Would you like to know how end users experience the Agent? Within Watermelon, there are several ways to collect feedback during or after a conversation with the Agent. In this article, we’ll explain the available options and how to set them up.
1. Thumbs up/down in the chat widget
End users can give a thumbs up or thumbs down for each response. This feedback is visible in the Inbox, linked to the specific message.
This gives you quick insights into which answers are well received and where there may be room for improvement.
2. Have the Agent ask for feedback
You can instruct the Agent to ask for feedback at the end of a conversation. This can be configured in Domain Knowledge.
Add an instruction such as: “Ask for feedback at the end of the conversation.”
Note: This question will only be asked if it is clear to the Agent that the conversation has ended. If a user closes the chat immediately after receiving an answer—without saying something like “okay, thanks”—the feedback request might not be triggered.
3. Request feedback in the welcome message
Another way to request feedback is via the welcome message of the chat widget. You can include a link to a feedback form, for example.
This allows you to invite users at the start of the conversation to share their opinion.
Example text: “Our AI Agent is still learning. Help us improve by sharing your feedback through this form after the conversation.”
You can update this text in the chat widget settings.
4. Automatically request feedback upon conversation closure (Feature Request)
A feature request has been submitted to automatically ask for feedback when a conversation ends.
Would you like to see this functionality added? You can cast your vote for this request via our feedback portal.