September 19, 2026.
Online reviews were supposed to give consumers something advertising couldn't: the experiences of other customers.
Today, reviews are everywhere.
Buy a roll of paper towels at Walmart or Home Depot and you may receive a request to review it. Book a hotel and Tripadvisor may ask about your experience. Search for a business on Google and its star rating can appear prominently. Visit a company's website and a Trustpilot score may be presented as evidence of customer satisfaction.
Stars have become a form of currency.
But how much should consumers—and businesses—trust them?
The answer isn't that online reviews are either trustworthy or fake. Major platforms have sophisticated systems for detecting fraud, and businesses generally cannot simply purchase a five-star rating.
But regulators and researchers have documented problems involving fake reviews, selective solicitation, incentives, AI-generated content, review bombing and reviews submitted before customers have actually experienced a product or service.
The problem also runs in both directions.
Fake positive reviews can mislead consumers. Fake negative reviews can damage legitimate businesses.
And even genuine reviews can produce a misleading picture if the people providing them aren't representative of the company's customers.
The better question is:
How were these reviews collected, and what does the rating actually represent?
Trustpilot: Paying for Services Isn't Buying Reviews.
Trustpilot is a useful case study because it operates a commercial business serving companies whose reputations appear on its platform.
Its current business pricing starts at $99 per month per domain, with higher-priced Plus and Premium plans and customized Enterprise pricing. Its services include tools for collecting reviews, displaying Trust Scores and testimonials, and analyzing review activity.
That does not mean companies can simply pay Trustpilot for five-star reviews.
Trustpilot's policies prohibit businesses from manipulating reviews, selectively soliciting positive reviews and offering incentives for particular outcomes.
But the commercial relationship remains relevant because businesses pay for services connected to collecting and presenting reviews.
In 2026, that relationship received significant regulatory scrutiny.
A €4 Million Warning.
On March 23, 2026, Italy's Competition Authority fined Trustpilot €4 million. €4 million is approximately $4.60 million U.S. dollars.
The authority said Trustpilot's review-collection practices could affect the representativeness of ratings because businesses using its services could select which consumers received invitations to submit reviews.
The regulator also raised concerns about information provided to consumers about Trustpilot's operation and the businesses using its services.
Trustpilot disputed the findings and said it would appeal.
The decision does not establish that Trustpilot reviews are generally fake.
It raises a different question:
Can thousands of genuine reviews still create a misleading impression if the people invited to provide them aren't representative of all customers?
Yes.
Genuine Doesn't Mean Representative.
Imagine a company that has 100,000 customers.
If every customer has an equal opportunity to provide a review, the resulting sample could provide a broad picture.
But what if only certain customers receive invitations?
Perhaps customers who completed successful transactions are more likely to be asked. Perhaps dissatisfied customers are less likely to receive an invitation. Perhaps very happy customers are simply more likely to respond.
The reviews can all be genuine while the overall rating still fails to represent the company's entire customer base.
Authenticity and representativeness are not the same thing.
Online reviews generally aren't scientific surveys. They are collections of people who chose—or were prompted—to provide an opinion.
How Many Fake Reviews Escape Detection?
This is one of the hardest questions to answer.
Platforms can report how many fraudulent reviews they identify and remove. They cannot directly count the fraudulent reviews they never identified.
You cannot reliably count what you don't know exists.
Independent research can, however, estimate the underlying problem.
A United Kingdom government-commissioned study examined reviews in three product categories—consumer electronics, home and kitchen, and sports and outdoors—on popular UK e-commerce platforms.
It estimated that approximately 11% to 15% of reviews in those categories were likely fake.
That figure needs an important qualification.
It does not mean that 11% to 15% of all online reviews are fake.
The research covered specific product categories and platforms and used a methodology for estimating likely fake reviews. It was not a universal measurement.
The study did not attempt to measure every form of review manipulation or the effect of inflated star ratings.
Its significance is that it attempted to estimate the underlying prevalence of fake reviews rather than simply counting those a platform detected.
What Platforms Catch Doesn't Tell Us What They Miss.
Tripadvisor illustrates the distinction.
For 2024, Tripadvisor reported 31.1 million reviews submitted and said it identified and prevented 2.7 million fraudulent reviews from appearing.
The company said about 54% of fraudulent submissions involved "review boosting," including attempts by business owners, employees or affiliated people to improve rankings.
Tripadvisor also reported approximately:
- 9,000 businesses warned about incentivized reviews;
- 360,000 removed reviews connected to employee incentive programs; and
- 214,000 AI-generated reviews identified and removed.
These are Tripadvisor's own figures, not an independent audit.
And the 2.7 million figure represents fraudulent reviews Tripadvisor says it detected.
It does not tell consumers how many fraudulent reviews existed but escaped detection.
That distinction applies to every platform.
The number of fake reviews a company removes tells us how many it found—not how many existed.
Google: Reviews Are Encouraged; Manipulation Isn't.
Google requires reviews to reflect genuine experiences.
Its policies prohibit businesses from paying for reviews, offering incentives for particular ratings, selectively soliciting positive reviews and manipulating review content.
At the same time, Google allows businesses to ask customers for genuine reviews.
That distinction matters because requesting reviews has become routine.
Businesses can send customers a link or QR code asking them to provide feedback.
The request isn't necessarily improper.
The problem is when the process is designed to produce a predetermined result.
The Review Request After the Purchase.
Consumers encounter review requests constantly.
Buy a product online and an email may arrive days later.
Purchase something in a store and the retailer may later ask for feedback.
Home Depot's privacy statement says the company may contact customers to ask them to review products or services they ordered.
That doesn't mean every Home Depot customer receives a request. It does show that post-purchase review solicitation is part of the company's customer-communications practices.
Walmart similarly solicits product reviews and operates a Recognized Reviewer Program through which selected customers can receive products at no cost in exchange for honest reviews.
Walmart says participants are expected to provide their honest opinions.
That doesn't make those reviews dishonest.
But it shows why consumers should pay attention to how a review was generated.
A customer who buys a product with their own money and someone who receives a free product through a review program have different circumstances, even if both provide honest opinions.
The FTC Draws a Line.
The Federal Trade Commission's Consumer Reviews and Testimonials Rule took effect in October 2024.
The rule addresses fake or false reviews, deceptive testimonials and other forms of review manipulation.
The FTC prohibits certain practices involving fake reviews and compensation conditioned on expressing a particular positive or negative sentiment.
But it does not prohibit every incentive associated with reviews.
An incentive for providing an honest review is different from paying someone specifically to provide a positive review.
A Real Person Doesn't Necessarily Have Real Experience.
One of the most revealing FTC cases involved Sitejabber.
In 2024, the FTC alleged that Sitejabber represented ratings as coming from customers who had experienced the products or services being reviewed.
According to the FTC's complaint, Sitejabber collected some ratings at the time of purchase—before customers had received or experienced the products or services.
The FTC said those ratings were subsequently used to increase displayed ratings and review counts.
The lesson is straightforward:
A real person can submit a real rating without having actually experienced the product being rated.
Fake Negative Reviews Can Be a Weapon.
Most discussions about fraudulent reviews focus on companies trying to make themselves look better.
But the opposite problem can be just as serious:
A legitimate business can be attacked through fabricated negative reviews.
Academic researchers have studied retaliatory reviews—negative reviews posted to punish or damage a business rather than accurately describe a customer's experience.
Research has also examined coordinated review attacks and "review bombing."
A 2016 study published in Information Systems Research examined more than 2.3 million reviews involving 4,709 hotels in 17 cities and studied how strategically placed fake reviews could affect a business's visibility.
The researchers found that relatively small numbers of strategically placed fake reviews could have disproportionate effects because rankings influence which businesses consumers see.
That means a business may not need thousands of fake reviews to suffer damage.
A small, coordinated attack can potentially have a much larger economic effect if it changes a company's rating or visibility.
Review Blackmail.
An even more troubling form of abuse is review blackmail.
Researchers have studied situations in which someone threatens a business with a negative review unless the business provides money, products, discounts or another concession.
The review becomes a bargaining weapon:
Give me what I want—or I will damage your reputation.
Academic research published in Management Science has examined this phenomenon and the economic incentives surrounding it.
That's very different from a customer saying, "I had a terrible experience and I'm going to tell other people."
The problem is when someone deliberately manufactures or threatens negative information for personal gain.
When Reputation Attacks Cross a Legal Line.
There are documented cases in which manipulation of online business information has moved beyond questionable marketing practices into conduct investigated or prosecuted by government authorities.
That does not mean every negative review campaign is criminal.
Nor does it mean most review manipulation involves criminal organizations.
But prosecutions and regulatory actions demonstrate that online reputation has become economically important enough to attract sophisticated attempts at manipulation.
For a small business, the stakes can be particularly high.
A company can spend years building a positive reputation. A sudden wave of fraudulent one-star reviews can potentially affect its rating, search visibility and customer acquisition before the business has an opportunity to respond.
The Problem Runs Both Ways.
The review system therefore has two major risks.
Fake positive reviews can harm consumers.
They can persuade people to buy products or services that don't deserve the rating.
Fake negative reviews can harm legitimate businesses.
They can discourage potential customers from considering a company that may have done nothing wrong.
Both undermine legitimate reviews.
The challenge for platforms is to protect consumers from dishonest businesses while also protecting honest businesses from dishonest reviewers.
Stars Are Not Scientific Measurements.
Perhaps the biggest mistake consumers make is assuming that a 4.8-star rating means 96 percent of customers are satisfied.
It doesn't.
A review score is not automatically a scientific survey.
The number depends on:
- who was invited;
- who responded;
- when they responded;
- what experiences they had;
- whether incentives were involved;
- what reviews were removed;
- what reviews were never detected;
- how the platform calculates its score; and
- how reviews are displayed.
The same applies to one-star reviews.
A one-star review may describe a genuinely terrible experience.
But it may also be an exaggeration, a retaliatory review, a review from someone who never purchased the product, a competitor's attack or part of a coordinated campaign.
Neither five stars nor one star automatically establishes the truth.
What Consumers Should Look For.
Consumers don't need to abandon online reviews.
They should look beyond the headline number.
Look for patterns:
- Are multiple customers describing the same problem?
- Are reviews detailed or repetitive?
- Are complaints consistent over time?
- Are there suspicious bursts of reviews?
- Do Google, Tripadvisor, Trustpilot and other sources tell similar stories?
- Does the company respond to complaints?
- Are reviewers identified as verified purchasers?
- What does "verified" actually mean on that platform?
- Are incentives disclosed?
- Were customers given a comparable opportunity to provide feedback?
The goal isn't to find the platform with the highest rating.
It's to understand why the rating looks the way it does.
What Businesses Should Watch For.
Businesses also have a legitimate reason to be concerned about the review economy.
A company can follow every rule and still become vulnerable to fabricated reviews.
A legitimate negative review is valuable information.
A fabricated negative review is something else entirely.
Businesses therefore face a difficult balancing act: responding to legitimate criticism without trying to silence genuine customers while defending themselves against potentially fraudulent attacks.
A negative review shouldn't automatically be treated as an attack.
But an unusual pattern—particularly a sudden wave of reviews from people who don't appear to have genuine customer relationships—may warrant investigation.
The Five-Star Illusion.
The modern review economy has created a complicated system.
Consumers need reviews to make informed decisions.
Businesses need reviews to establish credibility.
Retailers encourage customers to write reviews.
Review platforms sell businesses tools to collect and display reviews.
AI can now generate convincing review text at enormous scale.
And increasingly sophisticated technology is being used to detect fraud.
Yet consumers and businesses have limited visibility into what happens behind the rating.
That's why the most important question isn't simply:
"How many stars does this company have?"
It is:
"How did those stars get there?"
The evidence shows that online reviews can provide valuable information. It also shows that fake reviews, selective solicitation, incentives, review bombing and retaliatory reviews are real problems.
Trustpilot's 2026 regulatory case, Tripadvisor's reported detection of millions of fraudulent reviews, Google's review policies, Walmart's incentivized reviewer program, Home Depot's post-purchase review requests and the FTC's enforcement actions all point to the same larger reality:
The online review is no longer simply one customer telling another customer what happened. It has become part of a sophisticated commercial ecosystem.
That doesn't make every review unreliable.
It means consumers—and businesses—should understand the system behind the stars.
Because sometimes the most important information isn't the rating itself.
It's who was asked, who responded, what they actually experienced, what was removed—and what may have escaped detection.
