The Satisfaction tab in Smartsupp tells you how well Mira AI is doing — directly from your visitors. They rate each AI conversation with a positive, neutral, or negative reaction (and can add a comment). You'll see the trends in one dashboard, spot weak answers, and improve Mira AI based on real feedback.
We've broken the guide into the following sections:
Before you start
The Satisfaction tab gives you visibility into how Mira AI performs in real conversations. From the data you'll learn:
Whether the conversation with Mira AI was successful.
Whether Mira AI gave an accurate answer.
Whether the visitor was satisfied with the solution.
Feedback on improvements you could make — including your web scrape sources.
Whether the conversation led to a product purchase.
Prerequisite: The Satisfaction tab is part of Customer Satisfaction Rating. Make sure Customer Satisfaction Rating is enabled — see Basic chat box settings for setup steps. Without it, the Satisfaction tab won't collect feedback from Mira AI.
How visitors rate Mira AI
Once Customer Satisfaction Rating is on, every Mira AI conversation can be rated by the visitor. Here's how it works on the visitor's side.
If there's no activity from the visitor for 5 seconds after Mira AI's last message, a Rate AI Chatbot button appears in the chat box.
The visitor can pick one of three reactions — positive, neutral, or negative — and optionally add a written comment.
If the visitor only picks a reaction at first and decides to leave a comment later, they can click the Leave us feedback button to add it.
How to read the Satisfaction tab
In your dashboard, open Mira AI from the left sidebar and select the Satisfaction tab. The tab is split into three sections.
A. Summary
A high-level overview of what's been happening:
The number of ratings collected.
Overall feedback on Mira AI compared to the previous period.
The selected time period.
A chart showing positive, negative, and neutral reactions to Mira AI.
Filter the chart by clicking the matching color.
B. Average Satisfaction
The average satisfaction percentage for the selected period. Use this number to track how successful your changes to Mira AI have been over time — improving sources, persona, or training Q&A pairs should push this up.
C. Rated conversations
A list of every conversation a visitor rated, with the details you need to act:
Visitor's name.
Rating (positive / neutral / negative).
Written feedback (if any).
Full conversation preview — open it to review what Mira AI said and click Correct response to send a fix straight to the Training tab.
Tip: When a rated conversation reveals a wrong answer, click Correct response in the conversation preview. The new Q&A goes straight to the Training tab so Mira AI improves immediately.
What's next?
How to start training Mira AI — turn negative ratings into better answers.
Basic chat box settings — enable Customer Satisfaction Rating in the chat box.
Introduction to Mira AI Shopping Assistant — set up Mira AI from scratch.
Mira AI Shopping Assistant FAQ — common questions about Mira AI.