Chatbot Script Template: 5 Ready-to-Use Conversation Flows

What is a chatbot script?

A chatbot script is the pre-written dialogue and decision logic that controls how a bot converses with users. It maps every greeting, question, response branch, and fallback message into a structured flow. Without a script, your chatbot improvises. And improvising bots lose leads, frustrate customers, and create more support tickets than they resolve.

A chatbot script is a structured conversation map that defines what a bot says, what user responses it anticipates, what logic it follows to branch between paths, and where it hands off to a human agent.

The global AI chatbot market reached $15.6 billion in 2024 and is projected to hit $46.6 billion by 2029 (Tidio, 2026). Chatbot adoption across businesses grew roughly 4.7x between 2020 and 2025. Yet 51% of consumers say they prefer bots over humans only when they want immediate service, and 84% insist human interaction should remain an option (Zoho SalesIQ, 2026). This tells you something important: the script quality determines whether your bot feels helpful or infuriating. Businesses that deploy well-scripted chatbots see an average 340% first-year ROI, a 92% customer satisfaction rate, and 30-40% reduction in customer service costs (Jotform, 2026). The gap between those results and the bots people complain about on social media? Almost always the script.

What does this chatbot script template look like?

Each of the five scripts follows the same structural framework. Here’s the core layout you’ll work with for every use case:

Script Component	What It Contains	Example
Greeting	Opening message, tone-setting, purpose statement	“Hi! I can help you find the right plan. Want to answer 3 quick questions?”
Intent Detection	Initial branching question or button menu	Buttons: “Get a quote” / “Talk to support” / “Book a demo”
Conversation Flow	Main dialogue with question-response pairs	Bot asks company size, user selects range, bot adjusts recommendation
Decision Tree	Conditional logic branching based on user input	If budget > $5K, route to enterprise flow; else route to self-serve
Fallback Messages	Responses for unrecognized inputs	“I didn’t catch that. Would you like to: [Option A] [Option B] [Talk to a person]”
Handoff Trigger	Conditions that escalate to a human agent	After 2 failed fallbacks, or when user types “agent” or “help”
Closing	End-of-conversation message and next steps	“Thanks! A team member will email you within 2 hours. Anything else?”

Each script also includes a conversation flow diagram description in plain language, so you can map it in any chatbot builder (Tidio, Intercom, Drift, HubSpot, Voiceflow, or Botpress).

What does this template cover?

The template includes five complete chatbot scripts, each covering a different business use case. Here’s what you get:

• Lead Qualification Bot – 4-question qualification flow, scoring logic, routing rules for sales vs. nurture
• Customer Support Bot – Issue categorization, troubleshooting trees for 6 common issue types, escalation paths
• Booking and Scheduling Bot – Calendar integration prompts, time zone handling, confirmation and reminder messages
• Product Recommendation Bot – Needs assessment questions, recommendation logic for 3 product tiers, upsell triggers
• FAQ Bot – Category-based navigation, 15 pre-written FAQ responses, search fallback to knowledge base
• Fallback Message Library – 12 fallback message variations organized by context (unclear input, off-topic, repeated failure)
• Tone and Voice Guide – How to adjust scripts for formal, casual, or brand-specific voice
• Metrics Tracking Sheet – KPIs to measure per script: completion rate, fallback rate, handoff rate, CSAT

How do you customize these scripts for your business?

Start with the script closest to your primary use case. Don’t try to build all five at once. Here’s the process: Step 1: Pick your highest-impact use case. If you’re drowning in repetitive support tickets, start with the customer support bot. If your sales team is wasting time on unqualified leads, start with lead qualification. Data from Desk365 (2026) shows that 30% of service cases are now resolved by AI, so support is often the fastest win. Step 2: Replace placeholder text with your brand voice. Every script uses neutral, professional language. Swap in your company name, product names, pricing tiers, and any industry-specific terminology. Keep messages under 60 words per bubble. Users skim chatbot messages faster than emails. Step 3: Map your decision tree. The template gives you the logic structure. You need to fill in the specific conditions. For lead qualification, that means defining what “qualified” means for your business. For product recommendations, it means mapping your actual product catalog to the recommendation triggers. Step 4: Build in your chatbot platform. Copy the dialogue into your builder’s flow editor. Platforms like Voiceflow, Tidio, and Botpress all support drag-and-drop flow creation from script documents. Test every branch. Step 5: Test with 10 real users before full launch. Watch where they get stuck, where they drop off, and where the fallback messages fire. Fix those paths first. Then launch.

What does a lead qualification chatbot script look like?

A lead qualification bot replaces the “fill out this form” experience with a conversation that feels more natural and captures better data. The goal is to determine fit in 4 questions or fewer, then route the lead to the right team member or nurture sequence. B2B companies using chatbots for lead qualification report 60% of them already use chatbot software (Tidio, 2026), and well-built flows convert 2-3x better than static forms. Greeting: “Hi there! I’m [Bot Name]. I can connect you with the right person on our team. Mind if I ask a few quick questions? It’ll take about 30 seconds.” [Button: “Sure, let’s go”] [Button: “I’d rather talk to someone”] Question 1 – Company Size: “How many people are on your team?” [Button: “1-10”] [Button: “11-50”] [Button: “51-200”] [Button: “200+”] Question 2 – Primary Need: “What’s the main thing you’re looking to solve?” [Button: “Generate more leads”] [Button: “Improve our SEO”] [Button: “Run paid ads”] [Button: “Something else”] Question 3 – Timeline: “When are you looking to get started?” [Button: “This month”] [Button: “Next quarter”] [Button: “Just researching”] Question 4 – Budget (conditional, only if team > 10): “Do you have a monthly marketing budget in mind?” [Button: “Under $2K”] [Button: “$2K-$10K”] [Button: “$10K+”] [Button: “Not sure yet”] Routing Logic:

Condition	Action
Team 51+, budget $10K+, timeline this month	Route to senior account exec, book meeting link
Team 11-50, budget $2K-$10K, any timeline	Route to account manager, send case study
Team 1-10, any budget, any timeline	Add to email nurture, offer free resource
“Just researching” selected	Offer downloadable guide, add to newsletter
“I’d rather talk to someone” at greeting	Immediate handoff to live agent or show phone number

Closing (qualified lead): “Great, you’re a perfect fit for our [service]. I’ve booked you a 15-minute call with [Name]. Check your email for the invite. Anything else I can help with?”

What does a customer support chatbot script look like?

A support bot’s job is triage and resolution. It categorizes the issue, attempts a self-service resolution, and escalates to a human when it can’t solve the problem. AI chatbot interactions cost roughly $0.50 each, while human support tickets range from $6 to $40 depending on complexity (Hyperleap AI, 2026). The cost savings come from resolving the easy 60-70% of tickets automatically. Greeting: “Hey! I’m here to help. What can I assist you with today?” [Button: “Order issue”] [Button: “Account/billing”] [Button: “Product question”] [Button: “Technical problem”] [Button: “Something else”] Branch: Order Issue “I can help with that. What’s going on with your order?” [Button: “Where’s my order?”] [Button: “Wrong item received”] [Button: “Want to return/exchange”] [Button: “Cancel my order”] Sub-branch: Where’s my order? “Let me look that up. Can you share your order number? You’ll find it in your confirmation email.” [Text input field] Bot: [Pulls tracking info via API] “Your order #[number] shipped on [date] via [carrier]. Here’s your tracking link: [link]. Expected delivery: [date]. Does that help?” [Button: “Yes, thanks!”] [Button: “No, I need more help”] Escalation Trigger: If user selects “No, I need more help” twice, or if the order is marked as delivered but user reports non-receipt: “I’m connecting you with a support specialist who can investigate this further. Expected wait time: [X minutes].” Key design principle: Never dead-end the user. Every response should offer at least two forward paths. Chatbot flowchart best practices from Rasa (2026) confirm that users who hit dead ends abandon at 3x the rate of users who always see a next step.

What does a booking and scheduling chatbot script look like?

A booking bot eliminates the back-and-forth of scheduling by guiding users through date, time, and service selection in under 90 seconds. This script works for service businesses (salons, consultants, clinics, agencies) and integrates with calendar tools like Calendly, Cal.com, or Google Calendar. Greeting: “Hi! I can help you book an appointment. What type of appointment are you looking for?” [Button: “Consultation call”] [Button: “Service appointment”] [Button: “Follow-up visit”] Service Selection (if “Service appointment”): “Which service are you interested in?” [Button: “[Service A] – 30 min”] [Button: “[Service B] – 60 min”] [Button: “[Service C] – 90 min”] [Button: “Not sure, help me choose”] Date/Time Selection: “Here are the next available slots for [selected service]:” [Calendar widget or button list showing 3-5 available slots] “Don’t see a time that works? [Button: “Show more times”] [Button: “Check next week”]” Contact Info Collection: “Almost done! I just need a few details:” “Your name?” [Text input] “Best email for the confirmation?” [Email input] “Phone number (for reminders)?” [Phone input, marked optional] Confirmation: “You’re all set! Here’s your booking: [Service] on [Date] at [Time]. A confirmation email is on its way to [email]. We’ll send a reminder 24 hours before. Need to change anything? [Button: “Reschedule”] [Button: “All good!”]” Time Zone Handling: Always detect or ask for the user’s time zone before showing slots. Display times in the user’s local zone. A bot that shows 2:00 PM without specifying EST or PST will cause no-shows.

What does a product recommendation chatbot script look like?

A recommendation bot acts as a digital shopping assistant. It asks 3-4 questions about the user’s needs, matches their answers against your product catalog, and presents 1-3 options with reasoning. Well-built recommendation bots increase average order value by 10-30% because they guide users to the right product instead of letting them browse and bounce. Greeting: “Hi! Want help finding the right [product category]? I’ll ask 3 quick questions and give you a personalized recommendation.” [Button: “Yes, help me choose”] [Button: “I already know what I want”] Question 1 – Use Case: “What will you mainly use this for?” [Button options customized per product category – e.g., for software: “Personal use” / “Small team” / “Enterprise”] Question 2 – Priority: “What matters most to you?” [Button: “Price”] [Button: “Features”] [Button: “Ease of use”] [Button: “Scalability”] Question 3 – Budget: “What’s your budget range?” [Button: “$0-$50/mo”] [Button: “$50-$200/mo”] [Button: “$200+/mo”] [Button: “Flexible”] Recommendation Output: “Based on your answers, I’d recommend: [Product Name] – [1-sentence reason]. [Price]. [Link to product page] This matches your need for [use case] with a focus on [priority], and it fits your budget.” Follow-up: [Button: “Tell me more”] [Button: “Show me alternatives”] [Button: “Add to cart”] [Button: “Talk to a specialist”] Upsell Trigger: If user selects the mid-tier product and budget is “Flexible,” show a comparison with the next tier up: “For $[X] more per month, you’d also get [feature]. Worth considering if [condition]. Want to compare side by side?”

What does an FAQ chatbot script look like?

An FAQ bot is the simplest chatbot to build and often the first one teams deploy. It replaces your static FAQ page with a conversational interface that guides users to answers based on category, not scrolling. Over 67% of consumers worldwide have engaged with a chatbot for customer support in the past year (Zoho SalesIQ, 2026), and FAQ bots handle the repetitive questions that consume agent time. Greeting: “Hi! I can answer common questions about [your company/product]. What would you like to know about?” [Button: “Pricing”] [Button: “Features”] [Button: “Account & billing”] [Button: “Shipping & returns”] [Button: “Getting started”] Category: Pricing “Here are the most common pricing questions:” [Button: “What are your plans?”] [Button: “Is there a free trial?”] [Button: “Do you offer discounts?”] Answer: “What are your plans?” “We have 3 plans: [Starter] at $[X]/mo, [Professional] at $[X]/mo, and [Enterprise] with custom pricing. You can compare all features on our pricing page. Want details on a specific plan?” [Button: “Starter details”] [Button: “Professional details”] [Button: “Enterprise details”] [Button: “Back to topics”] Search Fallback: If user types a free-text question instead of using buttons: “Let me search for that…” [Bot searches knowledge base] “Here’s what I found: [Top 3 matching articles]. Was any of these helpful?” [Button: “Yes, thanks”] [Button: “No, I need something else”] Knowledge Base Miss: “I couldn’t find a matching answer. Let me connect you with someone who can help.” [Button: “Chat with support”] [Button: “Email us”] [Button: “Try a different question”]

How do you write effective fallback messages?

Fallback messages fire when the bot doesn’t understand the user’s input. They’re the most overlooked part of any chatbot script, and they’re responsible for the majority of bad chatbot experiences. A good fallback acknowledges the confusion, offers alternatives, and always keeps a path to a human open. 12 Fallback Message Templates: First failure (gentle redirect):

• “I didn’t quite catch that. Could you try rephrasing, or pick from these options?”
• “Hmm, I’m not sure I understood. Did you mean: [Option A] or [Option B]?”
• “I want to help but I’m not sure what you need. Here’s what I can do: [list]”

Second failure (offer structured options):

• “I’m still having trouble understanding. Let me give you some options: [buttons]”
• “Sorry about that. Try picking one of these instead: [buttons]”
• “Let me start fresh. What area can I help with? [category buttons]”

Third failure (escalate to human):

• “I think a real person would be better here. Let me connect you with our team.”
• “I appreciate your patience. I’m going to get you a human agent now.”
• “This one’s beyond me. Transferring you to our support team now. One moment.”

Off-topic inputs:

• “Ha, I’m just a bot with a narrow skill set. I can help with [X, Y, Z]. Pick one?”
• “That’s outside my expertise. I’m best at [topic]. Want help with that?”
• “Interesting question, but I’m only trained on [topic]. For anything else, here’s how to reach our team: [contact link]”

Why most chatbot scripts fail

We’ve reviewed chatbot implementations for clients across ecommerce, SaaS, and professional services at ScaleGrowth.Digital. The scripts that fail share three patterns. First, they try to sound human instead of being useful. Nobody is fooled by “I’m so glad you asked!” responses. Users want answers, not performative enthusiasm. Second, they don’t test the unhappy paths. Every script works when the user clicks the right buttons. The real test is what happens when they don’t. Third, they treat deployment as the finish line. The best chatbot scripts are rewritten 3-4 times in the first month based on actual conversation logs.

“The biggest mistake we see is teams spending weeks perfecting the happy path and zero time on fallbacks. In production, 20-30% of chatbot interactions hit a fallback within the first three messages. If those fallback messages are lazy ‘I don’t understand’ dead ends, you’ve just built a frustration machine. Script your fallbacks with the same care you script your main flow.” Hardik Shah, Founder of ScaleGrowth.Digital

Common mistakes to avoid:

• Too many questions upfront. Keep qualification to 3-4 questions max. Every additional question drops completion rates by 5-10%.
• No escape hatch. Every screen should have a “Talk to a person” option. 84% of consumers insist human interaction should remain available (Zoho SalesIQ, 2026).
• Walls of text. Keep each bot message under 60 words. Break longer explanations across 2-3 messages with a short pause between them.
• Ignoring mobile. Over 60% of chatbot interactions happen on mobile devices. Test your flows on a phone screen before launch.
• No analytics. Track completion rate, fallback trigger rate, handoff rate, and CSAT per script. Without data, you’re guessing.