As much as 90% of people read online reviews before making a purchase. And yet, research has shown that 30–40% of these are not genuine. For customer review platforms and online marketplaces, this is a major problem.
In 2022, Amazon reported more than 23,000 groups facilitating fake reviews, with over 46 million members and followers. They proactively blocked over 200 million suspected fake reviews from their stores in the same year. And, have taken legal action against 147 fraudulent review brokers across China, Europe and America.
And it’s not an issue exclusive to Amazon. Tripadvisor identified 1.3 million fake reviews on its platform in 2022. Additionally, a research team by ‘Which?’ found Facebook groups trading fake reviews for the likes of Google and Trustpilot with more than 62,000 members.
These bad actors distort the market and mislead consumers. So, how can platforms find fake reviews and stop them from damaging the integrity of their brand?
Sometimes, businesses buy or incentivize reviews to boost their visibility and ratings around the launch of a new brand, product or service. Equally, they may pay for fake reviews to falsify their rating after they’ve had a period of negative feedback from genuine customers who haven’t been happy with their purchases.
Often, these companies are looking for a quick fix. So you’ll see a spike in activity that’s much higher than their usual ratings frequency.
A user joined your platform less than a week ago but has left 500+ reviews. Either, they’re incredibly passionate about rating their favorite businesses (to the point where it takes up most of their day) or it’s a fake account.
Often, platforms sweep their user-base for suspicious activity and suspend thousands of accounts at a time. A brand new account with an abnormal number of reviews could be a sign of a repeat offender.
In late 2023, there was a scandal in the publishing world. Author, Cait Corrain was dropped from a lucrative publishing deal after creating multiple fake accounts on Goodreads to prop up the ratings for their own book, whilst down-rating titles from other authors in similar genres.
It was spotted by Canadian author and Goodreads user, Xiran Jay Zhao, who collated 13 pages of evidence and posted a video on TikTok with over 700,000 views.
This type of “review bombing” is unfortunately not uncommon.
Sometimes review fraudsters aren’t so subtle. They’re briefed by their clients to not only leave negative reviews on their competitors’ accounts, but to recommend their product or service as a better alternative.
In some instances, this could be a recommendation from a genuine customer. But if the same brand is mentioned in the comment sections of multiple competitors on multiple occasions, it’s a major red flag.
A lot of people review the hotels and restaurants they visit when they go on vacation, and that’s completely normal.
But, when an account is based in Lisbon, and they’ve reviewed a dentist in Melbourne, a repair shop in Brownsville, Texas, and a hairdresser in Hong Kong, all in the space of a month—it’s very likely those reviews have been bought and paid for.
In other cases, the opposite can happen. You might see a spike in reviews coming from the location where the business in question is based. This could be a sign that they’ve asked their friends, family and employees to flood their page with positive reviews and increase their rating.
If a barrage of negative reviews shows up on a company profile, and they’re all from the same location, it could mean a competitor has asked their network to down-rate the other business.
When a business briefs a review broker, they’ll often send details on the type of reviews they want to receive in return.
For example:
If there’s a flurry of reviews all with similar sentiment, they could be fake.
Not all fake reviews come from fake reviewers. Sometimes, you can have a fake review from a real customer. This is often the business’s fault. They ask the customer to leave a positive review to receive a free product or future discount with their purchase. The customer then thanks the seller for their generous gift in the content of their review.
Any business with any mention of gifts or discounts on their page should be investigated further.
These days, it takes a lot more to spot a fake review. AI technology is advancing at a rapid pace and making it all the more difficult to separate legitimate reviews from the fake ones.
In many guides on ‘how to tell if online reviews are fake’ consumers are advised to watch out for simple language, spelling mistakes and poor translations. But today, with tools like ChatGPT, all it takes is a few prompts for AI to deliver credible-sounding reviews in seconds. Fraudsters can work faster than ever before.
And it doesn’t stop there. Typically, reviews with photos attached are seen as more trustworthy. But as platforms like Midjourney and Dall·E become more advanced and produce hyper-realistic images, it’s even easier for review brokers to look legitimate.
They use the same technology to create fake profiles. In the past, a platform could spot fake reviewers if they used the same stock photo as their profile image across multiple accounts. Now, they can simply visit a website like thispersondoesnotexist.com to download images that have been generated from thousands of faces across the internet—giving them unique, incredibly convincing, avatars for every fake profile they make.
As fake reviewer’s tactics become more sophisticated with AI, so too must the tools that combat them. You need to analyze the content and behavioral patterns to weed out suspicious activity.
Pasabi combines AI, ML and behavioral analytics to continually analyze all your user profiles and reviews, detecting suspicious activity with great precision—giving you a fully-automated fake review spotter in one solution.
Our graph technology analyzes thousands of data points to draw links between businesses, individuals and groups of individuals that are working together to mislead your users. This could be a series of bots (automated systems using fake accounts) or people that are connected to the business in some way.
It answers questions like: Where did the review come from? Who posted it? Who does it benefit? When was it posted? Who is the reviewer connected to?
Our technology then examines multiple behavioral risk signals in your data, such as:
→ Learn more about our fake review detection solution.
Fake reviews chip away at consumer trust. And over time, this diminishes the reputation of your brand. When the story broke about author, Cait Corrain, people started questioning the credibility of the Goodreads platform. “The sense of community has been shattered” said one author whose work was targeted. And an article in The Guardian voiced concerns about its reliability. All from one fake review scandal that went viral.
Fake reviews can also cause serious harm to innocent businesses. In 2022, restaurants all over America were being targeted by a fake review scam. Out of nowhere, they’d receive numerous one-star reviews on Google, then days later, someone would email them demanding money to remove the negative reviews.
For not-so-innocent businesses, the consequences are high. E-commerce brand Fashion Nova was fined $4.2 million under FTC rules that charge guilty parties $50,000 for each fake review, every time a consumer sees it. Across the EU and UK, similar fines are being issued under recent legislation. Part of this legislation puts more responsibility on review platforms and marketplaces.
—
Fake review brokers are becoming more and more advanced, posing a significant threat to consumers, businesses and platforms like yours. Pasabi can help you deploy an easy-to-integrate solution that combats bad actors, to safeguard your users and your brand.
→ Looking for a way to tackle fake reviews on your platform? Find out how we can help you.