SensaBeauty Price Tracker - APIFY Actor

Automated competitor price tracking for SensaBeauty products.

Features

• 3 Competitor Sites: Jomashop, FragranceNet, Maxaroma
• Smart Filtering: Automatically scrapes 998 fragrance products (excludes skincare)
• Intelligent Matching: Fuzzy search with SKU prioritization
• Real-time Analytics: Automatic price statistics calculation
• Supabase Integration: Direct database updates with confidence tracking
• Match Quality Scores: Confidence scoring (0.00 - 1.00)
• Statistics Logging: Scrape run performance metrics and analytics

Deployment

Prerequisites

• APIFY account
• APIFY CLI installed: npm install -g apify-cli
• Supabase credentials

Deploy to APIFY

# Install dependencies
npm install

# Test locally
npm test

# Login to APIFY
apify login --token apify_api_1Z5ffM5PqQSRz2LOl5SWbdbsPvbkkg35ssUx

# Deploy
apify push

Environment Variables

Set these in APIFY Console:

• SUPABASE_URL = https://iuyxjwgklxhxqtqjwvrq.supabase.co
• SUPABASE_SERVICE_KEY = [your service role key]

Usage

Input Schema

{
  "mode": "full",           // "full", "sample", or "specific"
  "productIds": [],         // Array of product IDs (for "specific" mode)
  "competitors": [          // Which sites to scrape
    "jomashop",
    "fragrancenet",
    "maxaroma"
  ],
  "maxConcurrency": 3,      // Number of parallel pages
  "updateDatabase": true    // Whether to update Supabase
}

Run via API

curl -X POST "https://api.apify.com/v2/acts/YOUR_ACTOR_ID/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"mode":"sample"}'

Run via GitHub Actions

The actor is automatically triggered daily at 2 AM UTC by the GitHub Actions workflow in .github/workflows/price-sync.yml.

Output

Data Schema

Each product scrape outputs structured JSON:

{
  "productId": "uuid",              // Product UUID from Supabase
  "productName": "Product Name",    // SensaBeauty product name
  "productSku": "SKU123",          // Product SKU/UPC code
  "competitor": "jomashop.com",    // Competitor domain
  "competitorName": "Jomashop",    // Human-readable name
  "price": 89.99,                  // Current price (USD)
  "originalPrice": 89.99,          // MSRP price
  "currency": "USD",               // Currency code
  "availability": "in_stock",      // Stock status
  "url": "https://...",            // Direct product URL
  "matchedTitle": "...",           // Matched product title
  "matchConfidence": 0.85,         // Match score (0.00-1.00)
  "scrapedAt": "2025-10-24T...",   // ISO 8601 timestamp
  "status": "success"              // success|not_found|error
}

Storage Destinations

1. APIFY Dataset: All scraped price data (always saved)
2. Supabase Tables (if updateDatabase: true):
   • competitor_prices: Individual price records
   • product_price_analytics: Calculated statistics
   • scrape_runs: Run statistics and performance metrics

Run Summary

Available via Actor.getValue('OUTPUT'):

{
  "runId": "...",
  "mode": "full",
  "competitors": ["jomashop", "fragrancenet", "maxaroma"],
  "productsAttempted": 100,
  "productsMatched": 85,
  "matchRate": 85.00,
  "databaseUpdated": true,
  "completedAt": "2025-10-24T...",
  "status": "completed"
}

Product Filtering

The actor automatically filters products to only scrape fragrances:

• ✅ 998 fragrance products (category = "all")
• ❌ Excludes 2 skincare products (Serums, Cleansers)

This filtering ensures:

• Higher match rates (fragrances are sold on competitor sites)
• Faster run times (no wasted requests on skincare)
• More accurate pricing data

Performance

• Products: 998 fragrances (automatically filtered)
• Estimated Match Rate: 40-60% (niche fragrances have lower availability)
• Duration: ~20-25 hours for full catalog (998 × 3 competitors)
• Concurrency: 3 pages (adjustable, recommended: 3)
• Rate Limiting: Auto-adjusts based on CPU and response codes

Monitoring

Check scrape quality:

SELECT * FROM scrape_runs ORDER BY completed_at DESC LIMIT 5;

View price updates:

SELECT COUNT(*) FROM competitor_prices
WHERE scraped_at > NOW() - INTERVAL '24 hours';

Support

For issues or questions, contact support or check the integration documentation.