[ BDSN ] How fake Amazon reviews actually work

💰 INTRODUCING the BDSS ORACLE

When you turn to ChatGPT for help with your Amazon or ecom business, you'll often get useful information. But here's the problem: it can't distinguish between legitimate experts and self-proclaimed "gurus" spreading misinformation on YouTube or blog posts.

ChatGPT treats everything it reads as equally valid, even when it's flat-out wrong.

Here's a real example: Ask ChatGPT about Amazon's "A10 algorithm," and it'll tell you it's the successor to the A9 algorithm. That's completely false. Some people use "A10" as clickbait, but there's no such thing.

Ask the same question to the new BDSS Oracle, and you'll get the truth: "There is no official A10 algorithm for Amazon."

What makes the BDSS Oracle different?

The BDSS Oracle uses content exclusively from Kevin’s curated Amazon and ecom experts, not random internet sources.

It pulls answers only from:

• Reputable podcasts for ecom sellers, including Kevin & Norm’s Marketing Misfits

• The BDSN Newsletter back to the original

• Kevin's webinars and courses

• The BDSS WhatsApp Group chats

• BDSS events, like BDSS Iceland and the recent BDSS 12 Virtual.

• Presentations inside BDSC, like Dan Kurt’s AI ranking hacks

It's completely walled off from all the noise and misinformation flooding the internet, giving you access to exclusive tactics and strategies you won't find anywhere else.

Normally, the BDSS Oracle is restricted to members of the Billion Dollar Sellers Club (BDSC). But for a limited time, Kevin is opening it to the public so you can experience the difference firsthand.

Try the BDSS Oracle today and reply to this email to let me know what you think!

🕹️ AMAZON’S NEW PARTIAL REFUND: YOUR Q4 GAME PLAN

Last week, on October 1, Amazon rolled out expanded FBA partial refunds without returns, letting you refund customers while they keep the item.

The Q4 upside:

• Lower costs: Holiday returns are expensive. Eliminating return shipping and processing protects your bottom line.

• Less operational strain: Your team is already stretched thin. Fewer return shipments = fewer hours burned.

• Faster resolution: Instant refunds keep customers happy during the busiest shopping season.

• Revenue protection: Quick, frictionless fixes can salvage customer relationships and drive repeat orders post-holidays.

Sounds promising. But there's a catch.

The risks you can't ignore:

• Refund abuse: Some buyers will exploit this, claim issues, keep products, pocket refunds.

• Death by a thousand cuts: Even $5–$10 partial refunds add up fast across thousands of orders. Your logistics savings could evaporate.

• Blind spots in quality control: No returned items = no physical evidence of defects. You're flying blind on product issues.

• Shifting expectations: Once customers learn they can keep items and get refunded, there's no going back. Removing this option later risks backlash.

How to use partial refunds without getting burned:

1. Test on low-value, high-volume items first
Start with products under $20 where return shipping exceeds item value. The economics work here.

2. Protect your premium SKUs
Don't enable this for high-value items or categories prone to abuse—electronics, supplements, premium beauty. Brand reputation matters.

3. Cap your refund amounts
Set partial refunds at 15–20% of order value. High enough to resolve complaints, low enough to discourage gaming.

4. Track repeat offenders
Use Amazon's return reports and your CRM to flag serial refund seekers. Patterns emerge quickly.

5. Guard your metrics
Partial refunds still count as returns. Monitor SKU-level return rates religiously. You don't want the "frequently returned" label tanking your listings.

6. Reassess in January
After Q4 wraps, run the numbers. Did partial refunds save money or bleed margin? Keep what works, kill what doesn't.

🙊 INSIDE the WORLD of FAKE AMAZON REVIEWS

An Ecom Crew investigation went undercover into Amazon's fake review networks.

What they found was a sophisticated operation that Amazon's lawsuits haven't stopped.

How It Actually Works

Facebook is the marketplace. Dozens of region-specific groups (US, Canada, UK, password-protected VIP groups) connect sellers with reviewers. The process: customer buys product at full price, leaves review, gets PayPal refund, keeps free product. The investigator tested it, got a free webcam and confirmed the system works perfectly.

Professional operations charge $5 per review plus product cost. They primarily target Chinese sellers and require WeChat video verification. One seller provided a tracking dashboard showing real-time order status, customer IDs, and review posting (delayed 7 days to look organic).

The Amazon Lawsuit That Changed Nothing

The investigator discovered Amazon had "shut down" the exact website they bought from, except it hadn't. Court records showed defendants simply never responded to the lawsuit. Amazon won by default, issued a victory press release, while the website kept operating and taking orders.

The Scale

The upsell revealed everything: 100 five-star reviews in 30-40 days. Competitor sabotage with one-star reviews? Also available. The typical reviewer isn't a criminal, she's a local teacher and mother leaving dozens of overly enthusiastic reviews for free products, unknowingly violating FTC rules.

Amazon says they stopped millions of fake reviews in 2024 alone. That's not a solution—that's the scope of the problem.

Your competitors know about these systems. Some are using them. Meanwhile, buyer distrust affects everyone's conversion rates.

Watch the full Ecom Crew investigation to see the WeChat call footage, live dashboard screenshots, and reviewer profiles. This isn't happening in the shadows, it's happening on your listings right now.

🗜️ OPEN AI’S SORA 2: The MISSING COMMERCE PIECE

In a recent edition of her AI for E-commerce newsletter, Jo Lambadjieva breaks down why OpenAI's Sora 2 launch, which came just days after ChatGPT added native checkout, is actually a brilliant commerce play disguised as a TikTok clone.

The Two Modes of Shopping

Jo's key insight: We shop in fundamentally different modes depending on what we're buying.

• Research Mode (ChatGPT): When you need to compare laptop specs, read reviews, and make considered purchases

• Vibe Mode (Sora): When you impulse-buy that ridiculous cat sweater at midnight because it just felt right

ChatGPT dominates Research Mode but completely fails at Vibe Mode—those visual, emotional, "I didn't know I needed this" moments that drive billions in sales on TikTok Shop.

Closing the Loop

Now OpenAI has both pieces: the analytical research assistant AND the entertainment-driven discovery engine. More importantly, they control the entire journey: discovery happens in Sora's feed, and transactions happen through ChatGPT's checkout. It's a closed ecosystem that guides every type of purchase, from considered to impulsive.

For e-commerce sellers, Jo warns: static product photos are about to feel as outdated as GeoCities websites. The future demands dynamic, AI-generated content that works across this new discovery-to-purchase ecosystem.

📛  MASSIVE KEEPA UPDATE JUST DROPPED

Saul (@Saul) on X:

Amazon Sellers, massive Keepa update just dropped 🚨

Keepa quietly rolled out some new features, and this one's a game-changer. It's like storefront stalking on steroids.

You can now view the ENTIRE historical product catalog of any Amazon seller using just their seller ID.

Why does this matter?

When you're sourcing replenishments, seasonal items, or arbitrage products, you'll notice the same sellers popping up on your listings.

Before, you could only check their current inventory. Now? You can dig into their out-of-stock items - the ones they'll be restocking soon.

Found a seller hitting all your same retail stores? Check what they crushed with last holiday season and track those products down for yourself this year.

Doing wholesale? You just got handed your competitors' entire catalog to reverse-engineer their distributor relationships.

Want to test it? Go to Keepa > Data > Product Finder and look for the section shown in the screenshot above.

✌🏼 See you again Thursday …

The answer to today’s STUMP BEZOS is
95% say they get no return from investing in AI