> ## Documentation Index
> Fetch the complete documentation index at: https://docs.inceptia.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Labels

> How each conversation's results are classified

## What labels are

**Labels** are the possible outcomes of a conversation — payment promise, refusal, unavailable, voicemail, etc. They let you classify each call and later analyze the campaign's results in reports.

<Info>
  Labels are **predefined by Inceptia**. You can't create or modify labels from the platform — if you need specific labels for your use case, talk to your Inceptia representative.
</Info>

***

## Where to see them

The bot's **Labels** tab shows the full table of labels available for that bot, with:

| Column           | Description                                                       |
| ---------------- | ----------------------------------------------------------------- |
| **Label**        | The label's name.                                                 |
| **ID**           | Internal numeric identifier.                                      |
| **Group**        | The label's category or grouping.                                 |
| **Contact type** | Which type of contact it applies to.                              |
| **Final**        | Whether this label definitively closes the case (no retries).     |
| **Successful**   | Whether this label indicates a positive outcome for the campaign. |

<img src="https://mintcdn.com/inceptia/2xrQ5P1QkQTHz4UH/images/construccion/etiquetas/01-etiquetas.png?fit=max&auto=format&n=2xrQ5P1QkQTHz4UH&q=85&s=e38eeebc7d167433ea7e599ec337e4a8" alt="Table of predefined labels for the bot" width="1440" height="900" data-path="images/construccion/etiquetas/01-etiquetas.png" />

***

## How they're assigned

Labels are assigned automatically by the **Post-Conversation agent** (the red agent on the canvas). This agent runs after the call ends and analyzes the conversation's context to assign the correct label.

<Warning>
  For classification to work correctly, the Post-Conversation agent's instructions must include the list of available labels and explain when to use each one. If the labels aren't in the classifier's prompt, the bot can't assign them correctly — it's one of the [most common mistakes](/en/construccion/buenas-practicas/errores-comunes) when configuring a bot.
</Warning>

### Example instruction for the classifier

```
When the conversation ends, analyze the outcome and assign one of the following labels:

- PAYMENT_PROMISE: the contact committed to paying by a specific date.
- REFUSAL: the contact refused to pay or hung up.
- UNAVAILABLE: no answer, voicemail, or phone off.
- PAYMENT_MADE: the contact confirmed they already made the payment.
- TRANSFERRED_TO_HUMAN: the call was transferred to an operator.

Return only the label name, with no additional text.
```

<Tip>
  Always add an explicit rule that classification can only be based on what was literally said in the conversation — that the model shouldn't infer commitments that weren't mentioned, and that silence never counts as a commitment. This keeps the classifier from "hallucinating" results.
</Tip>
