Flows are your visual canvas for crafting dynamic customer interactions. Unlike simple triggers, Flows enable you to guide visitors through structured conversations, collect specific information, perform actions, and ensure they reach the right outcome, whether it’s self-service via Aura AI or a seamless handoff to a human agent.
A Flow is a sequence of modular blocks that define a conversational path. You design these paths using an intuitive drag-and-drop builder, allowing for complex logic and personalized visitor experiences.
Flows are typically initiated by a Start Condition (similar to a Trigger) and then progress through various Action Blocks based on visitor input or predefined logic.
Creating a Flow involves two main components: defining when the Flow should begin and assembling the series of actions and questions it will execute.
A Flow can be initiated by various visitor actions or environmental factors. These conditions determine when your automated journey begins.
Visitor Behavior & Presence
Scheduled & Time-Based
Visitor Identification & Attributes
Device & Browser Specificity
Agent Availability
Start conditions for Flows are very similar to those for Triggers. You can think of a Flow as a more advanced, multi-step “action” that a simple Trigger might initiate.
Once a Flow starts, you connect various modular blocks to create the desired user journey. These blocks represent different actions or interactions.
Send Message
This block allows the Flow to send a message to the visitor.
Options: Text, rich reply (with buttons), image, or quick reply options.
Ask a Question (via bubbles)
This block sends a message and then pauses the Flow, waiting for a user reply to proceed. Useful for gathering specific input.
Send Form
Present a predefined form to the visitor. This can be used to set or update a profile field based on the collected conversation data.
Assign Aura AI
Directs the conversation to Aura AI, allowing the AI agent to handle the next set of user queries based on its knowledge.
Send Data Update
Push user-provided data or other information collected in the Flow to an external system via an API call (e.g., update a CRM record). Requires a configured Data Update endpoint.
Branch Logic
Create conditional branches in your Flow using AND/OR
operators. This allows the Flow to take different paths based on visitor responses, profile data, or other criteria.
Tag Chat
Automatically apply specific labels or tags to the chat based on user behavior or the outcome of the Flow. Useful for segmentation and reporting.
Escalate to Agent
Routes the conversation to a human agent when certain conditions are met (e.g., if the user asks for human help, or if the Flow can’t resolve the issue).
End Flow
Cleanly stops the Flow once a desired outcome is met or the conversation concludes.
Flows empower you to automate complex interactions that go beyond simple trigger-response mechanisms.
Build a multi-step conversation to ask qualifying questions (e.g., company size, pain points), collect contact info via forms, and then assign qualified leads to sales.
Guide users through a series of questions to identify their issue, offer relevant KBA articles, and escalate to the correct department or agent if self-service isn’t sufficient.
After a support interaction or a successful purchase, trigger a Flow to ask for feedback, collect ratings, and identify areas for improvement.
Combined with Data Requests/Updates) Allow users to book meetings or appointments directly within the chat by collecting necessary details and updating an external calendar system.
Create interactive tours or mini-onboarding sequences. Ask users about their goals, provide quick tips, and direct them to relevant features or documentation.
We’ll add flow testing within the same page soon!
Follow these guidelines to create effective and user-friendly Flows:
Map Out the Journey First
Before building, sketch out the user’s journey, decision points, and desired outcomes. This helps avoid getting lost in the visual builder.
Keep it Concise & Conversational
Design messages that are easy to read and natural. Avoid overly long messages or too many questions at once.
Provide Clear Options
When using “Ask a Question” or rich replies, give users clear, easy-to-understand options to guide them.
Design for Fallback & Escalation
Always include Escalate to Agent
blocks at points where the Flow might not be able to resolve an issue, ensuring a smooth handoff.
Utilize Branch Logic Smartly
Use Branch Logic
to personalize paths based on visitor input or known profile data, making the Flow more relevant.
Test Extensively with Virtual Visitor
Simulate every possible path and user input. Test edge cases, incorrect inputs, and the escalation process to catch errors early.
Iterate & Optimize (with Analytics)
Once Flow performance analytics are available, use them to identify bottlenecks, optimize conversation paths, and improve conversion rates.