Fake reviews

There are two directions. Inflating reviews praise a product or business, often written by owners, employees or paid reviewers, or offered in exchange for incentives without disclosure. Attacking reviews target a competitor or a public figure's business with low ratings, sometimes as part of a coordinated campaign such as brigading. A third form is the review that is real but posted for a purchase that never happened or written about an event unrelated to the business, common in review bombing after a news event.

Review platforms use automated systems and human moderation, and they publish policies that describe what they remove. Outside analysts look for the same patterns used for other coordinated activity: bursts of reviews in a short window, reviewers with a single review, near-duplicate text, ratings that do not match the written content, and clusters of reviewers who all reviewed the same set of businesses. See astroturfing for the broader pattern.

In the United States the Federal Trade Commission has published rules and guidance on fake reviews and undisclosed endorsements, and other jurisdictions have their own consumer protection rules. Requirements differ by country, and organizations should check current legal guidance rather than rely on a summary.

Reviews are a reputation surface that customers, and increasingly AI assistants, read. A surge of hostile reviews after a news event can look like a customer verdict. Deciding whether the surge is genuine, coordinated or unrelated to actual customers determines whether the response is a product fix, a platform report or no action. Keeping a record of review volume, dates and the reviewers' patterns supports a takedown request.

Teams also face internal risk. A company that pays for or writes its own reviews, or lets an agency do so, invites regulatory and press exposure. Clear vendor contracts and a review policy reduce it.

Teams can track new review volume, rating distribution and the phrases reviewers use, with alerts when volume departs from the normal pattern. When a spike appears, compare the reviews with sales and support records to see whether the reviewers are customers. Note the dates of each spike alongside external events, because a news story often explains a wave of reviews that have nothing to do with the business.

A negative review is not a fake review, and asking to remove it because it hurts is not a reason platforms accept. Similarly, a run of five-star reviews is not proof of fraud. What matters is evidence about authenticity, not the rating.