Interview Stages

Think of stages as chapters in your interview. Each one has its own focus, topics to explore, and insights you're trying to gather. They help you structure the conversation so it flows naturally while hitting all your research goals.

What Are Interview Stages?

Stages are sequential sections of an interview. Each stage explores specific topics or gathers particular insights. It's like chapters in a book - each has its own purpose that builds toward your overall goals. The AI moves through them automatically as the conversation progresses.

You get structured conversations with logical phases (intro, exploration, deep-dive, wrap-up) without feeling robotic. Each stage targets specific information so you're not trying to cover everything at once. You'll get reports for each stage showing what you learned in that phase, which makes patterns way easier to spot. And if you need to reference a specific document in one part of the interview, you can attach it to just that stage.

What's in a Stage?

Name: Pretty straightforward. Call it something descriptive like "Introduction" or "Career Background" or "Technical Skills Assessment."

Focus: This is the stage's main goal. It tells the AI what it's trying to accomplish and how to behave. Example: "Build rapport and understand the candidate's current role and responsibilities"

Topics: What the AI will actually ask about. These are your conversation starters. Keep them specific and actionable - max 10 words each, 3-6 topics per stage. Frame them as conversation starters, not yes/no questions. Like "Daily workflow and priorities" or "Most challenging situations encountered" or "Tools and resources currently used."

Insights: This is what you're trying to learn. What do you want to understand about the person? Be specific - max 5 words each, aim for 3-6 insights per stage. Things like "Decision-making criteria" or "Pain points and frustrations" or "Success metrics and definitions."

Stage-Specific Knowledge Documents: You can attach reference docs to individual stages. The AI uses these for context during that part of the interview. Only the first 3 pages get used, pick from your knowledge repository, and each stage can have its own doc or share. See Reference Documents for details.

Duration: There's no hard time limit. Stages naturally take 3-8 minutes based on what you're covering and how engaged the participant is.

How Stages Flow

Stages connect in sequence - each one has a clear "next stage." Here's a typical structure: Introduction (2-3 mins) → Core Exploration (5-10 mins) → Deep Dive (5-10 mins) → Closing & Reflection (2-3 mins).

The AI figures out when it's time to move on. When it thinks the current stage's objectives are met, it transitions naturally. You won't see abrupt shifts - transitions are smooth and conversational. All transitions get tracked with timestamps so you can see exactly when each phase happened.

During a stage, the AI uses the stage focus to guide its conversation style, explores the topics you defined, tries to gather the insights you specified, references any knowledge documents you attached, and extracts structured data based on the stage's configuration.

Creating Good Stages

Start broad, get specific. Early stages build rapport and context, later stages can dig deeper. Give each stage a clear purpose - if you can't explain why a stage exists, you probably don't need it. Make them build on each other logically, not jump around randomly. Balance the depth - don't cram 12 topics into one stage and leave others empty. Think about transitions - can the AI smoothly move from this stage to the next?

Common Patterns

For user research: Introduction & Context → Current Experience & Pain Points → Goals & Aspirations → Feedback & Suggestions.

For job interviews: Background & Introduction → Technical Skills Assessment → Experience & Problem-Solving → Culture Fit & Closing.

For customer discovery: Current Situation → Problems & Challenges → Current Solutions → Ideal Outcomes.

Stage Reports

Each stage generates its own report. Interview Stage Reports show what you learned during that stage of one specific interview. Interviewer Stage Reports combine insights across all your interviews for that particular stage. Quantitative Data gets extracted based on what you configured for each stage.

AI-Assisted Creation

You don't have to build stages from scratch. The platform can generate them automatically based on your interviewer's focus and description. The AI can generate complete stage sequences, suggest topics and insights, improve existing stages, and make sure everything flows right. Look for "Generate Stages" or "Enhance Stage" in your interviewer settings.

What Works and What Doesn't

Create 3-6 stages for most interviews (covers 15-30 minutes). Give stages clear, descriptive names. Use specific, actionable topics. Define measurable insights. Test your stage flow with a practice run. Start broader, then narrow down.

Don't create too many stages (>8) - makes it choppy. Don't use vague topics like "general discussion." Don't overload one stage with 10+ topics. Don't make stages that overlap in purpose. Don't skip the intro or closing. Don't forget to define what you're actually trying to learn.

Next Steps

Learn about creating, editing, and reordering stages in Managing Interviewers. See how to add stage-specific knowledge in Reference Documents.