What Is a Fake Review?

A fake review is a customer rating or testimonial fabricated to mislead: praise for a product the writer never bought, an attack on a competitor's listing, or a five-star flood ordered by the thousand. Reviews are among the most trusted signals in online commerce — which is exactly why they attract manufactured input. The manipulation economy spans paid review farms, incentivized customers, and increasingly automation: bots operating fleets of fake accounts can seed thousands of reviews, complete with plausible purchase histories, at a cost no honest reputation-building can match.

How Fake Reviews Are Manufactured

The supply chain has distinct tiers. At the artisanal end sit brokered human reviews — sellers recruiting reviewers through groups and marketplaces, refunding purchases in exchange for stars. The industrial tier runs on account infrastructure: synthetic accounts aged and warmed to pass platform checks, automation to post at scale, and generative text to make each review unique enough to evade duplicate detection — the arrival of capable language models removed the last bottleneck, mass-producing fluent, varied, product-specific praise. Negative campaigns use the same machinery in reverse, review-bombing competitors. Platforms respond with purchase verification, velocity checks, and network analysis linking accounts that review in coordinated clusters.

The Cost of Corrupted Trust

For buyers, fake reviews convert a decision aid into a trap; for honest merchants, they tilt marketplaces toward whoever cheats best, since ranking algorithms amplify the manipulated signal. For platforms the stakes are becoming legal as well as reputational: consumer-protection regulators on both sides of the Atlantic now treat fake reviews as deceptive practice — the FTC's 2024 rule banning their purchase and sale, and EU consumer law requiring platforms to state whether reviews are verified, both put enforcement teeth behind what was once a moderation problem. A review section that cannot be trusted stops converting, and shoppers who feel deceived rarely distinguish between the platform and the fraudster.

Defending the Review Pipeline

Moderation that starts at the posted review starts too late; durable defense targets the infrastructure that mass manipulation requires. The account layer comes first — verification at registration that stops bot-driven signup removes the raw material of review fleets, and services such as CaptchaFox screen account creation and login invisibly, so building a thousand-account fleet stops being cheap. The submission layer adds gates where reviews enter: proof-of-purchase requirements, and bot checks on the review form itself that filter automated posting without burdening genuine customers. Content and graph analysis then handle what remains — linguistic fingerprints of generated text, timing bursts, and account clusters that always praise and attack together. No single layer suffices, but manipulation at scale needs all three layers to be cheap at once.