How do fake reviews get manufactured and planted on marketplace listings?

A review is bought long before it is written

Review brokers recruit account holders or operate account farms, place orders so the review carries a verified purchase badge, then arrange reimbursement plus a fee through channels the marketplace cannot see. The product ships, sometimes empty weight or a cheap substitute, and the review follows a script about quality and delivery.

Because the order is real in the marketplace records, the review survives checks that only look for purchases. The fraud sits in the side payment and the instruction, both of which happen off platform by design.

Why ratings move markets so quickly

Shoppers use star ratings as a shortcut under time pressure, and marketplace search ranking uses them as a quality signal. A product pushed from 3.8 to 4.5 stars gains both: more clicks from humans and more placement from the algorithm. That double effect is why a few dozen planted reviews can be worth thousands in sales.

The mirror tactic is the planted attack review. Honest competitors get hit with detailed one star stories that sound specific enough to believe, dragging their average down while the broker client rises. Both directions are the same business.

The patterns hiding inside review data

Fake reviews cluster in time, arrive in bursts after a broker campaign starts, and share phrasing habits across supposedly unrelated buyers. Reviewers in a network often review the same small set of unrelated products, because the broker reuses the same accounts across clients.

Language is a tell too. Planted praise tends to be generic about benefits and specific about nothing, or suspiciously focused on the exact keywords the seller wants to rank for. Real reviews complain about small things even when they are happy, and that unevenness is hard to script at scale.

Why marketplaces struggle to clean it up

Removing reviews at scale risks deleting genuine ones, and sellers appeal aggressively. Brokers also adapt: slower drip campaigns, varied text written by language models, and accounts with months of ordinary purchases before they ever post a paid review. Each cleanup teaches the network how it was caught.

The strongest marketplace signal is linkage, not language. Accounts, payment routes, devices, and broker client lists connect campaigns that look unrelated review by review, and acting on the network removes hundreds of planted reviews in one decision.

What honest sellers can do this quarter

Watch your category, not just your listing. A rival whose rating jumps in steps, whose newest reviewers review the same odd set of products, or whose praise arrives in bursts is showing you the campaign. Report the pattern with dates and reviewer overlap, not just a complaint about unfairness.

On your own listing, earn reviews the slow way and answer negative ones well. Shoppers read responses, and a seller who handles real criticism calmly reads as more trustworthy than a wall of perfect, identical praise.

Keep reading: How do price scraping bots help rival sellers undercut you within minutes?, How do fake seller accounts flood marketplaces with counterfeit listings?, and How do bots scalp limited marketplace deals before real buyers even see them? cover the nearby cases in this series.

Frequently asked questions