Add live chat support

We don't offer Live chat functionality ourselves. However you can use external live chat tools and use them in sync with our bots.

Easy Method

The easy way is to use the Links Type inside your conversation to leave links to an external live chat tool. The other tool opens up in a new tab and the user can continue the conversation the chatbot, incase all your agents are offline.

You can also leverage your social channels, like Facebook, Telegram, Slack for the live chat using the same method.

Developer Method

If you have a developer on the team, then you can invoke a live chat widget when user opts for such an option in the conversation.

First add a Multi Choice message to your conversation, to see if user needs immediate assistance.

In the Logical Jump settings of this message, point the answer "Yes" to "End of Conversation".

This ensures that the chatbot will send a completed response containing all the previous messages from the user.

Now, add a piece of a script to your webpage to check if the user responded to this message with an answer, "Yes" using the JavaScript method, collectchat.on for the message event.

Then you can remove the chatbot from the page, using the collectchat.unload method and invoke the live chat widget using it's API. Here is the example using Tawk.to API. details.question points to the ID of the question being asked, which will be unique for each question.

<script>
var collectchat = window.collectchat || {};
collectchat.ready = function() {
  collectchat.on('message', function(details) {
		console.log(details); // remove this after debugging
    if (details.question === "A1895B5D-0ABA-4636-93A1-4ED9C2E0B793" && details.answer === "Yes") {
          console.log('user opted for immediate assistance');
          // Remove chatbot launcher and widget
          collectchat.unload();
          // Toggle widget status of Live chat widget here
          Tawk_API = window.Tawk_API || {};
          Tawk_API.toggle();
    }
  });
}
</script>

Note

In both demos, you can see that it's a "Bot-First" approach. The bot interacts with the user and collects their details initially and only near the end of the conversation, we ask the user if they need live support. This way even if your agents are offline, the bot is able to capture user data and send it you.