What’s a Fake Product Review, Anyway?

Put simply, a fake product review is a non-genuine written review of a product or service. In some cases, fake reviews are purchased by a company in order to boost their average ratings on a site like Amazon. Sometimes, a company’s employees may leave fake reviews for the company’s products across various retailer and review sites.

However, fake reviews don’t have to involve payment — it’s also possible that a company’s competitors may engage in leaving fake negative reviews about their products or services, and even consumers may leave fake reviews for various reasons despite not being compensated for doing so. For example, sometimes a product gains traction in the media as being terrifically bad for one reason or another, prompting jokesters to concoct elaborate stories illustrating just how terrible the product really is.

One example is the Haribo Sugar Free Gummi Bears, a product which, after being mentioned in a few articles (see BuzzFeed and Slightly Viral, among others), has accumulated hundreds of reviews on Amazon, many of them admittedly hilarious. While it makes for good entertainment value (warning: reading those reviews is not for the faint of heart — or those with a weak stomach!), such trends give a skewed perception to consumers who aren’t there for a laugh but are actually considering purchasing the product. So, impressive writing talent among a few of those reviewers aside, fake reviews make it hard for consumers to get a reliable feel for the true quality of a product.

Incentivized Reviews and Bots Muddy the Waters

Incentivized reviews can blur the lines between blatantly false reviews and reviews that are a bit more generous than they might otherwise be if no incentives were involved. As TechCrunch reports (based on research from ReviewMeta analyzing 7 million reviews), the average rating for products with incentivized reviews is higher than that of products with non-incentivized reviews. Additionally, incentivized reviewers are 12 times less likely to give a 1-star rating compared to reviewers who are not receiving an incentive, and they’re also nearly four times less likely to leave a negative review overall.

“The compensated-review process is simple: Businesses paid to create dummy accounts purchase products from Amazon and write four- and five-star reviews. Buying the product makes it tougher for Amazon to police the reviews, because the reviews are in fact based on verified purchases,” explains Lauren Dragan of The Wirecutter. “The dummy accounts buy and review all sorts of things, and some of the more savvy pay-for-review sites even have their faux reviewers pepper in a few negative reviews of products made and sold by brands that aren’t clients to create a sense of ‘authenticity.’ In fact, for extra cash, a company can pay one of these firms to write negative reviews of a competitor’s product.”

Bots are also becoming a problem in the fake review sphere. “Multiple small companies report they’re seeing one-star reviews of unverified purchases on their pages that are written with bad grammar, coupled with remarks like, ‘Great product satisfaction guaranteed,’” reports Digiday in a September 2017 article. “The problem seems to run across both the first-party and third-party sellers on the platform.”

While bot reviews may be easier for the average consumer to detect, companies have little recourse in dealing with these bot-generated reviews, which are believed to originate at least in part from competitors. Outside of clicking “report abuse” and making great effort to actively engage with customer feedback, which can be a daunting task for smaller vendors who lack dedicated support teams, there’s not much a company can do to mitigate the damages caused by review-generating bots.

Facts and Stats on Fake Reviews

More than nine out of 10 shoppers (92%) say they read online product reviews, and 88% of online shoppers say they rely on reviews to inform purchase decisions. What’s more, 79% of consumers have read a fake review in the past year, but most consumers (84%) can’t always identify a fake review when they read one.

Despite retailers making a concerted effort to reduce the problem of fake reviews (more on those efforts below), studies show that inauthentic reviews remain a significant problem. According to research conducted by ReviewMeta.com, which analyzes reviews in order to arm consumers with information on which reviews to trust, the average weight of reviews on Amazon dropped considerably between June and August 2017, meaning the number of untrustworthy reviews is on the rise. At the same time, average product ratings increased dramatically during the same period. In other words, an increasing number of suspicious reviews are being posted.

According to tactac, anywhere between 10% and 35% of online reviews are fake or incentivized — meaning potentially more than one out of every three reviews. What’s more, more than half of all reviews on Amazon are incentivized. While 95% of reviewers suspect that some reviews are fake, it’s difficult for consumers to filter out the genuine from the non-genuine, and even reviews a consumer suspects are fake can serve to sway opinions.

On average, a fake review costs about $5, so newer products with few existing reviews can easily nudge their average rating to four out of five stars or better with a few hundred bucks. For consumers who don’t read reviews in detail and instead rely on star ratings displayed on product category pages, these practices can easily lead a buyer to evaluate (and sometimes purchase) products that aren’t actually as good as reviews make them appear.

How Amazon Combats Fake Reviews

Amazon has always prohibited paid reviews, and in fact, it has taken legal action against companies that purchased reviews in bulk or even on an individual basis. The exception is books, as advance review copies have long been standard in the publishing industry.

In 2016, Amazon introduced a machine learning algorithm that gives more weight to newer reviews as well as those deemed more helpful than others. Additionally, Amazon tightened the criteria for qualifying for the Amazon verified purchase badge and also banned or suspended (or in some cases, even sued) vendors who were found attempting to manipulate the ratings system.

In October 2016, Amazon put an end to incentivized reviews, which allowed vendors to provide free or discounted products to reviewers in exchange for reviews, unless vendors are enrolled in the Amazon Vine program.

“Here’s how Vine works: Amazon — not the vendor or seller — identifies and invites trusted and helpful reviewers on Amazon to post opinions about new and pre-release products; we do not incentivize positive star ratings, attempt to influence the content of reviews, or even require a review to be written; and we limit the total number of Vine reviews that we display for each product,” explains Chee Chew, Amazon VP of Customer Experience, in a blog post. For vendors selling products on Amazon, it costs about $2,000 for 30 reviews via the Vine program, making it considerably more expensive than the estimated $5 per review via other avenues.

The problem with incentivized reviews is that they tend to be more positive than those posted by reviewers who are not incentivized. ReviewMeta.com finds that incentivized reviews receive ratings an average of 0.38 stars higher than non-incentivized reviews. So, banning incentivized reviews was aimed at reducing the number of non-genuine positive reviews, but the problem hasn’t gone away since Amazon halted the program — it’s gotten worse.

According to FakeFilter, the act of banning incentivized reviews could be the very catalyst to the worsening of the problem, as reviewers now no longer have to disclose that they received a product for free or at a discount in exchange for the review (seeing how the practice is disallowed), and companies that emerged for the sole purpose of supplying fake reviews to vendors are already finding creative ways to work around the problem.

ReviewKick, for instance, simply responded to the ban by announcing that they “no longer request or require customers who purchase products through Review Kick at a discount to leave a product review.” As FakeFilter points out, the ban drives more users to the Amazon Vine program as it’s the only “allowable” program for getting discounted products, but even a resurgence of this program would lead to inherently biased reviews. Simply put, consumers who get a product for free or at a deep discount tend to feel an obligation to leave reviews that are at least neutral, if not overwhelmingly positive.

How Other Retailers and Review Sites Combat Fake Product Reviews

Of course, Amazon isn’t the only retailer facing the growing problem of fake reviews, nor is it the only entity to take action to combat the issue. Yelp, a popular review site for local businesses ranging from restaurants to hotels, has its own considerable fake review problem. In 2013, the New York State Attorney General caught 19 companies engaged in writing fake reviews for businesses in exchange for compensation on the platform. The perpetrators were fined in varying amounts totaling $350,000.

While many review sites have filters that aim to detect fake and disingenuous reviews and block them from being posted (such as Google and TripAdvisor, for instance), the New York State Attorney General stated at the time (2013) that Yelp’s review filter was the most aggressive. In fact, Yelp has had a review filter in place since two weeks after the platform’s launch, necessitated by the appearance of obviously fake reviews. (Yes, fake reviews started showing up just two weeks after Yelp’s launch.) While Yelp has garnered some criticism for a lack of transparency about how the filter works, revealing the intricate details of the algorithm would only serve to arm those who wish to circumvent it with the ammunition to do exactly that.

The fake review trend plagues digital products, too. In 2016, Google implemented measures to filter out disingenuous reviews in the Google Play Store as well as to prevent practices such as inflated install numbers that can mislead consumers. What’s more, if developers are caught using shady tactics that go against the Google Play Developer Policy, Google may revoke their account and remove all the developer’s apps from the Play Store. That’s precisely what happened to Dash in October 2016.

Still, even Google’s likely sophisticated review filters fail to correctly identify and filter out fake reviews. As recently as May 2017, for instance, sleuths pinpointed reviews appearing in Google My Business for various businesses all originating from the same reviewer — who happened to utilize an aquarium service in Royal Palm Beach, Florida, a tree trimming service in Evansville, Indiana, and a landscaping service in Colorado within the same month, in addition to selling a used car to a buyer in Louisiana during the same time frame. Either this guy moves around faster than he changes his underwear, or Google My Business’ review filter is less than perfect.

TripAdvisor is another platform that relies heavily on reviews to make its services more valuable to consumers, but like all such platforms, fake reviews work against those goals by making reviews less useful. The company has a clear set of traveler reviews that reviewers must abide by, along with an automated filter that flags questionable content for a closer look and a team of moderators to handle those potentially fraudulent reviews. Additionally, like nearly all review sites, TripAdvisor’s users have the ability to “report inappropriate content” on every published review.

Common Characteristics of Fake Product Reviews

Amazon aims to make incentivized reviews (those that are obtained via the Amazon Vine program) obvious to consumers by clearly displaying a message such as, “Amazon Vine Review of Free Product,” where reviews left by verified purchasers state, “Verified Purchase.” That doesn’t mean that other consumers who have purchased products elsewhere can’t leave reviews for products sold at Amazon, but these reviews don’t have the “Verified Purchase” badge, and in fact, users can filter reviews in order to show only those left by verified purchasers — boosting the odds that the reviews are based on genuine feedback rather than inflated due to incentives.

There are some other tactics you can use to spot fake reviews, as well. For example, if you come across products that have hundreds of reviews with high average ratings and few to none of those reviews are from verified purchasers, the odds are good that many of those reviews are fake. On Amazon, in particular, you might find several listings for the same product from various vendors. If one of those product listings has an average of 4.9 stars with a few hundred reviews, while others have average ratings of 1.5 stars or even 2.5 stars, it’s probably safe to assume that reviews for one of those listings have been inflated or deflated with the aid of fake reviews.

Fake reviews don’t always inflate an average rating favorably, so just because a product has a lower average rating doesn’t mean the reviews are genuine. Vendors can gain a competitive advantage by procuring fake negative reviews for competing product listings, which in turn paints their own listing in a more favorable light in the eyes of consumers.

Here are a few warning signs that indicate you might be reading fake product reviews:

• A large quantity of 5-star reviews from non-verified purchasers
• Several identical product listings with drastically different average ratings
• Hundreds to thousands of reviews from non-verified purchasers
• Most 4- and 5-star reviews from non-verified purchasers, coupled with 1- or 2-star reviews which are mostly (or all) from verified purchasers (or vice-versa)
• Strange syntax (odd phrasing) in reviews, which may indicate that the reviews are generated using bots or text-spinning software
• Little-known brands with no website presence that have garnered thousands of reviews in a short period of time
• A number of reviews with similar high or low ratings posted in a short amount of time, particularly for products that don’t otherwise garner dozens or hundreds of reviews that quickly
• Several or many reviews that have similar wording or similarly staged user photos
• Reviews that vary widely from site to site (a perfect 5-star rating on Amazon in thousands of reviews, while the same product at Walmart has a 2-star rating in a few hundred reviews)

While the above indicators can lead you to concluding that a good portion of reviews for a certain product are fake, there are also some indicators that can help you determine whether an individual review is authentic or a likely fake:

• Lacking in detail. “It’s hard to describe what you haven’t actually experienced, which is why fake reviews often offer general praise rather than digging into specifics,” says Jessica Stillman in an article for Inc.
• Heavy use of first-person pronouns. According to Stillman, “If you’re anxious about coming across as sincere, apparently you talk about yourself more. That’s probably why words like ‘I’ and ‘me’ appear more often in fake reviews.”
• More verbs than nouns. “Language analysis shows the fakes tend to include more verbs as their writers often substitute pleasant (or alarming) sounding stories for actual insight. Genuine reviews are heavier on nouns,” Stillman explains.
• Reviews from non-verified purchases. While non-verified purchase reviews aren’t always fake, they’re more likely to be inauthentic compared to reviews from verified purchasers.
• Staged photographs. Genuine reviewers may include photos with reviews when warranted, but fake reviews are more likely to include photos that appear to be intentionally staged to get a point across.
• Short, sloppy reviews. Paid reviewers may only be making a few bucks for each review, so they’re not always likely to take an hour to craft a compelling, highly detailed analysis of a product that they’re earning $5 to write.
• Reviews that read like an infomercial script. This may be an SEO tactic aimed at improving the visibility of a product or company, but if a review contains the complete, overly-detailed product name several times in a paragraph, you might be reading a fake.
• Obvious bias. Surprisingly, some reviewers will state outright that they haven’t actually used the product. Often, these reviewers have some other motive for writing a review, and they’re usually not shy about making it known (for instance, a book reviewer leaving a negative review of a book they haven’t read because they’ve read the author’s previous works and didn’t like them).

You should also evaluate the reviewers themselves. Most websites require users to sign in using Facebook or another social networking account (or even Google) so their reviews are tied to a real person’s identity. That makes it possible for consumers to confirm that reviews are coming from real, live people.

Plus, you can check out their feedback percentage on some sites (such as eBay and Etsy), and on most, you can review a reviewer’s review history (try saying that three times fast). If an individual has reviewed hundreds of gadgets over a two-week period, the odds are good that they haven’t actually used all of those products. Another warning sign is a reviewer who leaves only 5-star or only 1-star reviews.

The real trick to identifying fake reviews is to know the tricks of the trade. However, as Bing Liu, a professor in the department of computer science at the University of Illinois at Chicago, tells The Wirecutter, this is often an easier prospect for the platform itself (such as Amazon or Yelp) than it is for consumers. That’s because review platforms and retail sites have access to more data about the individual users visiting and interacting with their sites as well as elsewhere on the web, making it easier to pinpoint unusual patterns of behavior that can be used to identify potentially suspicious reviewers.

Liu also points out that there’s no sure-fire way to know for certain that a review flagged as suspicious is truly inauthentic, unless the reviewer confirms it as such. That’s pretty unlikely (someone being paid to post fake reviews isn’t likely to admit to the practice, as they’re essentially risking their livelihood by doing so), unless of course a review platform or company takes legal recourse.

What you can do, however, is become familiar with the common characteristics of fake product reviews above, and use that knowledge to give more weight to reviews that are more likely to be genuine. Additionally, there are a few third-party tools cropping up that can help consumers determine whether reviews are trustworthy or not.

Tools for Identifying Fake Reviews

Consumers are becoming increasingly discouraged by the prevalence of fake reviews — so much so that it’s hindering consumer trust and confidence. In response, a cottage industry has emerged with companies offering products and services aimed at helping consumers identify fake reviews or, in some cases, make sense of hundreds or thousands of reviews by identifying common threads and trends.

Fakespot — Analyzes Amazon, Yelp, TripAdvisor, and App reviews. Fakespot also offers free browser extensions for Chrome, Firefox, and Safari, making it even easier to spot fake reviews. According to Rick Broida in an article for CNET, “Fakespot analyzes both reviews and reviewers, looking for questionable spelling and grammar, number of reviews, purchasing patterns, mismatched dates and other telltale signs of suspicious review activity.” Fakespot offers a letter grade on review quality, which can be a bit confusing. A grade of “F,” for instance, doesn’t meant the product is bad — it means there are a lot of fake (or suspicious) reviews.

ReviewMeta — Analyzes reviews from Amazon and other platforms such as Bodybuilding.com. “Although it’s functionally similar [to Fakespot] — paste in an Amazon link or use one of the browser extensions — ReviewMeta merely strips out or reduces the weight of certain reviews, then leaves you with an adjusted rating,” says Broida.

FakeFilter — Allows you to shop on Amazon without the fake reviews. FakeFilter removes all sponsored, fake, low-quality, and unhelpful reviews, scores the remaining reviews based on quality and usefulness, then applies a weighted algorithm to create a more accurate star rating.

Review Skeptic — Based on research at Cornell University that uses machine learning to identify fake hotel reviews. “The tool can spot fake hotel reviews with almost 90 percent accuracy, which is better than most people,” says Patrick Allan in an article for Lifehacker. “We tend to suffer from ‘truth bias,’ meaning we assume something is true until we find evidence to the contrary, but a machine doesn’t have to worry about that. It simply analyzes the words used and spots the deceptive language.”

The Review Index — A neural network-based tool that analyzes Amazon reviews (electronics and gadgets only) and identifies recurring threads and patterns, displayed in easy-to-understand charts and graphs. “The Review Index does the hard work of removing the spam reviews,” says Saikat Basu in an article for Make Use Of. “It also removes all reviews older than two years because they might not be relevant for an updated product. Also, pay attention to the number of unverified reviews which is also another set you can ignore as the reviewers haven’t bought the product themselves.”

Of course, given that algorithm giant Google has yet to perfect the review filtering process, it’s probably safe to assume that these tools aren’t yet perfect, either. And as Skeptical Science shows, if reviews are written in a way that sounds authentic (instead of text-spinning or quick, generic comments) they’re more likely to pass filters, so as the quality of fake reviews improves, they’ll become increasingly challenging to distinguish from genuine reviews. That said, there’s certainly something to be said for having the ability to get two different perspectives on the quality and authenticity of the reviews you’re reading: one perspective from the platform itself, and another from a third-party analysis using the tool of your choice above.

Products Most Likely to Have Commission-Driven Reviews

Now that you’re armed with the knowledge you need to identify many of the fake product reviews you’re likely to encounter during a typical online shopping journey, there’s one more important thing to be aware of: Certain products are more likely to have commission-driven reviews than others. Knowing what those product categories are can help you become even more vigilant and make smarter buying decisions.

In addition to hotels and apps, consumer products tend to be heavily plagued by fake reviews, but a few categories stand out. Electronic gadgets, for instance, may be more likely to have fake reviews based on the value of a sale to the vendor. Vendors are more likely to drop some cash to pay for fake reviews if the sales are very profitable for them — it’s not usually worth it to shell out $5 per review for a product with a $0.50 profit margin, but it’s a more attractive investment for a product with a $50 profit margin. Likewise, reviews for kitchen gadgets are frequently inflated with fake or incentivized reviews.

Really, any product that a consumer is likely to read reviews about before making a purchase decision is a target for fake reviews. That can be true of any product. But, in reality, consumers are more likely to take risks on items with a lower cost (because it’s a less risky investment). Someone may be likely to buy a $10 scarf that looks attractive without reading hundreds of reviews, with the idea that they can simply give it away or donate it to Goodwill if it’s not to their liking when it arrives. A $1,200 laptop, on the other hand, isn’t an easily discarded item and therefore carries a greater financial risk to buyers.

Electronic gadgets are not only often more expensive, but they’re more prone to not meeting expectations if they fail to function as they should, so consumers tend to read more reviews on these types of products before buying. So, any product that has clear performance expectations is a potential target for fake reviews, from lawn equipment to paint, beauty products, TVs, smartphones, and more. And when these products have a lot of competition, either among competing manufacturers making similar products or among multiple vendors listing identical items, fake reviews tend to be even more prevalent as companies and vendors attempt to make their products stand out.

Further Reading on Fake Product Reviews

For more information on how to identify fake product reviews and make smarter buying decisions, visit the following helpful resources:

• Coming to terms with fake reviews
• Fake Online Reviews: Here Are Some Tips for Detecting Them
• How to Spot Fake Online Product Reviews
• Are fake online reviews killing consumer confidence?
• 6 Clues That an Online Review Might Be Fake
• How to Spot Fake Reviews on Amazon, Yelp, and Other Sites
• 30 Ways You Can Spot Fake Online Reviews
• Amazon Problems: Fake Product Reviews
• Scammers elude Amazon crackdown on fake reviews with new tricks
• Tackling a Problem of Fake Online Reviews
• Amazon’s Fake Review Problem
• A Surprisingly Large Amount of Amazon Reviews Are Fake

Armed with this knowledge, you have the ability to become a smarter online shopper. The downside is that it takes some critical thinking and a bit more time to thoroughly vet reviews when you’re just trying to purchase a product, but it’s time well spent when you can avoid being scammed by fake reviews of products that don’t live up to the hype.