Did you know, Amazon removed over 200 million fake reviews in 2024 alone? In this 7-minute episode, you will get a detailed technical explanation of exactly how Amazon's AI detects fake reviews and why so many reviews from review-swapping services were removed. We're doing a deep dive into the sophisticated AI systems Amazon is using to hunt down fake review networks in 2025. If you’ve ever been tempted by "review swap" groups on Facebook or services like Pubby, you need to hear this before you hit "publish." Amazon isn't just looking at the text of your reviews anymore – they are looking at the structure of your network. What We're Covering: The Heavy Penalties: From content removal and losing community features to the permanent termination of your entire Amazon account. The "Paper Trail": How Amazon links authors and reviewers through shared IP addresses, social media links, and even your personal gift-giving history on your address list. "Review Gating" Exposed: Why filtering for only positive reviews before sending readers to Amazon is a major violation that the AI is now trained to catch. The Science of the Snitch: A breakdown of how Amazon uses "Graph Theory" (Clustering Coefficients and Eigenvector Centrality) to identify "tight-knit" neighborhoods of suspicious activity. Unsupervised Clustering: How Amazon's AI can find these fraudulent groups automatically without even knowing which products to look at first. The Reality Check: Amazon's system is now so effective that just two small clusters of data accounted for 70% of all identified fake review activity in a recent study. Stop playing games with the algorithm. The "fingerprint" left behind by review services is too expensive and difficult to hide. We’re discussing how to build a sustainable, long-term publishing business without risking a lifetime ban. And the most critical question: How to obtain reviews under these conditions? We will answer it in January 2026, and you will definitely like the answer.

We had a great, deep conversation yesterday. @Christian Kursch asked very important question about getting out of authors' bubble to get reviews. There are answers ready, but we will announce it next week. The full video recording for Premium members is available in the Classroom. And those are my takeaways from the meeting: Amazon’s Aggressive Review Sweeps - Historical Context: While Amazon typically increases scrutiny during the holiday season, the recent "swipe" was exceptionally aggressive, with some publishers reporting 30% to 50% of reviews deleted. - Algorithmic Targeting: Amazon's systems have become almost entirely algorithmic, searching for "outliers" from normal human patterns rather than specific individuals. - The Virtual Assistant (VA) Problem: A primary cause for recent account terminations is the use of VAs from platforms like Upwork or Fiverr who post fake reviews using flagged or low-activity accounts. - Risk Indicators: A key indicator of a flagged account is the "pending approval" time. While the standard is three days, accounts under scrutiny will see longer wait times. Adriano suggests never having more than two reviews pending approval simultaneously. Best Practices for Review Safety - Prioritize Verified Paid Purchases: Verified purchases (at price points like $0.99 or $2.99) are the safest way to show Amazon a real transaction has occurred. - Avoid Unverified and KU Reviews: Standard unverified reviews and Kindle Unlimited (KU) reviews are currently heavily targeted. - KU Limitations: Kindle Unlimited reviews often do not show as "verified" unless the book is read on an official Amazon device or app; reading via the Kindle Cloud Reader (web browser) will not trigger verified status. - Natural Review Timing: Reviews should ideally be posted between 5 and 7 days after purchase to appear natural.

what was the name of the AI used (Igor mentioned that) that can be used to have the manuscript read too you by the AI?

I've been skimming discussions across several KDP groups this week. Here's a short snapshot of what keeps coming up. 1. "Why did this book stall?" Many posts are from authors whose books launched well, then flattened out. The common thread is not ads or reviews, but weak topic depth, thin interiors, or unclear audience fit. People are slowly realizing that a clean launch cannot fix a shallow product. 2. Low-effort books are quietly dying There is growing frustration from people who followed older playbooks: low content, fast publishing, and generic covers. These books are not banned or penalized. They just stop selling. The mood is less panicked, more accepting of the fact that the bar moved. 3. Fewer tactics, more questions about catalogs Instead of "How do I launch this book?", more people ask: - How many books do I need before things feel stable? - Does niche focus still matter after 10–20 titles? - Is it better to go deeper or wider now? This is a shift toward long-term thinking. 4. Ads confusion without panic Ads are discussed a lot, but the tone has changed. Less "this trick prints money," more "I’m spending, but I don't fully understand what’s working." Many admit they run ads before they understand their own books. 5. Quiet interest in boring niches Several threads circle around planners, logs, workbooks, and reference-style books. People are noticing that unsexy ideas often behave better over time than trend-based projects. 6. Burnout signals Not dramatic exits, but comments like: - "I published 12 books, and I'm tired." - "I'm not sure what to publish next." - "I feel busy but not effective." This shows a need for systems, not motivation. Open question for the Guild members: Which of these feels closest to where you are right now, and why? Reply with a number (1–6) or add what you are seeing that's missing here.

Let's take a break from tactics. Most long-term royalties often don't come from exciting ideas. Sometimes they come from books that feel almost… dull. Question: What is your most boring book that still makes money? Reply with: - Topic or niche (not exact) - Why you thought it would flop - What it does instead (steady, seasonal, slow growth, etc.) - One lesson it taught you Examples welcome. Bragging not required, but welcomed. If you have no boring earners yet, answer this instead: Which book surprised you by not working, even though you thought it was solid?