After Creating an Interview Project, How to Search and Select the Most Suitable AI Personas from the Public Persona Library? What Keywords Should I Use?

Question Type

User Manual Question

---

Search Strategies

Method 1: Keyword Search (Most Common)

Input Content:

• Demographic characteristics: Age, gender, occupation, city
• Behavioral characteristics: Fitness, coffee, skincare, financial planning
• Scenario characteristics: Commuting, office, home, social

Examples:

---

Method 2: Tag Filtering

Common Tag Categories:

Demographics:

• Age: 18-25, 25-35, 35-45, 45-55, 55+
• Gender: Male, Female
• Occupation: Product Manager, Developer, Designer, Student, Stay-at-home Mom
• Income: 5-10K, 10-20K, 20-30K, 30K+

Lifestyle:

• Health-conscious, Efficiency-focused, Social butterfly, Homebody
• Early riser, Night owl
• Exercise lover, Book lover

Consumption Characteristics:

• Price sensitive, Quality-first, Brand loyal
• Early adopter, Cautious observer
• Primarily online shopping, Offline shopping

Interest Areas:

• Tech, Beauty, Fitness, Food, Travel
• Gaming, Music, Reading, Photography

---

Method 3: Similarity Search

Use Case: Already have target user profile

Steps:

1. Input target user description (within 200 words)
2. System automatically matches similar personas
3. Sort by similarity
4. Select 3-10 closest matches

Example:

---

Selection Recommendations

Quantity Recommendations

Diversity Recommendations

Goal: Cover different user types

Example: Test Sparkling Coffee

---

Frequently Asked Questions

Q1: What if I can't find suitable personas?

Solutions:

1. Broaden scope: Relax filtering criteria
2. Use Scout: Let Social Media Scout observe relevant social media, generate new personas
3. Create custom personas: Import your own user interview data

Q2: Why do I get so many personas? How to filter?

Recommendations:

1. Sort by similarity
2. View persona detail pages, confirm 7-dimension information
3. Cluster: Group similar personas into categories, select 1-2 representatives from each

Q3: What is the quality of the Public Persona Library?

Quality Assurance:

• ✅ Built based on real research contexts (300,000+ real research questions)
• ✅ Built based on real social media observations (each persona has 3000-word observation record)
• ✅ Stable quality, consistency score of 80 (close to real human baseline of 81)
• ✅ Covers 90% of general research scenarios
• ✅ For higher quality or simulating specific users, can create custom personas (consistency score 85)

Data Source Transparency:

• Social media platforms: Xiaohongshu, Douyin, Twitter, Instagram, TikTok
• Observation depth: 10-15 rounds of in-depth observation
• Data scale: Each persona corresponds to 15 tool calls

---

Related Feature: AI Persona Three-Tier System Document Version: v2.1 Updated: 2026-02-02 Update Notes: Updated terminology and platform information