How to Have You Scan Social Media for Target Users?

Question Type

User Manual Question

Quick Answer

3 Steps to Start Social Media Scanning:

Create Scout Task: Describe what demographic you want to scan
AI Starts Scanning: Scout automatically scans social media for 5-10 rounds
View Output: Demographic profile + 3 high-quality custom AI personas (can be directly used for Interview)

Time: 1-2 days to automatically complete Output: Structured demographic profile + high-quality AI personas

Complete Operation Flow

Step 1: Access Social Media Scanning Feature

Entry Points:

Homepage → "Start New Research"
Or Sidebar → "Social Media Scanning"
Or directly visit /scout/new

Step 2: Describe Scanning Target

Input Box: "Which users do you want to scan?"

Example Inputs:

Example 1: Product Testing Scenario

"Scan users discussing sparkling coffee on Xiaohongshu,understand their attitudes toward this product category"

Example 2: New Market Entry

"Scan users in Chengdu discussing coffee consumption,understand differences between Chengdu and Beijing users"

Example 3: Demographic Insights

"Scan 25-35 year old women discussing healthy snacks,understand their consumer psychology and pain points"

Input Tips:

✅ Good Description:

Clear demographic: Age, region, characteristics
Clear topic: Discussing what content
Clear purpose: Want to understand what

❌ Avoid:

Too broad: "Scan young people" (What's young? What content?)
Too narrow: "Scan users who bought our product" (Cannot find specific users on social media)
Too vague: "See how the market is" (Need specific direction)

Step 3: Plan Mode Clarification (If Needed)

AI May Ask:

Question 1: "Which platforms to scan?"

Example Answers:- "Xiaohongshu and Weibo"- "Bilibili and Douyin"- "You recommend" (AI recommends based on demographic characteristics)

Question 2: "How many scan rounds?"

Example Answers:- "10 rounds" (Recommended, more comprehensive)- "5 rounds" (Quick understanding)- "You decide" (AI suggests based on complexity)

Question 3: "Any special requirements?"

Example Answers:- "Focus on price sensitivity"- "Want to know comparison attitudes with competitors"- "None"

Step 4: AI Starts Scanning

Scanning Process Proceeds Automatically:

Phase 1: Scanning (5+ rounds)

Scout is scanning...Round 1/10: Scanning Xiaohongshu #Sparkling Coffee→ Found 12 related posts→ Extracting key information...Round 2/10: Scanning Weibo #Sparkling Coffee→ Found 8 related posts→ Extracting key information......Round 5/10: Initial scanning complete→ Auto-triggering reasoning...

Phase 2: Reasoning (Auto-Triggered)

Reasoning about user characteristics...Initial Hypotheses:1. Core demographic: 25-32 year old first-tier city women2. Main motivation: Social sharing > Energy boost3. Price sensitivity: ¥20-30 acceptable...

Phase 3: Validation (Continue Scanning)

Continue scanning with hypotheses...Round 6/10: Validate price sensitivity→ Observed multiple price discussions→ Validation result: ✅ ConfirmedRound 7/10: Validate purchase scenarios→ Observed "afternoon tea", "shopping" high frequency→ Validation result: ✅ Confirmed...Round 10/10: Scanning complete→ Generating demographic profile and AI personas...

You Can:

View scanning progress in real-time
Pause or adjust midway
Wait for completion notification

Step 5: View Output Results

Output Content:

1. Demographic Profile (Structured)

Demographics:

- Age: 25-32 years old- Gender: Primarily female (about 80%)- Region: First-tier cities- Income: Mid-to-high income (monthly 8K-20K)

Psychological Characteristics:

- Willing to try new products- Focus on quality of life- Care about product appearance- Like sharing on social media

Behavioral Patterns:

- Purchase scenarios: Afternoon tea, shopping- Decision factors: Appearance > Taste > Health- Price acceptance: ¥25-28 optimal, hesitate above ¥30- Repurchase willingness: Will repurchase if tasty, but not necessity

Pain Points:

- Worry "carbonation too strong"- Worry "coffee flavor insufficient"- Hope "packaging looks good"

2. Highlighted Content Snippets

Examples:

User A (Xiaohongshu):"Tried new sparkling coffee today, packaging super pretty!Took photos and posted to moments, friends all asked where to buy.Taste is special, carbonation stronger than expected,but 0 sugar very healthy, ¥25 acceptable."→ Key insights: Appearance important, carbonation needs adaptation, ¥25 price acceptableUser B (Weibo):"Sparkling coffee bit disappointing, ¥30 per cup too expensive.And combination of bubbles and coffee is weird,prefer straight Americano or sparkling water."→ Key insights: ¥30 price sensitive, sparkling + coffee combination has acceptance issues

3. Auto-Generated High-Quality Custom AI Personas (3)

Persona 1: Price Sensitive

Linda (28, Product Manager, Shanghai)- Income: 15-20K- Consumption characteristics: Price sensitive, rational decision-making- Focus: Value for money, user reviews- Price acceptance: ¥12-15

Persona 2: Health Anxious

Emma (32, Accountant, Beijing)- Income: 18-25K- Consumption characteristics: Focus on health, worry about sweeteners- Focus: Ingredient safety, health certification- Price acceptance: ¥18-22 (willing to pay for health)

Persona 3: Social Early Adopter

Chloe (26, Designer, Shenzhen)- Income: 12-18K- Consumption characteristics: Like trying new things, care about appearance- Focus: Photo-worthy, social sharing- Price acceptance: ¥20-30 (don't mind price much)

Can Use Directly:

✅ For subsequent Interview deep conversations
✅ For Discussion group discussions
✅ Quality: High-quality custom personas

Advanced Settings (Optional)

Setting 1: Specify Scanning Platforms

Default: AI auto-selects platforms based on demographic characteristics

Manual Specification:

Platform Selection:- Xiaohongshu (Young women consumer goods)- Weibo (Tech products, hot topics)- Bilibili (Young people, anime, tech)- Douyin (Gen Z, short videos)- X/Twitter (Overseas market, tech circles)- Reddit (Overseas market, deep discussions)

Recommendation:

Unsure which to choose → Let AI recommend
Have clear target demographic → Manually specify 2-3 platforms

Setting 2: Scan Rounds

Default: 10 rounds (Recommended)

Adjustable:

5 rounds: Quick understanding (suitable for urgent situations)
10-15 rounds: Deep understanding (recommended)
20+ rounds: Not recommended (diminishing returns)

How to Choose:

Simple demographic (like young women consumer goods) → 10 rounds
Complex demographic (like enterprise decision-makers) → 15 rounds
Niche market (like figurine collectors) → 15 rounds

Setting 3: Focus Dimensions

Default: Comprehensive scanning

Can Focus:

Dimension Selection:- Price sensitivity (Focus on price-related discussions)- Usage scenarios (Focus on when/where used)- Competitor comparison (Focus on comparison attitudes with competitors)- Pain point digging (Focus on complaints and dissatisfaction)

Example:

User: "Scan sparkling coffee users, focus on price sensitivity"Scout will:- Prioritize scanning price discussions- Analyze acceptance across different price ranges- Output detailed pricing recommendations

Real-World Cases

Case 1: New Product Testing (Most Common)

Background: Company plans to launch 0-sugar sparkling coffee, wants to understand target user attitudes

Operation Flow:

Step 1: Create Scout task

Input: "Scan users discussing sparkling coffee on Xiaohongshu/Weibo,understand their acceptance and price sensitivity for this product category"

Step 2: AI auto-scans 10 rounds

Time: 1-2 daysScanning platforms: Xiaohongshu, Weibo, DouyinScanning content: User spontaneous discussions about sparkling coffee

Step 3: View output

Demographic profile:- 25-32 year old first-tier city women- Care about appearance and social sharing- Price acceptance ¥25-28- Worry carbonation too strongGenerated AI personas: 3 custom personas→ Directly use for subsequent Interview to test product concept

Value:

Quickly understand target users (1-2 days)
Based on real user content, not imagination
Generate usable AI personas

Case 2: Regional Expansion Research

Background: Beijing coffee brand wants to enter Chengdu market

Operation Flow:

Step 1: Create Scout task

Input: "Scan users in Chengdu region discussing coffee consumption,understand differences between Chengdu and Beijing users"

Step 2: AI scans Chengdu users

Platforms: Xiaohongshu (Chengdu region), Dianping (Chengdu)Focus:- Coffee consumption frequency- Price acceptance- Preferred coffee types- Consumption scenarios

Step 3: Discover differences

Chengdu vs Beijing differences:1. Price sensitivity: Chengdu more focused on value2. Consumption scenarios: Chengdu prefers "sit and chat", Beijing fast-paced takeout3. Product preference: Chengdu prefers "tea-coffee fusion", Beijing pure coffee4. Consumption frequency: Beijing daily, Chengdu 2-3 times/weekRecommendations:- Chengdu stores need larger space (suitable for socializing)- Launch tea-coffee product line- More affordable pricing (¥20-25 vs Beijing ¥25-30)

Value:

Avoid directly copying "Beijing experience"
Reduce market entry risk
Precise regional positioning strategy

Case 3: Competitor User Research

Background: Fitness app wants to understand competitor users' true attitudes

Operation Flow:

Step 1: Create Scout task

Input: "Scan users discussing Keep fitness app,understand why they choose Keep, what dissatisfaction"

Step 2: AI scans competitor discussions

Platforms: Xiaohongshu, Weibo, BilibiliScanning content:- Why users choose Keep- Users' experience- Users' complaints and pain points

Step 3: Discover opportunities

Keep user profile:- Core users: Fitness beginners (60%+)- Choosing reasons: "Many courses", "Free"- Dissatisfaction pain points: "Course quality varies", "Many ads", "Lack personalization"Differentiation opportunities:1. Focus on intermediate-advanced users (Keep leans toward beginners)2. Provide high-quality curated courses (vs Keep many but messy courses)3. Personalized training plans (Keep lacks)4. Ad-free experience (Keep many ads)

Value:

Understand competitor users' true attitudes
Discover differentiation opportunities
Avoid homogeneous competition

FAQ

Q1: How Long Does Social Media Scanning Take?

Time:

Scanning process: 1-2 days (10 rounds scanning + AI reasoning)
Can run in background automatically, no manual participation needed

If Need Rush:

5 rounds scanning: Half day to 1 day
But depth and accuracy will decrease

Q2: Can Pause or Adjust Midway?

Yes:

Pause scanning: Save current progress
Adjust direction: Modify scanning focus
Increase rounds: If feel not deep enough

Example:

User after 5 rounds scanning:"I want to add 5 rounds, focus on price sensitivity"Scout will:- Retain first 5 rounds' scanning results- Scan 5 more rounds, focus on price discussions- Integrate output

Q3: Is Scanned Content Public?

Completely Public:

Scout only scans publicly available social media content
Does not access any private information
No privacy concerns

Data Sources:

Xiaohongshu public posts
Weibo public discussions
Bilibili public comments
Other social platform public content

Q4: What's the Quality of Generated AI Personas?

Quality Level: High-quality custom personas

High consistency: Close to real person performance
Depth: Built based on real user content
Usability: Directly use for Interview/Discussion

vs Temporary Generation:

Temporary generation: Lower consistency, shallow feedback
Custom personas: High consistency, realistic feedback

Recommendation:

Critical research: Use Scout-generated custom personas
Quick validation: Can use temporary generated personas

Q5: Can Specify Scanning Specific Demographics?

Yes, through filtering conditions:

Example 1: Age filtering

"Scan 25-35 year old women discussing healthy snacks"Scout will:- Focus on content published by 25-35 year old women- Exclude other age groups

Example 2: Regional filtering

"Scan users in Chengdu region discussing coffee"Scout will:- Focus on users in Chengdu region- Analyze regional characteristics

Note:

Don't filter too narrow (can't find enough content)
Don't filter too broad (too much noise)

Q6: Are Scout-Discovered Trends Reliable?

Reliability:

✅ Based on real user content
✅ Multiple scan rounds for cross-validation
⚠️ But still qualitative research, not statistical significance

Suitable Scenarios:

✅ Quickly discover trend directions
✅ Understand user attitudes and motivations
❌ Cannot replace large-scale quantitative research

Recommendation:

Scout discover trends → AI Interview validate → (Optional) large-scale survey confirm

Q7: Can Scan Competitors' Users?

Yes:

Scan users discussing competitors
Understand competitor user characteristics
Discover differentiation opportunities

Example:

"Scan users discussing Starbucks"Scout discovers:- High-frequency words: "Convenient", "Stable", "Nothing special"- Opportunity: Users feel Starbucks "no surprises"- Positioning direction: "Coffee experience with character"

Best Practices

Recommendation 1: Scout First Then Interview

Recommended Flow:

Step 1: Social media scan 10 rounds → Generate demographic profileStep 2: Select AI personas based on profileStep 3: Interview deep conversations for validation

Value:

Scout ensures correct direction
Interview deeply digs based on real demographics
Avoid "interviewing into thin air"

Recommendation 2: Don't Over-Rely on Single Platform

Recommendation:

Scan 2-3 platforms
Cross-validate findings
Avoid platform bias

Example:

Xiaohongshu + Weibo + Bilibili
See consistent trends → High credibility
See contradictions → Need further validation

Recommendation 3: Regularly Re-Scan

Market is Changing:

Re-Scout after 3-6 months
Discover trend changes
Adjust strategy promptly

Value:

Avoid decisions based on outdated understanding
Seize emerging opportunities

Recommendation 4: Focus on "Why" Not Just "What"

Common Mistake:

Only record "users like A"
Don't dig deep "why like A"

Correct Approach:

Ask "why" when scanning
Analyze underlying motivations
Output insights not just phenomena

Final Takeaway

"Social media scanning is simple: Describe target → AI auto-scans 10 rounds → Output demographic profile + high-quality custom AI personas. Complete in 1-2 days, directly use for subsequent research."

Remember:

✅ 3 steps to start: Create task → AI scans → View output
✅ Time: 1-2 days automatically complete
✅ Output: Demographic profile + 3 high-quality custom AI personas
✅ Recommended flow: Scout → AI Interview → (Optional) real person validation
✅ Regularly re-scan: Update understanding every 3-6 months

Related Feature: Social Media Scanning Document Version: v2.1 Update Date: 2026-02-02 Update Notes: Updated terminology and platform information