Generate a complete conversation flow blueprint for your AI chatbot. Choose your industry and goal — get a ready-to-implement dialog tree.
Tip: This flow uses a branching tree structure. Each → arrow represents a user choice. Customize the placeholders ({{value}}) with your actual data before implementing.
## Chatbot Conversation Flow: E-commerce Support
### Entry Point
Bot: "Hi! 👋 How can I help you today?"
→ [Track My Order]
→ [Return or Exchange]
→ [Product Question]
→ [Something Else]
---
### Branch: Track My Order
Bot: "Sure! What's your order number?"
User: [enters order number]
Bot: "Let me look that up..."
→ IF found: "Your order #{{order}} is {{status}}. Expected delivery: {{date}}."
→ "Need anything else?"
→ [Yes] → restart
→ [No] → "Great! Have a wonderful day! 😊"
→ IF not found: "I couldn't find that order. Can you double-check the number?"
→ After 2 attempts: "Let me connect you with our team for help."
→ [Handoff to human agent]
---
### Branch: Return or Exchange
Bot: "I can help with that! Is this for a return or an exchange?"
→ [Return]
→ "When did you receive the item?"
→ IF within return window: "You're within our return period. Here's how to start:"
→ "1. Go to your order page
2. Click 'Start Return'
3. Print the label
4. Drop off at any {{carrier}} location"
→ IF outside window: "Unfortunately, this order is past our {{days}}-day return window. I can connect you with a team member to discuss options."
→ [Handoff to human agent]
→ [Exchange]
→ "What would you like to exchange it for?"
→ [Collect new item preference]
→ "I'll have our team process this exchange. You'll get an email with next steps within 24 hours."
---
### Branch: Product Question
Bot: "What product are you looking at?"
User: [describes product or pastes URL]
Bot: [Searches knowledge base]
→ IF answer found: "{{answer}}"
→ "Did that answer your question?"
→ [Yes] → "Awesome! Anything else?"
→ [No] → "Let me connect you with a product specialist."
→ [Handoff to human agent]
→ IF no answer: "I don't have details on that specific product. Let me connect you with someone who does."
→ [Handoff to human agent]
---
### Branch: Something Else
Bot: "No problem! Just type your question and I'll do my best to help."
User: [free text]
Bot: [AI responds from knowledge base]
→ [Follow-up or close]Copy this flow as a blueprint for your chatbot setup. Canary handles the AI routing automatically — just provide the knowledge base.
Effective chatbot conversations follow a tree-based dialog design. Every flow starts with an entry point — the initial greeting and intent detection — then branches into specialized paths based on what the user needs. The generator builds these flows using proven patterns from thousands of real support, sales, and lead generation conversations.
Each generated flow includes four key components:
Traditional chatbot flows require manual scripting of every branch. Modern AI chatbots like Canary handle routing automatically — the AI understands natural language and routes conversations based on your knowledge base content, eliminating the need to maintain rigid dialog trees. The flows generated here serve as a planning blueprint and a useful reference for understanding how conversations should progress. For a deeper comparison of scripted flows versus AI-powered routing, see our guide on chatbot vs live chat.
A conversation flow is the structured path a chatbot follows when interacting with a user. It defines the greeting, how the bot identifies intent, what branches exist for different topics, when to ask clarifying questions, and when to hand off to a human agent. Think of it as a dialog tree: each user input leads to a specific bot response, which may branch into further paths. Well-designed flows feel natural while guiding users toward a resolution.
Branching logic routes conversations based on user input. When a user selects an option or types a response, the chatbot evaluates the input and follows the appropriate branch. For example, a support bot might branch into 'billing issue,' 'technical problem,' or 'general question' — each with its own resolution path. Modern AI chatbots like Canary handle branching dynamically using natural language understanding rather than rigid keyword matching.
Handoff triggers should fire when the chatbot detects: (1) high-emotion language indicating frustration or urgency, (2) questions outside the knowledge base where the bot has low confidence, (3) explicit user requests to speak with a human, (4) complex multi-step issues like billing disputes or account recovery, or (5) high-value interactions like enterprise sales inquiries. The best practice is to hand off with full conversation context so the customer never repeats themselves.
Start with 3-5 primary branches based on your most common query categories. Each primary branch can have 2-3 sub-branches for specifics. Going deeper than 3 levels often creates a frustrating experience — if the user hasn't reached a resolution by the third branch, the bot should either provide a direct answer or escalate to a human. Most successful chatbots cover 80% of queries with 5-7 total conversation paths.
Review your flows monthly for the first 3 months after launch, then quarterly. Key signals that a flow needs updating: high drop-off rates at specific nodes, frequent handoffs from a particular branch (indicating the bot can't resolve those queries), new product launches or policy changes, and seasonal shifts in customer questions. AI chatbots that learn from your knowledge base adapt automatically, but the overall flow structure benefits from periodic human review.
Join businesses that have automated support, captured more leads, and cut response times to zero — no code required.
Free forever on Starter. No credit card required.