Outset gives you a few ways to do this so you can catch any issues, refine your language, and feel confident that your study is ready to go.
1. Use the "Test interview" button on the Guide page
On the Guide tab of your study, you'll see a "Test interview" button. This lets you simulate the interview experience from a participant's perspective.
You can choose to start from a specific question or run through the full guide.
This test helps you check the flow, tone, and pacing of the AI moderator.
It’s a great way to make sure your prompts are clear and engaging.
You can also send this link to teammates so they can try it out and give feedback.
2. Run a test with synthetic AI participants
On the Recruit tab, you can test your study by launching interviews with synthetic (AI) participants. This simulates how real participants might respond.
You’ll get full transcripts of the simulated conversations.
You’ll also get summaries of the responses to help you evaluate the depth and relevance of the answers.
You won’t see analytics, but you can see how the AI moderator asked follow-ups and handled the conversation.
Synthetic tests are complimentary and don’t count against your interview quota.
This is especially useful for seeing how your guide performs with different types of answers and identifying any places where the moderator might need clearer direction.
3. Share with colleagues for feedback
If you want human eyes on the flow and tone, you can send your teammates the test interview link from the Guide page. They can walk through the experience themselves and suggest edits.
This is helpful for catching unclear language, awkward transitions, or questions that might be misinterpreted.
Testing before publishing helps ensure your study is clear, well-structured, and ready to collect high-quality insights. You can use all three testing options together to fine-tune your guide and catch any gaps before going live.
Hope this helps! If you have any further questions, please reach out to our team at [email protected] or via chat.