Reviews are the richest public signal in retail: they encode quality, sentiment, and demand in the shopper's own words. But the value only emerges at scale - across products, categories, and retailers - which means product review data scraping of publicly visible ratings and reviews.
This report uses web scraping of public review data across Amazon, Walmart, and Target to measure rating distributions, review velocity, and sentiment patterns - the context any team relying on review data extraction for product or sentiment intelligence needs. It uses only public review content, never reviewer personal data.
Key findings at a glance
Three patterns stand out across the review data. (Figures are illustrative previews - the full report breaks them down by category and retailer.)
Key finding 1: ratings cluster high but category matters
Most listings sit above four stars, so raw averages barely separate products. The signal is in the distribution and the outliers, not the mean.
This is why product review data scraping must capture the full rating distribution and review count, not just the headline star - a 4.2 from 50 reviews means something very different from a 4.2 from 50,000.
Key finding 2: review velocity is a demand signal
How fast a product accumulates reviews is a leading indicator of demand. The sample shows velocity tiers (illustrative).
| Velocity tier | Reviews/week | Signal |
|---|---|---|
| Breakout | 50+ | Rising demand |
| Steady | 5-50 | Established |
| Stale | <5 | Fading / niche |
Tracking review velocity over time - only possible with repeated review data scraping - turns a static rating into a demand trend.
Key finding 3: sentiment lives in the text, not the star
The star is a blunt instrument; the review text carries the actionable detail - what breaks, what delights, what shoppers compare against. Structured product review data scraping that captures review text, rating, verified-purchase flag, and date lets teams mine sentiment and themes at scale. Reviewer identities are never needed - only the public content and metadata.
What the underlying data looks like
The report is built from public review records like the one below - the structure buyers receive in a sample.
{
"retailer": "Amazon",
"product_id": "B0EXAMPLE12",
"product_name": "Brand C Blender 700W",
"category": "home",
"rating": 4,
"review_title": "Great value, a bit loud",
"review_text": "Works well for smoothies, motor is noisy...",
"verified_purchase": true,
"review_date": "2026-06-20",
"captured_at": "2026-06-29T12:00:00Z"
}
Aggregated to a product-and-category view, the data rolls up into a flat file analysts can model on:
retailer,category,avg_rating,review_count,reviews_per_week,pct_negative
Amazon,home,4.4,2140,32,12
Walmart,apparel,3.9,860,11,24
Target,grocery,4.3,540,7,9
Who this report is for
This report is built for the teams that mine reviews for product and sentiment intelligence via product review data scraping.
- Rating distributions by category & retailer
- Review velocity as a demand signal
- Sentiment & theme extraction approach
- Verified-purchase and recency effects
- Complete methodology, sample size and sources
Methodology & data
The findings are based on product review data scraping of publicly visible ratings and reviews across Amazon, Walmart and Target in 2026, capturing rating, review text, verified-purchase flag and date, aggregated by product, category and retailer. It uses only public review content - never reviewer personal data. The full report details the categories, method and how each metric is calculated.
The numbers and charts shown on this page are illustrative previews of the kind of analysis in the report. They are based on publicly available, non-personal web data in aggregate and do not represent any single named company. The full report contains the complete dataset, methodology and sources.
Frequently asked questions
Yes. Enter your details and we will email you the PDF.
No. It uses only publicly visible review content and metadata - never reviewer names, emails or personal data.
They are illustrative previews of the report's analysis. The full PDF contains the complete dataset, methodology and sources.
Yes. We deliver review datasets - rating, text, date, verified flag - for your products and competitors via API or file.
Configurable - reviews can be captured on a recurring schedule so velocity and sentiment trends stay current.