Skip to main content
Analytics allows you to track your chat’s performance and understand how users are interacting with it. Through the available reports and metrics, you can identify improvement opportunities, discover which content generates the most interest, and ensure your chat is meeting your audience’s needs.
To access the Analytics dashboard, go to the Overview tab in the left-hand sidebar menu.
Analytics Attlas

Time Filters

In the top-right corner of the page, you’ll find the time range selector. It allows you to analyze data across different periods, making it easier to compare results and identify trends. Examples:
  • Today
  • Last 7 days
  • Last 30 days
  • Last 90 days
  • All time

Key metrics

The key metrics provide a quick overview of your chat’s health and performance. Attlas Analytics

Coverage rate

The coverage rate measures your chat’s ability to answer user questions using the information available in its knowledge base.

How to interpret

  • 🟢 Above 80%: Your chat is well-trained and effectively serves your users’ needs.
  • 🟡 Between 50% and 79%: There are knowledge gaps. Consider adding new sources or expanding existing content.
  • 🔴 Below 50%: The user experience is being impacted. A review of your training content is strongly recommended.

How to improve coverage

  • Add new knowledge sources.
  • Update outdated documents.
  • Review unanswered questions.
  • Create content around frequently requested topics.

Covarage rate guide

Learn more about coverage rate in our complete guide

Views

Represents the total number of times your chat was viewed during the selected period.

What does this metric show?

The number of people who reached your chat.

How to interpret

  • High views with few messages may indicate low engagement.
  • High views with many messages usually indicate strong interest and relevance.

Messages

Represents the total number of messages sent to your chat during the selected period.

What does this metric show?

The level of interaction generated by your chat.

How to interpret

  • A growing volume of messages indicates increasing usage.
  • A decline may suggest lower interest or reduced traffic.

Clicks

Represents the total number of clicks on triggers, links, buttons, and content shared by your chat.

What does this metric show?

The level of engagement generated by the recommendations and content presented.

How to interpret

  • A high number of clicks indicates strong interest in the shared content.
  • Low click rates may suggest the need to improve triggers or calls to action.

Performance Charts

Charts help you visualize trends and user behavior over time.

Engagement

The Engagement chart shows how users interact with your chat over time. You can switch between two different views to analyze your audience’s behavior.

Volume

The Volume view shows the relationship between the number of conversations started and the total number of messages exchanged. Metrics displayed:
  • Conversations
  • Messages
A single conversation can contain multiple messages, which is why the number of messages is usually higher than the number of conversations. Volume Example: On a given day, your chat may receive:
  • 20 conversations
  • 100 messages
This means that each conversation generated an average of 5 messages.

How to interpret

  • Many conversations with few messages may indicate quick interactions or low engagement.
  • Many conversations with many messages usually indicate interested users and deeper discussions.
  • An increase in message volume is often a sign of stronger user engagement.

Funnel

The Funnel view helps you understand the user journey from viewing your chat to starting a conversation. Metrics displayed:
  • Views
  • Clicks
  • Conversations
This view helps you understand how many people reached your chat, how many interacted with the available triggers, and how many actually started a conversation. Funnel Example:
  • 100 views
  • 20 clicks
  • 8 conversations
This means that not every click resulted in a conversation. This happens because some triggers, such as external links, videos, or files, generate clicks without creating a conversation. Question-based triggers, on the other hand, start conversations and generate messages.

How to interpret

  • Many views and few clicks may indicate that your triggers are not attractive enough.
  • Many clicks and few conversations may indicate that most triggers are directing users to external content.
  • Consistent growth in conversations is usually a sign that your chat is relevant and engaging.

Answer Quality

The Answer Quality chart helps you monitor the quality of interactions and your chat’s ability to meet user needs. You can switch between two different views.

Coverage

The Coverage view shows the average coverage rate of user messages during the selected period. This metric indicates how effectively your chat was able to answer questions using information from its knowledge base. Coverage Example: On a given day:
  • 3 conversations were started
  • 20 messages were exchanged
If the chart shows a coverage rate of 95%, it means that the vast majority of those messages were answered successfully using the trained content, although small gaps may have occurred in some responses.

Conversation Depth

The Conversation Depth view shows the average number of messages exchanged per conversation during the selected period. This metric helps you understand the depth of user interactions. Conversation Depth Example: If the chart displays a depth of 7, it means that each conversation generated an average of 7 messages between the user and the chat.

How to interpret

  • Deeper conversations often indicate stronger engagement and user interest.
  • Very short conversations may indicate that users quickly found the information they needed or lost interest.
  • A gradual increase in conversation depth is generally a positive sign of relevance and usefulness.

What is a good conversation depth?

There is no universal benchmark. A support chat may require only a few messages to solve a problem, while an educational or advisory chat may naturally generate longer conversations. The most important thing is to track how this metric evolves over time and compare it with your coverage rate.

Audience insights

Learn where your users come from, what they click on, and how they access your chat. Audience Insights

Most clicked content

The Most clicked section shows which triggers, links, and content received the most interactions.

How to use this information

  • Identify the most popular topics.
  • Understand what captures your audience’s attention.
  • Create new content related to the most accessed subjects.

Countries

The Countries section displays the geographic distribution of your users.

How to use this information

  • Identify relevant markets.
  • Discover new audiences.
  • Adapt content for specific regions.

Cities

The Cities section shows the cities where your users are located.

How to use this information

  • Plan local campaigns.
  • Identify regional opportunities.
  • Better understand your audience.

Devices

The Devices section shows which types of devices users are using to access your chat. The main device categories are:
  • Mobile
  • Desktop

How to use this information

If most of your traffic comes from mobile devices, prioritize experiences, content, and sharing strategies optimized for smartphones.

Turn data into improvements

Analytics is more than just a collection of numbers. Use these insights to identify opportunities, improve your content, expand your knowledge base, and create increasingly valuable experiences for your users. The more you understand your data, the smarter your chat becomes.
Last modified on June 3, 2026