Write your interview guide

Your interview guide is the ordered list of questions, choice questions, and topics your agent covers in every interview. You write and edit it in the Interview guide section of the study page in the Study Builder, and you can steer the follow-ups and attach images, a video, or a Figma prototype item by item. Before you start

• The study must be a draft. On a live study, you can read the guide under Settings > Brief and steer the agent with research notes.

​

Choose questions or topics

When you create a study, the builder asks “How do you want to build your Interview guide?” and drafts the guide in the style you pick. Each item is one of two types:

• Question: asked as written, so every participant hears the same wording and you can compare answers across interviews.
• Topic: the agent writes its own questions about the area, which leaves room for answers you didn’t expect.

A guide can mix types, including Choice items: see Set up a choice question. To switch an item, open the Question, Topic, or Choice select on the item. Switching rewrites the item: a topic becomes a question a participant can answer directly, a question becomes the area to explore behind it, and a choice question gets options drafted for it.

​

Edit the guide

• Edit an item: click its text. Changes save automatically.
• Add an item: click Add question, then switch the new item to a topic or a choice question with the type select if you need one.
• Edit from the chat: ask the builder, for example “Merge Q4 and Q5 into one question”. Questions you paste into the chat are kept word for word as Question items.

​

Set up a choice question

Use a choice question when you want to count the answers, such as which payment method a shopper tried. To add one, set an item’s select to Choice, or ask in the chat:

Add a single-choice question on which payment method they tried first

Then shape the answer on the item:

• Single choice or Multiple choice: Single choice when exactly one option applies to each participant, Multiple choice when several can.
• Options: between 2 and 20, each with its own label. Participants see them in the order you set and can answer only from your list, so on Single choice cover every likely answer, adding an option such as “None of these” where it applies.
• Question text: write the question alone. Participants see the options under it.

The guide saves once every choice question has 2 to 20 options with distinct labels; until then, the item says what is missing.

​

Steer the follow-ups Optional

Study depth sets how many follow-ups the agent asks on every item. To change that for one item, click Follow-ups on the item:

• Follow-up instructions: where the agent should dig on this item, for example:

  If they hesitate at the payment step, ask whether it is the price, the payment methods, or trust.

  Instructions apply to this item only.
• Follow-up depth: how many follow-ups this item gets, whatever the study depth. The default, Study default, follows the study depth. With No follow-ups, the agent acknowledges the answer and moves on.

When an item gets no follow-ups, because the study depth is Express or the item is set to No follow-ups, the agent ignores its follow-up instructions. On an Express study, set a follow-up depth on the item to apply its instructions.

​

Show images, a video, or a Figma prototype to participants Optional

Attach images, a video, or a Figma prototype to an item to get reactions to screens, mockups, ads, packaging, or a product demo. The agent shows them to participants when it asks the question or explores the topic, and participants play a video in the interview.

1. On the item, click Concept testing.
2. Click Upload and pick either up to 4 images (PNG, JPEG, or WebP, up to 20 MB each) or one video (MP4, MOV, or WebM, up to 100 MB). Each item holds up to four images, one video, or one prototype. To use an attachment already in the study, click it under Reuse from this study.

The agent sees the images, so it can ask about what they show. It doesn’t watch a video: it asks participants for their reactions and follows up on what they say. To have the agent probe a specific moment of the video, name it in the item’s follow-up instructions. Clicking Concept testing on an item that has attachments turns concept testing off for that item and removes them, after you confirm. Attachments you add to a builder chat message are also added to the items they belong to. See Create a study.

​

Attach a Figma prototype

1. Open the prototype in Figma’s presentation view. Choose Share prototype, enable Anyone with the link can view, and copy the link.
2. On an empty guide item, open Concept testing and click Prototype.
3. Paste the link and click Preview. Check it in a signed-out browser too: your Figma session may give you access that participants don’t have.
4. Click Attach prototype. To reuse one, select it under Reuse from this study, preview it, and attach it.

Participants explore the prototype inside the interview and describe their reactions. In video interviews, they click Start recording and talk through their reactions while Pillow records their camera, microphone, and prototype interaction events. They submit their answer to finish. On phones, the question appears above the prototype in the same scrolling area. The selected starting screen and flow are preserved. The agent uses your question, follow-up instructions, and participants’ descriptions. It does not inspect the recording or interaction events, or score task completion. Researchers can watch a reconstructed prototype replay alongside the video answer. Replay shows the recorded cursor and click circles alongside screen and component changes. Scrolling, overlays, and animations may differ from the original session. Describe the parts you want feedback on in your question or follow-up instructions. The prototype stays live in Figma. Changes to its content or sharing settings affect the link in interviews and results. Remove it with Remove prototype, or turn off Concept testing for the item.

​

Related