We'll cover some guidelines on how to write discussion guides on Marvin so you can gather relevant insights during interview. Make sure your discussion guides are optimized to get you the best results while tagging and surfacing insights.
Add context
Include research goals, opening script, potential follow-up questions, prompts for the interviewer, and any other relevant context in the discussion guide. Marvin may analyze this data while generating Ask AI results, creating auto notes, and matching notes to labels. Adding detail helps it locate the right information each time.
Distinguish the questions by typing a "/" before them. Marvin will know these are questions, and you can tag them easily while making auto notes.
Link questions to research goals
Check if each question helps you answer something about the eventual research questions of the project.
However, It's best to frame open-ended questions, so users have a chance to explain their experience, workflow, and what they want to achieve. For example, it's better to ask "What kind of impact does research have in your organization?", instead of "Does the product team use your research reports?"
Arrange questions
Make sure all the questions in the discussion guide follow a logical order. For example, you may want to begin by understanding a user's role and goals.
While asking for feedback about your product, you may want to break up the questions by features or functions.
Create relevant labels
Think of how you would categorize the answers to each question. This will help you understand which existing label template will work for you. Or, it will help you identify new labels you need to create.
Be specific
If a participant mentions a specific feature while answering your question, try to repeat the name of the feature in follow-up questions. This will help Marvin correctly track the conversation and make better auto notes.
Ideate and create first drafts with Marvin's AI
Use Marvin's AI to create discussion guides so you can brainstorm, and access first drafts quicker. You can build on the AI-generated discussion guide to make sure it covers all your goals.
Consistent structure
Consider keeping the structure and even some questions across discussion guides consistent. This will particularly help if you do cross-project analysis on Marvin. You can also use the same labels and surface data more effectively in Ask AI and Deep Research.