# Run an interview (/tutorials/running-an-interview)

# Run a user interview [#run-a-user-interview]

An interview is a structured study Stunt Double runs with persona
participants, at a scale a manual research process can't match. This tutorial
takes one from a research question to a shareable report: set it up, write the
guide, recruit participants, run a round and read what comes back.

## Set it up [#set-it-up]

Open **Interviews** in a project and choose **New interview**. Setup has three
steps: **Details**, **Discussion guide** and **Recruitment**.

On **Details**, give the interview a name, the URL participants should use and
a **Research brief**: what you are trying to learn and who you want to hear
from. **Generate video** (in beta, not on the Free plan) also cuts a short
video of the study once the report lands.

## Write the discussion guide [#write-the-discussion-guide]

Choose **Draft guide with AI** to have the brief turned into a guide you then
edit, or **Write guide manually**. A guide is organized into sections, and
each section holds a series of items. An item is either:

- A **question**, asked directly, useful for opinions, priorities, or
  recalling past experience.
- A **task**, where the participant is asked to attempt something (in your
  live product or a Figma prototype) and then reflect on it.

Mixing the two in a section works well: a task that puts the participant in a
flow, followed by questions that surface what confused them or what they
expected instead.

## Recruit participants [#recruit-participants]

Every participant is an AI persona standing in for a real user. On
**Recruitment**, **Add actors** brings in actors you already have in the
workspace, with their knowledge and history. **Generate audience** creates
one to five personas to match an audience you describe. Mix the two in the
same round: a panel of personas gives you a fast, repeatable read on a
direction before you take it to real customers.

## Measure usability [#measure-usability]

After the guide, each participant answers a short **usability questionnaire**
in character, and the round reports a usability score and grade you can
compare between rounds. It is on by default with **UMUX** (four statements,
comparable with SUS scores). Switch it to **UMUX-Lite** (two statements) or
**Off** under the interview's setup. A round that ran with it off can still be
scored afterwards with **Score usability**, from the transcripts it already
has.

## Run a round [#run-a-round]

When the guide and participant list are ready, select **Run interview**. Each
round is a complete pass through the guide with every participant. As you
make product changes, select **Re-run interview** to run another round against
the same guide and see how perception shifts, rather than starting from
scratch each time.

The pages participants visit also feed the project's topology, so an interview
fills in the map of your product the same way checklist runs do.

## Read the report [#read-the-report]

Every round produces a report with:

- **Overview**: four short blocks, each opening onto its own tab. The summary
  with the counts beneath it, **Highlights** (the findings that reached the
  most participants), **Actions** (the highest priority recommendations) and
  **Evidence** (a quote for each highlight, with the screen it was said on).
- **Findings**: the study cut, a short video of the moments behind the report,
  then the patterns that came up across participants, each backed by quotes
  drawn straight from the interview.
- **By question**: what most participants did, where they differed, and the
  outliers.
- **Recommendations**: ranked by priority, ready to track as goals.
- **Usability** and **Transcripts**: the questionnaire scores, compared with
  the previous round, and every participant's session in full.

<Screenshot
  name="interview-report"
  alt="An interview round report on its Overview tab: counts of participants, findings, questions analysed and recommendations, a summary, and the participant list."
  caption="A round's Overview: the summary and counts first, then Findings, By question and Recommendations in order."
/>

Because every finding and recommendation is tied to evidence quotes, and each
quote shows the screen it was said on, you can check a claim against what a
participant actually said and saw before acting on it.

To share a round with people outside the workspace, use **Share** on the
report and **Get share link**. The link shows the report; recordings and
transcripts stay behind sign-in. **Email report** sends it to a list of
addresses, and **Download PDF** saves a copy.

## Next steps [#next-steps]

- **[Review runs and act on them](/tutorials/reviewing-runs)**
- **[Automate multi-step workflows](/tutorials/building-workflows)**