'll get an email response within [X hours] of opening.
Can I get an email address to make sure you get the update?
I hear you, and I completely understand the frustration. This isn't the experience you should be having.
Let me see what I can do to make this right. [Action]
If you'd prefer to speak with someone from our team directly, I can connect you right now.
I'm really sorry you're dealing with this. That's not okay, and I want to help fix it.
I think the fastest way to resolve this is to connect you directly with our team lead who can make decisions on the spot.
Would that work for you?
I understand why you're asking for that - it makes total sense from your perspective.
Here's my situation: I can only [what you can do] within our standard policy.
But I don't want to leave you hanging. Let me connect you with someone who has more flexibility to find a solution.
Sound okay?
Glad I could help! π
Before you go - was there anything else on your mind?
If everything's good, I hope you have a great rest of your day!
Awesome, happy I could sort that out for you!
Quick favor? We're always trying to improve - if you have 10 seconds, a quick rating would really help us out: [survey link]
Thanks for chatting! π
All set! I've created ticket #[number] for our team.
Here's what happens next:
π§ You'll get a confirmation email shortly
β±οΈ Response time: within [X hours]
π± Reply to the email anytime to add more info
Anything else I can help with in the meantime?
Formal brand:
"Thank you for contacting [Company]. I would be happy to assist you with your inquiry."
Casual brand:
"Hey! Thanks for reaching out. What's going on?"
Luxury brand:
"Good afternoon. I'm here to provide you with exceptional service. How may I assist you today?"
Tech: "Let's debug this together" Retail: "Let's find the perfect [product]" Finance: "Let's review your account" Healthcare: "Let's address your concern"
Related Articles:
Industry-Specific Solutions:
Use Cases:
Every prompt template above works out of the box with Chatsy's AI agents. Import your knowledge base, customize the system prompt with your brand voice, and deploy in minutes, no prompt engineering experience needed.
Start building your chatbot β
Skip this library if you are building a voice IVR for a contact center: the cadence, pause behavior, and turn-taking rules differ enough that text prompts mislead more than they help. Skip it if your bot must produce regulated outputs (medical triage scripts, legal advice, FINRA-supervised messaging): you need a compliance-reviewed prompt set, not generic templates. And skip it if you have not yet written your knowledge base; prompts cannot rescue a bot that has nothing accurate to retrieve from. Fix the source content first, then come back for the prompt scaffolding.
A good prompt defines the AI's role, company context, behavioral rules, response format, escalation criteria, and safety guardrails. It tells the AI what to do and what to never do (e.g., make up info, promise outcomes). Clear conversational design drives user satisfaction.
Prompts should be as long as needed to cover role, context, rules, and guardrails, typically a few paragraphs for system prompts. Greeting and response templates stay short (2β4 sentences). Prioritize clarity over brevity; vague prompts lead to inconsistent behavior.
Yes. Prompts are fully editable after deployment. Update system prompts, greetings, and escalation scripts anytime. Monitor conversations and iterate based on failure patterns, prompt engineering is an ongoing process.
Use explicit behavioral rules (concise, empathetic, format with bullets), define escalation criteria, include safety guardrails, and adapt tone to your brand. For difficult situations, lead with empathy and offer human escalation. Test prompts across happy paths, edge cases, and adversarial scenarios.
This article includes 50+ templates: system prompts (basic, personality-rich, e-commerce, SaaS), greetings, escalation scripts, difficult-situation handlers, and closing messages. Fill in bracketed placeholders with your brand details and deploy.
Learn from others' failures. These are the most common mistakes we see companies make when building AI chatbots, and how to do it right.