Skip to main content

Message recognition

How to use a smart trigger in automations

David P. avatar
Written by David P.
Updated over a week ago

The Message recognition trigger is a smart trigger that uses a neural network to analyze user messages or comments under posts or during live streams. It enables automation to be triggered not just by specific words but by the context of the message.

To use this feature, you’ll need the Unlimited plan or the Pro Plan with the additional ChatPlace AI package. This package enables access to all neural network features in your project.

More about plans and pricing: ChatPlace plans.

How it works:

  1. Context Configuration. Instead of specifying exact words or phrases, you set a topic that the automation should respond to. For example, you can configure automation for topics like "the client is interested in the prices" or "the client wants to learn about the company’s products".

  2. The neural network recognizes the context. The trigger analyzes the user's text and initiates the automation if the message matches a specified topic or context.

  3. The automation is triggered by context, not just words. You don’t need to list all possible phrases, because the neural network analyzes the message and activates the appropriate automation on its own.

How to set up:

  1. In the automation builder, click the "Set start triggers" button in the top right corner.

  2. Select "Message recognition".

  3. Define the key topic or context that the automation should respond to (e.g., "interested in pricing" or "wants to learn about products").

  4. The system will trigger automation every time a customer sends a message with the specified context.

Important! When using the smart trigger, the limitations that apply to triggers from post comments or live stream messages still remain. Therefore, the automation workflow should be configured with all nuances in mind.​

More about the limitations of comment-based triggers in a separate article.

This smart trigger makes it much easier to set up automations, allowing them to be triggered based on context, rather than on words accidentally appearing in a conversation.

Did this answer your question?