Churning out fake reviews on Amazon has become an industry. Here’s how to spot one.
Described as an "organized fake reviews scam," the server contained over 13 million records (about 7 GB of data), potentially involving more than 200,000 people in "unethical activities," the SafetyDetectives report said.
"The information found on the…server outlines a common procedure by which Amazon vendors procure ‘fake reviews’ for their products," the report said, adding that this underscores the "scale and impact" of fake scams on Amazon.
With this kind of activity potentially happening on Amazon, here are some guidelines for spotting fake reviews:
--Recent reviews: Before you read the reviews, be sure to sort them by the most recent, according to Saoud Khalifah, the CEO of Fakespot, which provides an AI-based app for detecting fake reviews. "Amazon will show you the ‘Top reviews’ but sometimes they are very old or even worse, they came from Amazon's Vine Program which is very biased," Khalifah said.
--Unknown brands: If you come across a brand that you’ve never heard of, be cautious. Especially if you see too many 5 star reviews, Khalifah says. In that case, look elsewhere.
--"Naturally-occurring" reviews -- or legitimate reviews -- tend to have a diverse sample of reviewers, according to ReviewMeta which provides an app to filter out fake reviews. Naturally-occurring reviews have "some newer reviewers and some experienced reviewers. You’d see short reviews, some long reviews and some reviews in the middle," according to ReviewMeta. Lots of very brief reviews can be a red flag.
--Timing: Product reviews should appear over a long period of time, says ReviewMeta. "You’d expect them to appear evenly over the life of the product, and not just on a few different days."
--Suspicious language: "Fake reviews often use less emotional language," according to SafetyDetectives. "A fake review may even read like an advert, badmouthing the product’s competitors in the process."
--Generic statements: Several five-star reviews may highlight the same points, "or the reviews could generally lack variance – not revealing anything about each individual’s specific experience," according to SafetyDetectives. Fake reviews may also contain many generic keywords.
Irrelevant information: A common practice for suspicious vendors is "review merging" says SafetyDetectives. In this case, a vendor may republish reviews from other products for their own product.
Check the reviewer’s account: If they have left, for example, a lot of positive reviews on a vendor's products, they could be fake, SafetyDetectives said. And if their account lacks personal information, that could be a sign of a fake reviewer.
Earlier this month Amazon, in a blog post, said that it is working to combat fake reviews.
"To help earn the trust of customers, we devote significant resources to preventing fake or incentivized reviews from appearing in our store," Amazon said.
"Over many years, we have continued to invent, deploy, and continually improve sophisticated technology powered by machine learning and couple that with expert human investigators to proactively prevent fake reviews from ever being seen in our store."
"We have clear policies for both reviewers and selling partners that prohibit abuse of our community features, and we suspend, ban, and take legal action against those who violate these policies," Amazon told Fox Business earlier this year.