Product review guide
How to spot a fake product review
No single phrase proves a review is fake. A stronger check compares the review’s specific claims, purchase context, account history and independent evidence.
Start with what the review actually claims
Separate opinion (“I loved it”) from measurable assertions about dimensions, ingredients, battery life, delivery, certification or performance. Check whether the reviewer identifies the exact model and describes details consistent with genuine use.
Look for context, not magic words
Very short praise, promotional language, repeated phrases and many reviews posted close together can be signals. They can also have innocent explanations. Writing style cannot establish whether a reviewer is real, paid, assisted by AI or mistaken.
Examine the account over time where the platform permits it. A varied history with product-specific detail can add context; a new account posting near-identical reviews deserves more scrutiny, but neither pattern proves authenticity.
Check claims outside the review
Use the manufacturer for specifications, but not as independent proof of performance. For safety, health, certification and comparative claims, look for the named regulator, standards body, test report or independent study. Confirm that evidence concerns the exact product, conditions and outcome in the review.
Compare patterns carefully
Search a distinctive sentence in quotation marks. Exact reuse across unrelated products may indicate templating or syndication. Common phrases such as “great value” reveal little. Reviews repeating one original source are not independent confirmation.
Make a proportionate decision
Describe what you found: unsupported performance claim, duplicated wording, missing purchase context, or inconsistent model details. Compare multiple independent sources, check the returns policy and avoid a high-risk purchase when important evidence remains unresolved.
Further guidance
Try a product review check
Check its specific claims and public wording
Paste a review or product claim. The result separates evidence from language signals and does not label the reviewer.