> 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

The Analytics page shows how your AI agent is doing: customer satisfaction, resolution rate, message volume, live chat, closed tickets, channels, locations and the questions customers ask most.

## Open Analytics

1. Click **Analytics** in the left sidebar.
2. Pick an agent from the agent menu.
3. Pick a date range. Click the date picker and choose a preset (**Today**, **Yesterday**, **Last 7 days**, **Last 30 days**, **Last 90 days**, **Last 12 months**, **Month to date**, **Last month** or **Year to date**), or select your own start and end dates on the calendar.

Every card on the page updates for the agent and dates you picked.

<figure><img src="https://3760420285-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0WUvfmkdMPXz3x7NXP8i%2Fuploads%2Fgit-blob-39394b726a6ab34069e96d755cc9f739273d2c72%2Fanalytics-overview.png?alt=media" alt="The Analytics page with the agent menu, date picker, Export button, and the Customer satisfaction, Resolution rate and Messages cards"><figcaption><p>The top of the Analytics page</p></figcaption></figure>

***

## Customer satisfaction

The average satisfaction score for the period, with a rating (**Outstanding**, **Good**, **Ok** or **Poor**) and a bar showing how chats split across **Very satisfied**, **Satisfied**, **Neutral**, **Dissatisfied** and **Very dissatisfied**. Satisfaction is predicted from your chat logs by AI.

## Resolution rate

The share of chats the AI resolved, with a rating and the number of resolved chats.

To learn how both scores are worked out, see [Understanding Resolution Rate and Satisfaction](https://docs.wonderchat.io/setup-guides/understanding-resolution-rate-and-satisfaction.md).

***

## Messages

The total number of messages in the period, with a graph of messages per day so you can spot busy days and changes after an update.

Below the graph, tiles show:

| Tile                       | What it counts                                                                                                    |
| -------------------------- | ----------------------------------------------------------------------------------------------------------------- |
| **Chat sessions**          | Conversations started.                                                                                            |
| **Avg. messages per chat** | The average number of messages in a conversation.                                                                 |
| **Avg. chat duration**     | How long a conversation lasts on average.                                                                         |
| **Escalated chats**        | Conversations handed to your team.                                                                                |
| **Message credits used**   | Credits spent. Advanced models and longer context use more per message.                                           |
| **Unique visitors**        | Visitors who loaded a page with your widget, counted once per visitor per day.                                    |
| **Widget opens**           | Clicks on the widget button, one per session.                                                                     |
| **Rated good**             | Chats the visitor rated as good.                                                                                  |
| **Rated bad**              | Chats the visitor rated as bad.                                                                                   |
| **First response**         | The average time until the first words of a reply appear. Hover the info icon to see how long a full reply takes. |

{% hint style="info" %}
**Unique visitors** is counted from April 23, 2026 and **Widget opens** from January 26, 2026. Earlier days count as 0.
{% endhint %}

***

## Live chat

How your team handled [live chats](https://docs.wonderchat.io/setup-guides/human-handover-for-your-chatbots/how-to-set-up-live-chat-within-wonderchat.md):

* **Resolved**: live chats closed in the period.
* **Open**: live chats waiting for a reply right now.
* **Ongoing**: live chats in conversation right now.

**By agent** splits resolved or open chats by teammate. Chats nobody on the team replied to are shown last, as **Visitor resolved** or **No agent yet**.

If live chat is off for the agent, the card says so with a **Turn on live chat** button.

***

## Tickets closed

How many tickets and live chats your team closed in the period, and **By channel**: the channel each one came in on, such as Live chat, Email handover, AI escalation, Connected inbox, Hosted address, WhatsApp or SMS. Spam, merged duplicates and sample tickets aren't counted.

<figure><img src="https://3760420285-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0WUvfmkdMPXz3x7NXP8i%2Fuploads%2Fgit-blob-a713d775190b29546f6550ccfeb461cf9e60413f%2Fanalytics-tickets-closed.png?alt=media" alt="The Tickets closed card with the total and a By channel breakdown"><figcaption><p>Tickets closed, by the channel they came in on</p></figcaption></figure>

See the [Inbox](https://docs.wonderchat.io/setup-guides/inbox-feature.md) guide for working tickets.

***

## Chats by location

Where conversations came from, ranked by country. Click a country to see its states or regions.

## Chats by channel

How conversations split across the places your agent is installed, such as Web, WhatsApp, Slack or MS Teams. Web chats are split further into Desktop and Mobile.

## Messages per chat

How many chats had 1, 2, 3, 4, or 5 or more messages, with the average. Many one-message chats can mean visitors aren't getting what they need from the first answer.

***

## Top questions

The questions customers ask most, with how often each was asked. Use it to find gaps in your knowledge base and to write FAQs. Click **Export CSV** to download the full list.

## Most visited links

The links in your agent's answers that visitors click most, with the click count for each and the total across all links. Click **Export CSV** to download it.

***

## Export your analytics

Click **Export** at the top of the page. Under **This agent** or **All agents**, choose:

* **Totals**: one row per agent for the selected dates.
* **By week** or **By month**: one row per agent per week or month.

The file downloads as a CSV.

***

## Plans

**Customer satisfaction** and **Resolution rate** are on every plan. The other cards are on the **Starter** plan and above. On the Free plan they're blurred, with a button to see plans.

***

## Tips

* Check **Customer satisfaction** and **Resolution rate** regularly to see how the agent is doing.
* Read chats that were **Rated bad** or escalated to find where the agent needs more training.
* Use **Top questions** to add missing content to your knowledge base.
