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

# What the agent collects

> Qualify buyers during the conversation and deduce intent afterwards.

Open **Agent → Settings → Data to Collect**.

<Note>
  **Inkly AI can design this for you.** Click **Inkly AI** and try: “Ask about the visitor's role and current workflow when it feels natural. After each session, deduce buying intent, buying stage, main objection, and pain point using clear rules.”
</Note>

## During the conversation

Add two or three questions the guide should ask naturally.

| Question                        | When to ask                               |
| ------------------------------- | ----------------------------------------- |
| What are you trying to improve? | After they show interest in a workflow.   |
| How large is your team?         | When rollout or pricing becomes relevant. |
| What do you use today?          | When migration or comparison comes up.    |

Use **Description** to tell the guide when and how to ask. Make a question required only when the answer is essential.

## After the conversation

Use deductions for buyer intelligence the guide should never ask directly.

| Result         | Rule                                                                  |
| -------------- | --------------------------------------------------------------------- |
| Buying intent  | High when they mention a timeline, budget, procurement, or next step. |
| Buying stage   | Separate casual research from evaluation or procurement.              |
| Main objection | Capture their strongest concern.                                      |
| Pain point     | Capture the problem that started the evaluation.                      |

Write rules another person could apply. Results appear in [Sessions](/measure/sessions) and [People](/measure/people), alongside the conversation that produced them.