DEV Community

Cover image for I Built a Chrome Extension to Detect Fake Amazon Reviews
Dilan Weerasinghe
Dilan Weerasinghe

Posted on

I Built a Chrome Extension to Detect Fake Amazon Reviews

As developers, we spend a lot of time reviewing tools, libraries, and products online, and Amazon reviews are often the first thing we check.

But lately, I noticed something frustrating:

⭐ Products with thousands of reviews but poor real quality
🧠 Reviews that look human, yet feel suspicious
🤖 Repeated phrases, patterns, and timing that don’t add up

So I built a Chrome extension to answer one simple question:

“Can we algorithmically judge how trustworthy Amazon reviews really are — without sending data to any server?”

That project became TrustRadar.
❓ The Problem With Amazon Reviews

Amazon reviews are valuable — but they’re also easy to manipulate:

  • Paid review farms
  • Incentivized “verified” reviews
  • Burst patterns (hundreds of reviews in short time windows)
  • Generic praise with no real product usage

For a human, spotting this across hundreds or thousands of reviews is exhausting.

For a program? Much easier.

🛠️ What TrustRadar Does

TrustRadar analyzes reviews directly inside your browser and gives you:

✅ Trust Score (0–100)
⚠️ Risk Flags (patterns commonly linked to fake reviews)
📊 Review distribution analysis
🧠 Buy / Consider / Not recommend

🔍** How the Analysis Works (High-Level)**

Without going into proprietary details, TrustRadar looks at:

  1. Review Authenticity Signals
  • Review length vs substance
  • Generic wording frequency
  • Repeated phrase patterns
  • Review timing clusters
  1. Reviewer Credibility
  • Verified purchase ratio
  • Reviewer activity patterns
  • Media-backed reviews (images/videos)
  1. Behavioral Patterns
  • Sudden review spikes
  • Rating distribution anomalies
  • Helpful-vote consistency

Each signal contributes to a weighted trust score, not a binary “fake / real” decision.

TrustRadar works entirely offline after installation.

Your Amazon page → your browser → your decision.

🚀 Current Status

✅ Approved on Chrome Web Store
✅ Live and usable
🚧 Actively improving detection logic
🚧 UI/UX refinements in progress

This is still an evolving side project, and I’m actively listening to feedback from both developers and regular shoppers.

🤝 What I’d Love Feedback On

If you’re a dev, I’d especially love thoughts on:
Signal weighting ideas
Edge cases you’ve seen with reviews
UI clarity (does the score feel trustworthy?)

False positives vs false negatives

🔗 Try It & Share Feedback

If you want to try it:
(https://chromewebstore.google.com/detail/hgnlhnjkenedenmhngkfdnfclmmocock?utm_source=item-share-cb)

Search TrustRadar on the Chrome Web Store

Or test it on any Amazon product page

If you’re building something similar or thinking about local-first browser tools, I’d love to connect.

Top comments (0)