Remember the old adage “don’t believe everything you read”? Today the axiom remains true: Despite more effective policing, in many ways the internet still resembles the wild west, with scamsters peddling snake oil in the form of bogus reviews. Products and services that are actually inferior are given false four- or five-star ratings to trick the trusting consumer. This digital sleight of hand can also be used to trash competitors with deceptively negative reviews, robbing a quality brand of hard-earned business.
The problem of fake reviews is almost as old as the internet, but one that has intensified over time.
“It started out on a much smaller scale,” says Rob Gross, COO at Fakespot, a data analytics company dedicated to detecting and pointing out fraudulent evaluations. “Let’s say you develop your own product and list it online. You might ask a few friends for a favor to post a few positive reviews to help get your business started. That’s all pretty innocent.”
But the growth of third-party ecommerce beginning around 2005 changed things. “When Amazon developed the third-party marketplace system, it incentivized the gaming of the system,” Gross says.
He says ranking algorithms like Amazon’s consider both the number of reviews and the distribution of their star ratings. To exploit this, an unscrupulous vendor could produce an inferior item (perhaps a knock-off version of a popular product), create a third-party seller account, then post dozens of fake reviews. This will move the item up in search algorithms virtually overnight.
“We’ve seen situations where a product will start out with zero reviews one day,” Gross says, “and the next day it’ll have 3,000.”
Manipulation of this kind proved highly lucrative — and affected everyone, including the honest merchant. With virtually every business having an online presence, from large retailers to the corner bakery or locksmith, maintaining a strong star ranking is essential to capturing business. “If any business has less than four stars, consumers have been trained not to even look at it,” Gross says.
‘LAST LINE OF DEFENSE’
Fakespot was born to root out these deceptions. In 2016 company founder Saoud Khalifah became frustrated with the rampant quantity of misleading or even fraudulent reviews online and decided to take action.
Fakespot functions as a consumer service, operating through Fakespot.com, a Chrome extension and a mobile app. Each analyzes reviews that appear on the sites for Amazon, Best Buy and Walmart, plus hospitality sites TripAdvisor and Yelp.
The proprietary technology the company uses was developed internally and consists of several closely guarded algorithms. “By maintaining our secrecy, we stay one step ahead of the sellers (who sometimes use fake reviews) and the fake review farms,” Gross says. “Every night we see people trying to game our system.”
Effective security also prevents would-be competitors from reverse-engineering Fakespot’s intellectual property.
Fakespot’s system analyzes reviews inputted by consumers and runs them through its artificial intelligence-based algorithms. The algorithms are trained through machine learning, which draws from a database of some 3.7 billion reviews. Natural language processing enhances the procedure, helping to discern common word patterns that often appear in fake reviews.
After the analysis is completed, the system assigns the review a letter grade, “A” through “F.” The higher the grade, the more trustworthy the review. Unreliable reviews are removed from the site, and the star rating of the reliable reviews is reevaluated in light of the findings.
Fakespot also offers other information — a product overview, details from the most positive and least positive vetted reviews, a representation of the most and least authentic reviews, review count history, price history and comparisons with similar products.
“We like to look at Fakespot as a one-stop shop for consumers, and a last line of defense before the customer hits the ‘buy’ button,” Gross says.
As Fakespot continues to evolve technologically, the business model changes along with it. When the company began, the consumer site was supported by advertising and Fakespot charged for the Chrome extension. It recently received $1.3 million in venture funding, however, which enabled the company to remove the advertising and make the Chrome extension free.
“Soon we plan to introduce a B2B product which will help brands better understand what’s going on within their reviews,” Gross says. That will allow future clients to obtain a clean view of customer sentiments once the influence of incentivized reviews is removed.
“People are moving toward more authentic experiences on the internet, and we want to promote that for all honest players.”
While the people at Fakespot have done the public a favor, what about the unfortunate merchants negatively impacted by fraudulent online appraisals? What recourse do they have?
“If a business owner or company perceives a problem of fake online reviews, there’s a full gamut of legal remedies that can be used to obtain relief,” says Matt Cavanaugh, partner in the Cleveland office of the McDonald-Hopkins law firm. “The best method to use will depend upon the nature of the problem. How big an issue is it? How large and influential is the offending party?
“Also, is the source of the problem in the U.S. or is it abroad? If it’s a rogue actor in Russia, U.S. law can be of limited usefulness.”
Within American borders, it’s a different story: Cavanaugh says every state has laws prohibiting individuals from engaging in deceptive conduct in commerce. Lawsuits can be filed on grounds of commercial disparagement or false advertising. On the federal level, suits can be brought under the Trademark Act, which covers false advertising.
If the plaintiff is successful, the court can order injunctive relief, which will force the offending party to take down the fraudulent review. Monetary damages may also be ordered.
In order to prevail, however, it may be necessary to prove causation — the plaintiff must demonstrate a particular review was directly responsible for the merchant’s lost business. Such proof may be difficult to show.
An especially effective strategy may involve some diligent investigative work.
“Try to find the hub,” Cavanaugh says. “Look for companies that use a network of reviewers. The group is approached by the unethical company which makes an offer to pay them to boost their ratings on Amazon or Trip Advisor. They have their minions post phony positive reviews, or negative reviews to bash a competitor. Finding the hub allows the plaintiff’s counsel to use the subpoena power from filing a lawsuit to undercover and expose many of the spokes in that network.”
Detroit-based Paul Vachon writes for various trade publications, in addition to feature stories for consumer magazines and books on Michigan history and travel.