Intent Flow Steps
Learn about each step type available in Intent Flows and how to configure them.
Get User Input
Collect information from the customer through a question-and-answer interaction.
Configuration:
- Prompt: The question to ask the customer
- Field Name: A name to store the response (e.g.,
order_number,email,issue_description)
How It Works:
- The flow sends your prompt to the customer
- The flow pauses and waits for the customer's reply
- Their response is saved under your field name
- The flow continues to the next step
Example:
- Prompt: "Could you please provide your order number?"
- Field name:
order_number - Customer replies: "ORD-12345"
- The value "ORD-12345" is now available as
{order_number}
Tips:
- Ask one question at a time for clarity
- Use clear, specific prompts
- Choose descriptive field names you'll remember
Update Conversation
Modify conversation properties or add information for your team.
Available Actions:
Add Tags
Apply one or more tags to the conversation. Useful for categorizing and filtering.Set Priority
Change the conversation priority:- Low
- Normal
- High
- Urgent
Higher priority conversations appear at the top of filtered views.
Set Category
Assign the conversation to a ticket category. This can affect agent routing if you have category-based assignment rules.Add Internal Note
Add a note visible only to agents. Use{field_name} to include collected data.Example note: "Customer requesting refund for order {order_number}. Reason given: {reason}"
Escalate to Human
Transfer the conversation to a human agent. Options:- Automatic: Let assignment rules determine the agent
- Specific Agent: Choose a particular team member
When escalated, the bot marks the conversation for human handling.
Send Message
Send a response to the customer.
Configuration:
- Message: The text to send (supports
{field_name}placeholders)
Example:
Thanks for providing your order number! I've found order {order_number}
and our team will review your {request_type} request shortly.Tips:
- Personalize messages with collected data
- Keep messages concise and helpful
- Use this to confirm what the bot understood
Using Placeholders
Throughout your flow, you can reference collected data using curly braces:
| Placeholder | Description |
|---|---|
{order_number} | Value from a "Get User Input" step with field name "order_number" |
{reason} | Value from a "Get User Input" step with field name "reason" |
{email} | Value from a "Get User Input" step with field name "email" |
- Send Message content
- Internal notes
- Any text field in Update Conversation steps
Step Order Best Practices
- Collect information first - Use Get User Input steps early to gather what you need
- Update conversation in the middle - Add tags, notes, and set properties
- Escalate before final message - If escalating, do it before your closing message
- End with confirmation - Use Send Message to confirm what happened
Example Flow Structure
1. Get User Input → "What's your order number?"
- Get User Input → "What issue are you experiencing?"
- Update Conversation → Add tag "Support", Add note with details
- Update Conversation → Escalate to human
- Send Message → "Thanks! An agent will help you shortly."