> For the complete documentation index, see [llms.txt](https://docs.wonderchat.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.wonderchat.io/setup-guides/analytics.md).

# Analytics

### How to Access Analytics

To open the Analytics page:

1. Log in to your Wonderchat dashboard.
2. From the left-side navigation menu, click **Analytics**.

<figure><img src="/files/Ov0xKkvAWgHNmYaIEfN0" alt=""><figcaption></figcaption></figure>

3. Select the chatbot you want to analyze (or choose **All Chatbots**).
4. Use the date picker or one of the quick date filters to view analytics for a specific time period.

<figure><img src="/files/tJRXlpqzoSjbk6yhpflE" alt=""><figcaption></figcaption></figure>

## Overview

### Overall Customer Satisfaction

Displays customer satisfaction based on feedback collected after conversations.

#### Metrics

**Average Score**

* Shows the average satisfaction rating received during the selected period.
* Ratings range from dissatisfied to very satisfied.

**Satisfaction Distribution**\
The colored bar shows the percentage of users in each satisfaction category:

* 😍 Very Satisfied
* 😊 Satisfied
* 😐 Neutral
* 🙁 Dissatisfied
* 😭 Very Dissatisfied

This widget helps monitor overall customer sentiment.

<figure><img src="/files/s7Vq82tCqhJRDZ5fySjO" alt=""><figcaption></figcaption></figure>

***

### Resolution Rate

Shows how effectively your chatbot resolves customer conversations without requiring additional assistance.

#### Metrics

**Overall Resolution Rate**

* Percentage of chats successfully resolved.

**Resolved Chats**

* Total number of conversations marked as resolved.

<figure><img src="/files/cdqppn0EcUGzNfybpQtX" alt=""><figcaption></figcaption></figure>

***

### Number of Chat Messages

Displays the total number of chat messages exchanged over time.

#### Purpose

This graph helps identify:

* Usage trends
* Peak traffic days
* Seasonal fluctuations
* Changes after updates or new content

<figure><img src="/files/WAh0ZhuoohU9szckAtAS" alt=""><figcaption></figcaption></figure>

***

### Confidence Level

Shows the AI confidence score for chatbot responses.

When confidence tracking is unavailable, the widget displays:

Confidence metrics help determine how certain the AI was when generating responses.

***

## Usage Metrics

### Number of Chat Sessions

Displays the total number of chatbot conversations started during the selected period.

A session begins when a visitor starts interacting with the chatbot.

***

### Unique Visitors

Shows the number of individual visitors who interacted with the chatbot.

Each visitor is counted only once regardless of how many conversations they initiate.

***

### Widget Engagements

Measures how many times users interacted with the chatbot widget.

Examples include:

* Opening the widget
* Clicking suggested questions

This metric indicates how attractive and visible your chatbot is to website visitors.

<figure><img src="/files/1fJajE1KeX1LQGhoQpbQ" alt=""><figcaption></figcaption></figure>

***

## Conversation Quality

### Good Chats

Displays the number of conversations that received positive feedback.

This metric helps identify successful customer interactions.

### Bad Chats

Displays the number of conversations that received negative feedback.

A high number may indicate:

* Missing knowledge
* Incorrect answers
* Poor conversation quality

Review these conversations to improve chatbot performance.

***

## Response Speed

Measures how quickly the chatbot responds to users.

Two response metrics are displayed:

### TTFT (Time To First Token)

Measures how long it takes before the chatbot starts generating a response.

Lower TTFT results in a more responsive user experience.

### Full Response

Measures the total time required to generate the complete responses.

For both TTFT and Full Response, the following statistics are shown:

| Metric | Description                                 |
| ------ | ------------------------------------------- |
| AVG    | Average response time                       |
| P50    | Median response time                        |
| P90    | 90% of responses are faster than this value |
| P99    | 99% of responses are faster than this value |

These metrics help identify slow responses and monitor system performance.

<figure><img src="/files/6ScbTyNWHAnJveFnN1zh" alt=""><figcaption></figcaption></figure>

***

## Geography

### Chats by Location

Displays where chatbot conversations originated.

Countries are ranked by total number of chats.

This helps identify:

* Primary customer markets
* Geographic trends
* Regional adoption

<figure><img src="/files/AG8KGrc2PZPkqpELUR68" alt=""><figcaption></figcaption></figure>

***

## Channels

### Chats by Channel

Shows the distribution of conversations across available communication channels.

Example channels include:

* Web App
* Web Call

This helps understand where customers prefer interacting with the chatbot.

<figure><img src="/files/ITvxzdooGJNPBbPJzdoR" alt=""><figcaption></figcaption></figure>

***

## Escalations

### Number of Escalated Chats

Displays the number of conversations that were transferred from the chatbot to a human agent.

Escalations may occur when:

* The chatbot cannot answer the question.
* A human handover is requested.
* Business rules require agent intervention.

Monitoring escalations helps identify knowledge gaps.

***

## Average Chat Duration

Shows the average time users spend in chatbot conversations.

Longer conversations may indicate:

* Complex support requests
* High engagement

Shorter conversations may indicate:

* Quick issue resolution
* Users leaving early

This metric should be evaluated alongside customer satisfaction.

***

## Sessions by Message Count

Displays the number of sessions grouped by how many messages were exchanged.

Example:

| Messages per Session | Meaning                     |
| -------------------- | --------------------------- |
| 1                    | User sent only one message  |
| 2                    | Two-message conversations   |
| 3                    | Three-message conversations |
| 4                    | Four-message conversations  |
| 5+                   | Longer conversations        |

This chart helps understand conversation depth and engagement.

<figure><img src="/files/MMlXPrX57U2ABQy6PEDs" alt=""><figcaption></figcaption></figure>

***

## Most Visited Links

Lists the links shared most frequently by the chatbot.

For each link, the following information is shown:

* URL
* Number of clicks

An additional summary displays:

**Total clicks across all links**

Use this report to determine which resources customers access most often.

The **Export as CSV** button allows downloading the complete report.

<figure><img src="/files/pFBm9YfU0ftg6KSgljW7" alt=""><figcaption></figcaption></figure>

***

## Top Questions from Users

Shows the questions most frequently asked by users.

For each question, the report displays:

* Question text
* Number of times it was asked

This report is useful for:

* Identifying common customer needs
* Discovering missing knowledge
* Improving chatbot training
* Creating FAQs
* Optimizing documentation

The **Export as CSV** button downloads the full list for further analysis.

<figure><img src="/files/gOEzqxstxjWbIm0xUILj" alt=""><figcaption></figcaption></figure>

***

### Best Practices

* Monitor **Customer Satisfaction** and **Resolution Rate** regularly to evaluate chatbot quality.
* Review **Bad Chats** and **Escalated Chats** to identify areas where the chatbot needs additional training.
* Use **Top Questions** to expand your knowledge base and FAQ content.
* Track **Response Speed** to ensure users receive timely answers.
* Analyze **Chats by Location** and **Chats by Channel** to understand your audience and optimize deployment strategies.
* Export reports periodically for historical tracking and business reporting.

This documentation provides a complete overview of the Analytics page and how each metric can be used to measure and improve chatbot performance.