In this article we explain what context transfer is for and how you can apply it in your automated conversations.
Context transfer is a key functionality in Atom that allows Smartons (AI) and other components to continue the conversation using information the user has already provided, avoiding annoying repetitions and dramatically improving the conversational experience.
📌 What is context transfer?
This functionality allows a Smarton or a Response Evaluator to receive the customer's original message (or information obtained in previous steps of the flow) and continue the conversation without losing the previous thread.
This happens transparently to the user, without them having to repeat what they already said.
✨ What does this functionality do in practice? It automatically transfers the customer's message to the connected Smarton.
🚗 Simple example:
The customer writes: "I'm interested in the Mazda CX50".
The message matches an intent that redirects to a catalog Smarton.
The Smarton receives that original message and responds: "Perfect, here's all the information about the Mazda CX50. Would you like to schedule a test drive?"
👉 The user doesn't perceive the component change and doesn't need to repeat their query. The AI already knows which car they want.
🧭 How is it configured?
Context transfer is active by default for all supported platforms in the Flowbuilder input (Inbound): WhatsApp, Instagram, Facebook Messenger, and Web Plugin.
Where do I see it? You can view this configuration from the Bot Start component, where you'll see an indicator or toggle showing that the functionality is active.
How do I turn it off? If for any strategy you want to disable it, simply turn off the corresponding toggle.
🔍 How does it work between components?
To master this function, you must first understand how components are classified in Flowbuilder:
📥 Input (Entry): Require a message or user action to execute (Smarton, Evaluate response, Intents, Location, Save field if the data is entered by the user).
✅ Action: Execute automatically in the background (Stage, Label, Conditionals, HubSpot, Save field with pre-assigned value).
⏳ Wait: Only add time to the flow (Wait time).
🔄 Scenarios where context IS transferred
Input ➡️ Smarton
Smarton ➡️ Action ➡️ Smarton: (Example: Generic Smarton ➡️ Stage ➡️ Field saving Smarton). In this case, the context travels seamlessly through the background action.
🛑 Scenario where context is NOT transferred
⚠️ The only action component that does NOT transfer context is the Message component.
Smarton ➡️ Message ➡️ Smarton: 👉 The flow stops at the Message component. The user will need to send a new text (input) for the second Smarton to wake up and activate.
⚠️ Golden Exception: The user's first message
When it comes to the very first message a user sends to open the chat, and context transfer is active, that message will always travel until it finds the first input component, skipping initial messages.
First message ➡️ Stage ➡️ Message ➡️ Evaluate Response: The initial message correctly reaches the evaluator.
Note: If it were not the first message, or if the transfer were turned off, the flow would remain paused at the Message component waiting for the customer to speak again.
🔕 When should you turn off context transfer?
There are scenarios where it's not recommended that the AI or flow remember the previous message:
When you want to force a new question and prevent the Smarton from interpreting past messages.
When the flow requires that the user absolutely responds to a specific message or selects a button before continuing.
In flows where an informative message should intentionally stop the conversation so the user reads it.
💡 In these cases, turning off context transfer gives you strict step-by-step control of bot behavior.
❓ Frequently Asked Questions (FAQ)
Why does my bot stop responding after the user's first message? This is the most common error and usually occurs for two reasons related to this topic:
Context transfer is disabled: If the user sends a message and the flow passes through a Message component, the conversation stops there and doesn't continue to the Smarton until the user writes again.
Poor flow design: Make sure your flow has correctly defined input components and their outputs.
👉 Quick review: Verify that the toggle on the "Bot Start" node is turned on.
Make your bots seem human! 🚀
Context transfer is the secret to preventing your customers from repeating information and getting frustrated. Master how data flows between your components and create flawless conversational experiences. ✅