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

# Write your interview guide

> Build the list your agent works through, as scripted questions, exploratory topics, or a mix of both.

Your interview guide is the ordered list your agent covers in every interview. Each item is either a **Question**, the literal wording your agent sends, or a **Topic**, an area it explores in its own words. You write the guide in the Study Builder and can keep reworking it until the study goes live.

## Choose a style

Choose between questions and topics when you create a study, right after the interview type question and before Pillow drafts anything:

> How do you want to build your Interview guide?

* **Questions**: You control the exact questions, the agent keeps smart follow-ups
* **Topics**: You set topics, the agent comes up with its own questions based on each answer

Your answer decides what the builder drafts from there. Guides can mix both kinds, and every item has a dropdown to switch it.

## Edit your guide

Open **Interview guide** in the builder sidebar. Items are cards you can edit, drag to reorder, or delete, and changes save automatically. You can also rework the guide from the chat, for example "Merge the last two questions into one".

Three behaviors are worth knowing:

* **Switching an item's kind rewrites it**: a topic becomes wording a participant can answer directly, and a question becomes the area to explore behind it.
* **Your own wording is kept**: questions you paste in are preserved exactly as written.
* **The main goal locks once you edit it**: Pillow keeps that one-line summary in step with your guide until you write your own.

## Steer the follow-ups <Badge>Optional</Badge>

Click **Follow-ups** on a card to tell your agent how to dig into that item:

> If they hesitate, dig into pricing.

Instructions apply to that item alone and override the general rules for it. They steer where your agent digs, while [interview depth](/guides/create-study#interview-depth) still sets how much. Leave the field empty and the item follows your depth setting.

## What's next

<CardGroup cols={2}>
  <Card title="Interview guide" icon="target" href="/concepts/interview-guide">
    How questions and topics shape each interview.
  </Card>

  <Card title="Create a study" icon="plus" href="/guides/create-study">
    Configure your agent, outreach, and reward.
  </Card>

  <Card title="Notes" icon="notebook-pen" href="/guides/agent-notes">
    Set rules your agent follows throughout.
  </Card>

  <Card title="Test vs. live mode" icon="flask-conical" href="/guides/test-vs-live">
    Try your questions before you launch.
  </Card>
</CardGroup>
