FragranceX Reviews Scraper

Overview

This project extracts customer review data and rating analytics for products listed on FragranceX using publicly available review data sources.
Given a FragranceX product URL, the tool returns structured, analytics-ready review data enriched with product-level rating metrics.

The output is designed for direct use in analytics, reporting, dashboards, and data pipelines.

---

Key Capabilities

• Accepts a FragranceX product URL as input
• Collects customer reviews with associated ratings and metadata
• Returns one row per review with embedded product-level rating statistics
• Produces clean, normalized, JSON-serializable output
• Ready for export to CSV, Parquet, BigQuery, Snowflake, or BI tools

---

Input

Example input payload:

{
  "URL": "https://www.fragrancex.com/products/dolce-and-gabbana/light-blue-perfume#power-reviews-section",
  "maxreviews": 50
}

Fields

• URL: FragranceX product page URL
• maxreviews: Maximum number of reviews to return

---

Output

Example output record (one review):

{
  "1_star_count": 70,
  "2_star_count": 28,
  "3_star_count": 73,
  "4_star_count": 252,
  "5_star_count": 1609,
  "rating_count": 2032,
  "average_rating": 4.63,
  "recommended_ratio": 0.93,
  "review_id": 572327682,
  "comments": "This is one that cost more than the other perfume I use most of the time",
  "headline": "I would buy this again",
  "nickname": "Tj",
  "property_values": [
    "Great Smell",
    "Long-Lasting",
    "Anytime",
    "Daily Use",
    "Natural Style"
  ],
  "location": "Heyburn, Idaho",
  "created_date": "2026-01-09",
  "product_page_id": "884W",
  "is_staff_reviewer": false,
  "is_verified_buyer": false,
  "is_verified_reviewer": false,
  "helpful_votes": 0,
  "not_helpful_votes": 0,
  "rating": 5,
  "helpful_score": 1431
}

Each run returns a list of such review objects.

---

Data Included

Product-Level Metrics

• Star distribution (1–5 stars)
• Total rating count
• Average rating
• Recommendation ratio

Review-Level Fields

• Review ID
• Rating
• Review headline and comments
• Reviewer nickname and location
• Review date (YYYY-MM-DD)
• Product page ID
• Review tags / attributes (flattened)

Engagement & Trust Signals

• Helpful and not-helpful votes
• Helpful score
• Verified buyer / reviewer flags
• Staff reviewer indicator

---

Types of Analysis You Can Perform

1. Rating & Quality Analysis

• Average rating trends
• Star distribution comparison across products
• Identification of rating polarization

2. Sentiment & Text Analysis

• Sentiment scoring of review comments
• Keyword and phrase extraction
• Feature-level feedback analysis using review tags

3. Customer Experience Insights

• Common positive and negative themes
• Usage patterns (daily use, occasion-based, longevity, etc.)
• Regional differences in feedback

4. Product Performance Tracking

• Review volume over time
• Changes in rating distribution
• Impact of new launches or promotions

5. Competitive Benchmarking

• Compare ratings and sentiment across brands
• Identify strengths and weaknesses relative to competitors

---

Business Value

• Improved Product Decisions: Understand what customers like or dislike at a granular level
• Marketing Optimization: Leverage positive themes and customer language in campaigns
• Reputation Management: Monitor review quality and emerging issues early
• Merchandising Strategy: Identify high-performing products and attributes
• Data-Driven Roadmaps: Support product and pricing decisions with real customer feedback

---

Output Format

• Python list of dictionaries
• Fully JSON-serializable
• Easy integration with analytics, ML, and BI workflows

---