Review fraud is the posting of fake positive or negative reviews to distort how much a product, seller, or business can be trusted. It is bought and sold like a service, and it quietly steers buyers, rankings, and marketplace trust away from what the ratings claim to measure.
What is review fraud?
Review fraud is the manipulation of ratings and written reviews so they no longer reflect real customer experience. It runs in two directions. Positive review fraud inflates a seller or product with paid five-star reviews to lift rankings and win the buy box. Negative review fraud tears down a competitor with coordinated one-star attacks, sometimes paired with false policy complaints to get a rival suspended.
Most of it is a paid service. Brokers run networks of accounts, often called review farms, that will post to order. Common tactics include incentivized reviews where a buyer is refunded in exchange for a five-star post, review hijacking where fake reviews are attached to a listing that was swapped after the reviews accrued, and brushing where a seller ships cheap unordered parcels so they can post as a verified buyer.
For a fraud or trust-and-safety team, the reviews are a signal layer that fraud targets directly. Because so many downstream decisions, from search ranking to buyer confidence, lean on that signal, corrupting it is cheap leverage on the whole marketplace.
How a review campaign works
A paid positive campaign usually moves through the same stages:
- OrderSeller buys reviews A seller hires a broker and specifies how many reviews, what star rating, and how fast to post them.
- StageMake it look verified Farm accounts place real or brushed orders so the reviews carry a verified-purchase badge.
- PostDrip the reviews Reviews land over days from many accounts and devices to dodge simple burst detection.
- ReimbursePay the reviewers Reviewers are refunded off-platform or paid a fee, closing the loop and hiding the incentive.
Who is involved?
Who | Their role |
The buyer of reviews | A seller or business that wants a rating boost or a rival taken down. |
The broker | Runs the account network and sells reviews by the batch, often through chat groups. |
The reviewer accounts | Real people paid per post, or farmed accounts that exist only to review. |
The platform | Owns the ratings system and the buyers who rely on it; carries the reputational and legal risk. |
What it looks like in practice
A new third-party seller launches a phone accessory and joins a chat group where a broker offers verified reviews. Over two weeks, dozens of accounts buy the item, get privately refunded through a payment app, and post glowing five-star reviews with near-identical phrasing.
The listing climbs to the top of search on the strength of its rating. Weeks later the same broker is hired by a competitor to bury it under a wave of one-star reviews complaining of a safety defect, and a linked flood of policy reports pushes the platform to suspend the listing while it investigates. Real buyers on both sides were steered by reviews that were never genuine.
Why it matters to operators
Ratings are a core trust signal, and once buyers learn they can be bought, the whole platform loses credibility, not just one listing. Review fraud also crosses into real financial harm: inflated products underperform and drive returns and chargebacks, while negative-review attacks can wrongly suspend honest sellers and trigger appeals, refunds, and lost sales.
Regulators increasingly treat fake reviews as deceptive practice, so unmanaged review fraud is a compliance exposure as well as a trust problem. The abuse rarely lives in isolation; the same account clusters that post fake reviews often overlap with promotion abuse, brushing, and seller collusion, so it is worth investigating as part of a wider ring rather than post by post.
What to watch in the data
- Bursts and templates. Many reviews in a short window with repeated phrasing, similar length, or the same misspellings.
- Account overlap. Reviewer accounts that share devices, IP ranges, or payment instruments with each other or with the seller.
- Brushing signals. Verified-purchase reviews tied to unusually cheap orders, odd shipping, or recipients who never engaged before.
- Rating whiplash. A listing that swings from a clean record to a sudden one-star pile-on, often alongside matching policy complaints.
- Off-platform reimbursement. Buyers steered to a chat app or gift card to receive a refund after posting, a classic incentivized-review tell.
Quick questions
Is a paid review always fraud?
A disclosed sponsorship is not fraud. It becomes fraud when the incentive is hidden and the review is presented as an independent verified-purchase opinion, which is what deceives buyers and the ranking system.
What is brushing?
Brushing is when a seller ships a cheap or empty parcel to a real address using an account they control, so the resulting order counts as a verified purchase and the fake review looks legitimate. It also inflates the seller's sales volume.
How do negative-review attacks work?
A competitor hires a broker to flood a rival with one-star reviews, often paired with false safety or policy complaints. The goal is to crater the rating or get the listing suspended while the platform investigates, costing the rival sales.
Why is it hard to detect one review at a time?
A single fake review looks plausible in isolation. The fraud shows up at the network level: shared devices, coordinated timing, reused text, and reimbursement trails, which is why link and cluster analysis beats reading reviews one by one.
Is review fraud illegal?
In many markets, yes. Consumer-protection regulators treat undisclosed and fake reviews as deceptive advertising, and platforms can face enforcement, so it is a compliance issue and not only a trust nuisance.
Review fraudと併せて知っておきたい用語
2026年 不正・AMLレポート
予測は不要です。このレポートは、不正・AMLチームが実際に直面していることと、その対応方法を分解して解説します。