> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs-parschat.parstechai.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs-parschat.parstechai.com/_mcp/server.

# Reports dashboard

> Track conversations, messages and AI performance over time.

Open your robot and go to **Report** → **Statistical Report**. **Manage Reports** gives a
live overview of how your conversations are being handled.

![Statistical Report](/_fern-files/_img/6f16ec5e26ee9615ca53d06bf69df09743065a240854f4960479794acf69c24b.webp)

## The headline figures

Three cards sum up the robot:

* **Response Time**: the average time to a reply, in minutes and seconds.
* **Total Messages**: every message sent and received.
* **Total Conversations**: how many conversations were opened.

## Filter

**Filter** at the top right narrows the whole page to one service type (such as Widget, Telegram or
Goftino) and a date range. By default the report covers the last 30 days. The badge on the button
shows how many filters are active.

## Who sends the messages

**Number of messages** splits the messages by sender:

* **Client**: your visitors
* **Operator**: replies from your team
* **AI**: the robot’s answers
* **System**: automatic messages such as welcome and auto-close notices

## AI performance

**AI performance** shows what happened to the questions the robot received:

* **Successfully processed**: the robot answered
* **Referred to operator**: the robot handed the conversation to a person
* **Not processed**: the robot did not answer
* **Other**

> **Tip**
>
> The share of **Referred to operator** and **Not processed** is the most important number on this
> page. When it is high, the robot’s knowledge doesn’t cover what your visitors actually ask.
> [Answer sources](/user-manual/conversations/answer-sources) helps you find the gaps.

## Trends over time

The chart has three tabs: **Conversation status**, **Message status** and **Cost status**. The
dropdown next to them sets the resolution: **Daily**, **Weekly** (the default), **Monthly** or
**Yearly**.

**Conversation status** draws one line per outcome: **Answered**, **Referred to operator**,
**Unanswered** and **Closed**. A rising **Referred to operator** line is the earliest sign that
your operators will need help.

## Bot performance

The **bot performance** table at the bottom lists the questions the robot reported, with the
time, the reason and the state of the conversation. It stays empty until the robot has something
to report.

> **Note**
>
> Each operator also has a personal report, under **Team** → **Members**. See
> [Operators](/user-manual/account/operators).