How to Start an Interview?
Question Type
User Manual Question
Quick Start (3 Steps)
Step 1: Create Interview Project
Entry Point: Homepage → "Create New Interview"
Fill in Information:
- Project Name: e.g., "Sparkling Coffee New Product Test"
- Research Objective: Brief description (within 100 characters)
- Project Type:
- Interview Chat (one-on-one in-depth interviews)
- Discussion Chat (group discussion)
Step 2: Select AI Personas (5-10 people)
Method 1: Search from Public Library
- Enter keywords: e.g., "25-35 years old female white-collar coffee"
- Filter criteria: age, gender, occupation, interests
- Recommended quantity:
- Interview: 5-10 people
- Discussion: 3-8 people
Method 2: Use Recommended Personas
- System automatically recommends based on your research objective
- Covers different user segments
Method 3: Choose from "My Personas"
- Previously saved or created Tier 3 personas
- Quick reuse
Step 3: Prepare Interview Questions (3-5 questions)
Sample Question Templates:
Tips:
- No need to design all questions; AI will intelligently follow up
- Focus on core questions (3-5 only)
Click "Start Interview"
During Interview Execution
Interview Chat (One-on-One)
AI Auto-Executes:
- Parallel interviews with 5-10 personas
- 7-10 conversation rounds per persona
- Real-time progress display
You Can:
- Monitor interview progress in real-time
- Adjust questions mid-way (for specific personas)
- End certain interviews early
Estimated Duration:
- 5 people: approximately 50-70 minutes
- 10 people: approximately 80-100 minutes
Discussion Chat (Group Discussion)
AI Moderates Discussion:
- Brings together 3-8 personas simultaneously
- Facilitates opinion exchange
- Controls speaking pace
You Can:
- Watch discussion in real-time
- Insert new questions mid-way
- Adjust discussion direction
Estimated Duration:
- 30-40 minutes
After Interview Completion
Auto-Generated Report
Content Includes:
- Interview Summary: Key findings from each persona
- Highlight Dialogue Snippets: As evidence
- Consensus and Divergence: Auto-identified
- User Profiles: 7-dimension detailed information
- Improvement Recommendations: Actionable suggestions
View Location:
- Interview project detail page → "View Report"
- Or Homepage → "My Projects" → Select project
Detailed Step Instructions
Creating Interview Project
Project Name Suggestions:
- ✅ Descriptive: e.g., "Sparkling Coffee New Product Test"
- ✅ Include timing: e.g., "2024Q1 Membership System Redesign"
- ❌ Avoid vague: e.g., "Project 1", "Test"
Research Objective:
- Briefly describe what you want to learn
- Example: "Understand acceptance and pricing sensitivity of sparkling coffee among 25-35 year old working women"
Project Type Selection:
| Type | Use Case |
|---|---|
| Interview Chat | Explore "why", deep dive |
| Discussion Chat | Compare viewpoints, quickly identify consensus and divergence |
Selecting AI Personas
Method 1: Keyword Search
Search Tips:
- Demographics: age, gender, occupation, city
- Behavioral traits: fitness, coffee, skincare, finance
- Scenario traits: commuting, office, home, social
Example:
Filter Criteria:
- Age: 18-25, 25-35, 35-45, 45-55, 55+
- Gender: Male, Female
- Occupation: Product Manager, Programmer, Designer, Student, Stay-at-home Mom
- Interests: Tech, Beauty, Fitness, Food, Travel
Method 2: System Recommendations
Auto-recommend Based on Research Objective:
- Enter research objective
- System analyzes keywords
- Recommends best-matching 5-10 personas
Example:
- Research Objective: "Test sparkling coffee acceptance"
- Recommended Personas:
- 2 health-conscious types (focus on ingredients)
- 2 price-sensitive types (focus on value)
- 1 social trendsetter type (focus on appearance)
Method 3: My Personas
Quick Reuse:
- Select from "My Personas" - previously saved or created personas
- Suitable for repeated research or continuous tracking
Scenarios:
- Certain research frequently uses the same set of personas
- Build "My Commonly Used Persona Library"
Preparing Interview Questions
Question Design Principles
Open-Ended Questions:
- ✅ "What's your first impression of this product?"
- ❌ "Do you like this product?" (closed-ended)
Specific Scenarios:
- ✅ "In what scenarios would you use it?"
- ❌ "Would you use it?"
Avoid Leading:
- ✅ "What do you think of this price?"
- ❌ "You think this price is very reasonable, right?"
Question Template Library
New Product Testing:
- What's your first impression of [product name]?
- When you hear this name/see this product, what comes to mind?
- In what scenarios would you use/purchase it?
- Can you accept this [price] point? Why or why not?
- Compared to products you usually use, what are the advantages/disadvantages?
Feature Testing:
- What pain points/needs do you have in daily life?
- How do you currently solve this problem?
- Would [feature description] help you?
- Compared to existing solutions, which do you think is better?
- Any suggestions for improvement?
Pricing Testing:
- What's your usual budget for this category?
- What do you think of this [price] point?
- At what price would you "buy immediately"?
- At what price would you "hesitate"?
- If the price increases/decreases, would your decision change?
Interview Execution Tips
Interview Chat Tips
Parallel Management:
- AI automatically conducts parallel interviews with 5-10 people
- You can monitor each interview's progress in real-time
- One failed interview doesn't affect others
Mid-Way Adjustments:
- If a persona's responses are particularly valuable, you can extend the interview
- If a persona goes off-topic, you can end early
Save Draft:
- You can save as draft during the interview
- Resume next time without losing progress
Discussion Chat Tips
Moderator Guidance:
- AI moderator actively guides discussion
- Controls speaking pace, ensures everyone has opportunity to express
- Identifies opinion conflicts, actively facilitates in-depth discussion
Insert Questions Mid-Way:
- If discussion gets stuck on a point, you can insert new questions
- Adjust discussion direction
End Conditions:
- Auto-ends after opinions fully collide
- Or manually end discussion
Common Questions
Q1: Must I prepare interview questions?
Recommended but not mandatory.
- Prepare 3-5 core questions; AI will automatically follow up and expand
- If unprepared, you can choose "Exploration Mode" where AI auto-designs questions based on persona characteristics
Q2: How many personas should I select?
Recommended Quantity:
| Research Type | Recommended Number |
|---|---|
| Interview | 5-10 people |
| Discussion | 3-8 people |
Reasons:
- Less than 5: Incomplete coverage, may miss key viewpoints
- More than 10: Information overload, difficult to analyze
Q3: Can I replace personas mid-interview?
Interview Chat:
- Yes. Failed interview personas can be replaced
- Doesn't affect other completed interviews
Discussion Chat:
- No. Cannot replace personas after discussion starts
- Recommend confirming persona selection beforehand
Q4: When is the interview report generated?
Auto-Generated:
- Interview: Auto-generated after all interviews complete
- Discussion: Auto-generated after discussion ends
Estimated Duration:
- 5-10 minutes
Best Practices
Recommendation 1: Start with Small-Scale Testing
First-Time Use:
- Select 5 personas for Interview
- Familiarize with process and output results
- Then expand to 10 people
Recommendation 2: Combined Use
Best Strategy:
- Phase 1: Interview deep dive (5-8 people)
- Phase 2: Discussion comparison validation (3-5 people)
Advantages:
- First understand individual motivations
- Then observe group opinion collisions
- More comprehensive insights
Recommendation 3: Continuous Tracking
Build "My Commonly Used Persona Library":
- Save frequently used high-quality personas
- Directly reuse for subsequent research
- Observe same set of personas' reactions to different products
Scenarios:
- Certain research frequently uses same set of personas
- Can compare reaction changes across different time periods
Related Feature: Interview vs Discussion Document Version: v2.1