Review Manipulation: What It Is and How It Works
Review manipulation is a deliberate attempt to influence which reviews, ratings, or customer experiences people see in a way that creates a misleading overall picture. A review does not have to be fabricated to contribute to that distortion. Customers may describe real experiences, yet the reviews visible to the next shopper can still overrepresent certain opinions, leave other experiences out, or suddenly shift because an unusual group of people starts posting at the same time.
Online review manipulation is therefore broader than simply publishing fake reviews. The key distinction is between the authenticity of an individual review and the representativeness of the overall review picture. A genuine review may describe a real experience without showing whether the reviews consumers can see reflect the wider range of customer experiences.
That matters because reviews influence how people judge quality, risk, and credibility — part of why customers trust reviews. Review manipulation can interfere with that judgment when feedback is created, collected, filtered, displayed, or amplified.
Review Manipulation Can Change More Than Individual Reviews
Online reviews are not a random sample of every customer experience. Some people never leave feedback, while others are more motivated to post after an unusually good or bad experience. That imbalance can exist without deliberate manipulation, but it shows why an average rating or visible review feed should not automatically be treated as a direct reflection of the entire customer base.
A Management Science study comparing private customer satisfaction ratings with public reviews found extremity bias in unsolicited online word-of-mouth: customers with more extreme experiences were more likely to participate. Soliciting reviews increased participation among customers with moderate experiences and made rating distributions more representative, although the bias did not disappear entirely.
More recent research published in Cornell Hospitality Quarterly adds another layer. Researchers studying customers invited to post on Google Travel or TripAdvisor identified selection bias because not every invited customer ultimately posted, as well as measurement bias because some public ratings differed from earlier survey responses. The patterns also differed between platforms.
These findings show why authenticity and representativeness are different questions. Manipulation can exploit the same weaknesses deliberately by influencing which experiences enter the review pool, which reviews remain visible, or how much weight a particular group receives.
Where Distortion Enters the Review Process
The most obvious manipulation happens when false experiences are added to the system, but distortion can also occur before or after a genuine customer writes anything. Looking at the full process makes it easier to see why review manipulation cannot be reduced to fake text alone.
Fabricated Reviews Add Experiences That Never Happened
Fake positive reviews can create the impression of satisfied customers who never existed or never used the product. Fake negative reviews can be used to damage a competitor or lower another business's rating. Buying reviews through brokers or coordinated services can extend the same behavior across larger numbers of accounts.
Fabricated feedback can also include AI-generated reviews when the text falsely represents a customer experience that never occurred. The use of AI is not what makes the review deceptive; the problem is presenting a fabricated experience as genuine customer feedback.
The FTC's Consumer Reviews and Testimonials Rule, which took effect on October 21, 2024, addresses fake or false consumer reviews and several other deceptive review practices. It also prohibits compensation or other incentives when the reward is expressly or implicitly conditioned on a review expressing a particular positive or negative sentiment.
This type of review fraud changes the visible picture by introducing experiences that did not occur or by paying for a predetermined opinion. Fake review detection tools can help surface suspicious patterns, but they cannot explain every form of manipulation. Selection, suppression, and coordinated activity can distort a review profile even when individual reviews appear genuine.
Selective Requests Change Who Gets a Chance to Speak
A subtler problem appears when the review collection process favors customers expected to leave positive feedback. Imagine that every customer receives a satisfaction survey. Customers who select “excellent” are encouraged to post publicly, while customers who report problems are directed only to private support.
The resulting public reviews may be genuine, but positive experiences have been given a clearer route into the public review pool than negative ones. This is review gating: using a customer's sentiment to determine whether they are encouraged or directed to leave a public review.
A private feedback channel by itself does not prove gating. Businesses have legitimate reasons to collect complaints privately and resolve service issues. The important question is whether customer sentiment determines who is encouraged or directed to publish feedback publicly.
The FTC notes that its Consumer Reviews and Testimonials Rule does not specifically prohibit requesting reviews only from customers a company expects to be happy. Depending on the circumstances, however, the practice can still raise concerns under the broader FTC Act.
Platform policies can go further. Google Maps' content policy prohibits merchants from discouraging negative reviews, selectively soliciting positive reviews, or offering incentives such as payments, discounts, or free goods or services in exchange for reviews.
That distinction matters for incentivized reviews. An incentivized review is not automatically fake under the FTC rule: an incentive may be offered for an honest review if it is not tied to a particular sentiment, although disclosure and other requirements can still apply. Google Maps has a stricter platform-specific policy and does not allow incentivized reviews.
Businesses should therefore check both applicable law and the policy of the destination platform. A safer collection process focuses on how to ask customers for honest reviews without making the invitation depend on an expected rating.
Suppression Changes What Remains Visible
The review picture can also be distorted after feedback has been submitted. This is where legitimate moderation and review suppression need to be separated carefully.
Platforms and businesses may reject spam, fake engagement, harassment, irrelevant material, duplicate submissions, or other policy violations. Removing that content can improve review quality. Suppression is different: unfavorable feedback is prevented from appearing, removed, or made harder to find because it is unfavorable rather than because it violates a neutral rule.
The FTC rule prohibits certain suppression practices, including intimidation, false accusations, or unfounded legal threats used to prevent or remove negative reviews. Ordinary organization is not automatically suppression. Sorting higher-rated reviews above lower-rated ones, for example, is different from making negative feedback effectively undiscoverable.
A real enforcement case shows why that distinction matters. In the Fashion Nova case, the FTC alleged that the retailer represented its product reviews as reflecting submitted customer opinions while suppressing reviews rated below four stars. The final settlement required the company to pay $4.2 million and prohibited it from suppressing customer reviews.
The case illustrates why visible positive reviews do not all need to be fake for the overall picture to mislead consumers. Moderation and display policies are therefore part of what makes a review platform trustworthy, because evaluating individual reviews alone cannot reveal every decision shaping the final review feed.
Coordinated Activity Can Change the Weight of Real Opinions
Review manipulation can also work through volume and timing. Review bombing usually involves a concentrated wave of reviews intended to influence a public rating or perception. The trigger may relate to the product itself, but it can also stem from a company decision, controversy, social issue, creator, employee, or another event that is not representative of ordinary customer experience.
Crucially, the participants do not all have to be fake accounts. A peer-reviewed study of review bombing around The Last of Us Part II analyzed 51,120 English-language reviews. The researchers found that fake reviews played a relatively small role in the case and identified ideology-driven polarization followed by counter-bombing from users with opposing views.
This case shows how a temporary concentration of highly motivated real users can change the weight of opinions in a public rating. A genuine person can leave a genuine opinion while still participating in a broader pattern that makes the overall review picture less representative.
Steam has addressed a related problem by identifying periods of anomalous review activity and investigating whether they qualify as off-topic review bombs. Reviews from such periods can remain accessible while being excluded from the calculated Review Score.
A spike alone, however, is not proof of review bombing. A sudden increase in reviews can follow a major product update, service outage, price change, viral campaign, influx of new customers, or another legitimate event. Timing can justify investigation, but it cannot establish manipulation on its own.
How to Spot Review Manipulation Without Treating Every Anomaly as Proof
The difficulty with spotting review manipulation is that most warning signs also have legitimate explanations. A new reviewer account could belong to someone posting for the first time. Similar wording may appear because many customers encountered the same problem. A burst of negative reviews may follow a real failure affecting a large number of customers.
That is why isolated signals should be treated as clues rather than conclusions. FTC consumer guidance recommends looking at factors such as timing, sudden bursts of activity, reviewer histories, and information from multiple sources. It also cautions that consumers often cannot determine whether a review is genuine simply by reading it.
When something looks unusual, examine several dimensions together:
- Timing: Did review volume change suddenly, and was there an identifiable event that could explain it?
- Reviewer history: Are many accounts new, inactive elsewhere, or showing unusual patterns?
- Language and themes: Do large groups of reviews repeat distinctive phrases or talking points?
- Rating distribution: Did the mix of ratings change sharply compared with the longer-term pattern?
- Collection method: Were some customers offered incentives or different review paths based on their experience?
- Visibility: Are certain types of feedback consistently harder to find without a clear moderation reason?
- Cross-platform context: Does the same shift appear elsewhere, or is it concentrated on one review source?
No single percentage, rating level, account age, or review-volume threshold universally proves manipulation. A review spike becomes more informative if it coincides with repeated language, unusual reviewer histories, and no comparable increase in customer activity. A rating difference between two platforms becomes more meaningful when their collection methods or audiences also differ.
The goal is not to label every anomaly as fraud. It is to identify what changed between the underlying customer experiences and the review picture people can actually see.
Businesses Should Audit the Process Behind the Rating
Businesses trying to protect review integrity should examine the whole path from customer experience to visible review rather than watching the average star rating alone. That means treating collection, incentives, moderation, publication, and unusual changes in review activity as connected parts of the same system.
Start with collection. Who receives a review request, and why? If only certain customers are being invited, the business should understand the reason. Review campaigns should not quietly give satisfied customers a better route to public feedback than dissatisfied ones.
Then examine incentives. Determine what customers receive, what they are asked to do in return, how the arrangement is disclosed, and whether the destination platform allows incentivized reviews. A campaign can comply with one requirement and still violate another platform's policy.
Next, review moderation and publication. There should be a defensible reason for removing or withholding content, and the same standard should apply regardless of whether the review is positive or negative. Resolving a customer's complaint is different from making the complaint disappear.
Finally, monitor changes in volume, rating distribution, and reviewer behavior alongside real business events. A sudden negative wave may reveal coordinated activity, but it may also expose an operational problem that needs attention. The same caution applies to unusually positive growth.
The objective is not to create a perfectly uniform distribution of opinions, because customer experiences are rarely uniform. The stronger standard is whether the process systematically gives one type of experience a better chance of entering, remaining in, or dominating the visible review picture.
