​

Glood’s Trending Products: Real-Time Momentum Tracking

​

Algorithm Architecture

​

Core Components

​

1. Multi-Signal Trend Detection

​

Velocity Metrics

trend_score = calculate_trend({
    'view_velocity': views_t / views_t-1,
    'purchase_acceleration': (sales_t - sales_t-1) / time_delta,
    'cart_momentum': cart_adds_rate_change,
    'search_spike': search_volume_derivative
})

​

Signal Sources

• Direct Interactions: Views, clicks, cart adds
• Purchase Data: Order velocity, unit sales acceleration
• Search Behavior: Query frequency, search-to-purchase rate
• External Signals: Social media mentions, influencer coverage

​

2. OpenAI GPT-4: Context Understanding

​

Semantic Trend Analysis

• Category Context: Understands why products trend together
• Seasonal Relevance: Identifies season-appropriate trends
• Event Correlation: Links trends to real-world events
• Description Mining: Extracts trend-worthy features from product text

​

Social Signal Processing

• Analyzes review sentiment for viral potential
• Identifies products mentioned in trending content
• Understands cultural and regional trend variations
• Predicts trend longevity based on context

​

3. Time Series Analysis

​

Trend Patterns We Detect

​

Advanced Techniques

• ARIMA Models: Forecast future trend trajectory
• Changepoint Detection: Identify trend inflection points
• Fourier Analysis: Detect periodic patterns
• Wavelet Transform: Multi-scale trend analysis

​

4. Glood Predictive Models

​

Trend Lifecycle Stage

1. Emerging: Just starting to gain traction
2. Accelerating: Rapid growth phase
3. Peaking: Maximum velocity
4. Plateau: Sustained high interest
5. Declining: Losing momentum

​

Trend Sustainability Score

sustainability = model.predict({
    'acceleration_rate': trend_acceleration,
    'breadth': unique_user_percentage,
    'depth': repeat_interaction_rate,
    'category_fit': category_trend_alignment,
    'inventory': stock_availability
})

​

Trend Detection Strategies

​

1. Micro-Trend Identification

• Influencer Early Signals: Products shared by trendsetters
• Geographic Hotspots: Regional trends expanding nationally
• Demographic Pioneers: Age/interest group leading adoption
• Category Crossovers: Products trending outside typical category

​

2. Macro-Trend Alignment

• Sustainability Trends: Eco-friendly products gaining traction
• Health & Wellness: Items aligned with wellness movements
• Technology Adoption: New tech categories emerging
• Cultural Moments: Products tied to cultural events

​

3. Viral Prediction

• Social Shareability: Visual appeal and uniqueness
• Price Point Analysis: Accessible luxury or incredible value
• Review Sentiment: Exceptional satisfaction indicators
• Influencer Affinity: Products likely to be promoted

​

Real-time Processing Pipeline

​

Data Ingestion (Every Second)

Events Processed:
  - Product views: 10,000/sec
  - Cart additions: 1,000/sec
  - Purchases: 100/sec
  - Searches: 5,000/sec

​

Aggregation Windows

• 1-Hour Window: Immediate spike detection
• 24-Hour Window: Daily trend confirmation
• 7-Day Window: Weekly pattern recognition
• 30-Day Window: Monthly trend validation

​

Scoring Algorithm

final_score = (
    0.3 * velocity_score +
    0.25 * acceleration_score +
    0.2 * breadth_score +
    0.15 * sentiment_score +
    0.1 * prediction_confidence
) * freshness_decay

​

Personalized Trending

​

Segment-Specific Trends

​

Age Demographics

• Gen Z: Fast fashion, tech gadgets, viral items
• Millennials: Sustainable products, home improvement
• Gen X: Premium brands, family items
• Boomers: Health products, traditional brands

​

Interest-Based

• Fashion Forward: Latest runway-inspired trends
• Tech Enthusiasts: Newest gadget releases
• Home Decorators: Seasonal decor trends
• Fitness Buffs: Workout gear and supplements

​

Geographic Trends

• Regional: Climate-specific trending items
• Urban vs Rural: Different lifestyle product trends
• International: Global trend adaptation

​

Quality Control

​

False Positive Prevention

• Bot Detection: Filter automated traffic
• Manipulation Detection: Identify coordinated behavior
• Inventory Checks: Don’t trend out-of-stock items
• Price Anomaly Detection: Exclude pricing errors

​

Diversity Requirements

• Category Balance: Trends from multiple categories
• Price Range: Mix of price points
• Brand Distribution: Not dominated by single brand
• Style Variety: Different aesthetics represented

​

Performance Metrics

​

Accuracy Metrics

• Trend Prediction Accuracy: 78% correct 7-day forecasts
• False Positive Rate: < 5% non-trending items shown
• Coverage: 95% of actual trends detected
• Lead Time: Detect trends 3-5 days before peak

​

Business Impact

• CTR on Trending: 52% click-through rate
• Conversion Rate: 31% higher than average
• AOV Impact: 23% higher order values
• Discovery Rate: 67% of users discover new products

​

Trend Categories

​

Seasonal Trends

Spring: Garden supplies, outdoor furniture, spring fashion
Summer: Swimwear, BBQ accessories, travel gear
Fall: Back-to-school, Halloween, cozy apparel
Winter: Holiday decorations, winter sports, gifts

​

Event-Driven Trends

Super Bowl: Team merchandise, party supplies, TVs
Valentine's Day: Jewelry, flowers, romantic gifts
Black Friday: Electronics, fashion, home goods
Back-to-School: Supplies, backpacks, dorm essentials

​

Viral Trends

TikTok Made Me Buy It: Products featured in viral videos
Instagram Worthy: Photogenic products
Celebrity Endorsed: Items worn/used by celebrities
Meme Products: Items that become internet sensations

​

Implementation Examples

​

Fashion Trending

{
  "trending_now": [
    {
      "product": "Oversized Blazer",
      "trend_score": 0.89,
      "velocity": "+145% this week",
      "stage": "accelerating",
      "reason": "Featured in viral fashion posts"
    }
  ]
}

​

Tech Trending

{
  "trending_now": [
    {
      "product": "Smart Home Hub",
      "trend_score": 0.92,
      "velocity": "+200% this week",
      "stage": "emerging",
      "reason": "New release gaining rapid adoption"
    }
  ]
}

​

Continuous Improvement

​

Learning Mechanisms

1. Outcome Tracking: Monitor if predicted trends materialize
2. Feedback Integration: Learn from user engagement
3. Pattern Evolution: Adapt to changing trend dynamics
4. Seasonal Calibration: Adjust for seasonal variations

​

Model Updates

• Hourly: Trend scores recalculated
• Daily: Model weights adjusted
• Weekly: New patterns incorporated
• Monthly: Full model retaining

​

Glood Dashboard Configuration

​

Setting Up Trending Products

1. Basic Configuration
   Section Name: "Trending Now"
   Algorithm: Trending Products
   Time Window: 24 hours
   Number of Products: 8
   Update Frequency: Every hour
   Minimum Views: 10
   
2. Trend Window Options
   • 1 Hour: Flash trends, viral products
   • 24 Hours: Daily bestsellers
   • 7 Days: Weekly momentum
   • 30 Days: Monthly trends
   • Custom: Define your own window
3. Manual Trend Boosting
   product_id,boost_factor,start_date,end_date
   PROD123,2.0,2024-01-15,2024-01-22
   PROD456,1.5,2024-01-15,2024-01-31
   

​

Glood’s Trend Calculation Settings

// Configure in Glood Dashboard
{
  "trend_config": {
    "algorithm": "velocity_acceleration",
    "signals": {
      "views": 0.3,
      "cart_adds": 0.25,
      "purchases": 0.35,
      "searches": 0.1
    },
    "filters": {
      "min_inventory": 5,
      "exclude_categories": [],
      "price_range": [0, 99999],
      "new_product_boost": true
    },
    "display": {
      "show_trend_badge": true,
      "show_velocity": true,
      "highlight_new_trends": true
    }
  }
}

​

Glood-Specific Trend Features

​

Trend Filtering Controls

1. Category-Specific Trends
   • Show trends only from specific collections
   • Cross-category trending
   • Department-level trends
2. Price Band Trends
   • Budget trends (under $50)
   • Mid-range trends ($50-200)
   • Premium trends ($200+)
3. Geographic Trends
   • Regional trending products
   • Country-specific trends
   • Local store pickup trends

​

Manual Trend Management

​

Pin Trending Products

• Force products to appear in trending
• Set custom trend scores
• Schedule trend appearances
• Create marketing-driven trends

​

Exclude from Trending

• Block specific products
• Hide entire categories
• Exclude low-margin items
• Remove problematic products

​

Trend Analytics in Glood

• Trend Velocity: Speed of popularity growth
• Peak Prediction: When trend will max out
• Conversion Rate: Sales from trending section
• Discovery Rate: New vs returning viewers
• Revenue Impact: Direct attribution

​

Glood Trend Windows Explained

​

1-Hour Window (Flash Trends)

Use Case: Viral products, social media spikes
Update: Every 5 minutes
Best For: Fast-fashion, limited drops
Example: "300% surge in last hour!"

​

24-Hour Window (Daily Hot)

Use Case: Today's bestsellers
Update: Every hour
Best For: General retail
Example: "Today's most popular"

​

7-Day Window (Weekly Trends)

Use Case: Sustained popularity
Update: Every 6 hours
Best For: Considered purchases
Example: "This week's favorites"

​

30-Day Window (Monthly Leaders)

Use Case: Seasonal trends
Update: Daily
Best For: Seasonal products
Example: "Month's top picks"

​

Glood Trend Customization

​

Visual Badges & Labels

• 🔥 “Hot” badge for velocity > 200%
• 🆕 “New Trend” for first 24 hours
• 📈 “Rising Fast” for acceleration
• ⏰ “Limited Time” for flash trends
• 🌟 Custom badges per product

​

Trend Notification System

Alert Settings:
  New Trend Emerges: Email + Slack
  Trend Peaks: Dashboard notification
  Inventory Low: SMS alert
  Viral Detection: Immediate alert

​

Success Stories with Glood Trending

​

Fashion Retailer

• Setup: 24-hour window with social signals
• Result: 156% increase in discovery
• Key: Instagram integration for trend detection

​

Electronics Store

• Setup: 7-day window for new releases
• Result: 89% of new products viewed
• Key: New product boost enabled

​

Home Goods

• Setup: Seasonal 30-day window
• Result: 45% higher seasonal sales
• Key: Manual boost for seasonal items

​

Implementation Best Practices

​

Quick Setup with Glood

1. Install & Configure (5 minutes)
   • Add Glood app to Shopify
   • Create Trending section
   • Select time window
   • Choose display location
2. Optimize Settings
   • Start with 24-hour window
   • Enable all signals (views, carts, sales)
   • Set minimum threshold (10+ interactions)
   • Add visual badges
3. Advanced Strategies
   • Create multiple trending sections
   • Different windows for different pages
   • Category-specific trending
   • Personalized trending per segment

​

Common Configurations

​

Homepage Trending

Window: 24 hours
Products: 8
Layout: Carousel
Badges: Enabled
Update: Hourly

​

Category Page Trending

Window: 7 days
Products: 4
Layout: Grid
Filter: Same category only
Update: Every 6 hours

​

Mobile Trending

Window: 1 hour
Products: 6
Layout: Vertical scroll
Badges: Minimal
Update: Every 10 minutes

​

Monitoring & Optimization

​

Glood Analytics Dashboard

1. Engagement Rate: CTR on trending sections
2. Conversion Impact: Sales from trending
3. Trend Accuracy: Predicted vs actual
4. Revenue Attribution: Direct revenue impact
5. Customer Segments: Who engages with trends

​

A/B Testing Options

• Test different time windows
• Compare AI vs manual curation
• Badge styles and messaging
• Number of products shown
• Update frequencies