Counterfeit sellers are careful about the product photos and careless about the reviews, because reviews feel like someone else's problem. They are not. Bought reviews, transferred reviews, and review pools shared across seller accounts leave patterns that are harder to fake than a product photo, and those patterns connect listings that otherwise look unrelated.
Reviews are also where counterfeit operations are most vulnerable to scale. One seller can photograph one product convincingly. Faking a credible review history across dozens of listings requires coordination, and coordination leaves traces: the same reviewer accounts, the same phrasing, the same unnatural timing. Detection starts by reading reviews as data, not as opinions.
First, velocity spikes. A listing that accumulates hundreds of five-star reviews in days, especially right after launch, is not experiencing organic enthusiasm. Real review curves are slow and noisy. Look for the step function: flat, then vertical, then flat again. That shape is purchased.
Second, rating distribution inversion. Genuine products show a spread: mostly positive, with a tail of complaints about shipping, fit, or expectations. Counterfeit listings propped up by bought reviews often show an unnaturally clean distribution, nearly all five stars, with the few real reviews (the actual buyers) clustered at one star describing fakes. When the one-star reviews say "counterfeit" and the five-star reviews say "great!!!", believe the one-stars.
Third, text reuse. Bought reviews are written at scale, and scale means templates. Identical or near-identical phrasing across reviews, generic praise that could apply to any product, and reviewer accounts whose entire history is five-star reviews of unrelated products are the tells. Copy a suspicious phrase into a search and watch it appear on other listings. That is the network revealing itself.
The real value of review analysis is not flagging one listing; it is mapping the operation behind it. Reviewer accounts that praise the same cluster of sellers connect those sellers. Identical review text across storefronts suggests a shared review vendor. And when a seller account is shut down, watch where its reviewer pool migrates: the same accounts praising a "new" seller weeks later is the operation reincarnating.
Review timing adds another layer. Coordinated review campaigns often run in bursts aligned with product launches or sales events. Charting review dates across suspect listings frequently shows synchronized bursts, the fingerprint of a single campaign serving multiple storefronts. One burst is a marketing push. The same burst shape across ten sellers is an organization.
Build review checks into the regular monitoring cadence, not as a one-off investigation. For priority products, track review velocity weekly: sudden spikes get flagged the same way price anomalies do. Sample the text of new five-star reviews monthly and search distinctive phrases across the marketplace. Maintain a list of reviewer accounts that appear across your suspect sellers, and check new listings against it.
When packaging evidence, reviews are supporting material, not the case itself. A velocity chart plus text-overlap examples plus the underlying listing evidence (photos, pricing, seller history) makes a report that platforms act on. Reviews alone rarely trigger action. Reviews that corroborate a pattern of counterfeit behavior do. Read them as data, document what they show, and let them connect the listings your other signals already flagged.
Marketplaces have their own review fraud detection, and it catches the clumsy operations: the accounts reviewing fifty products a day, the obvious template text. What it misses is the patient kind, reviews dripped in slowly across many accounts over months, timed to look organic. That is exactly the kind a brand-side review monitoring program catches, because you are looking at your own products with context the platform does not have. Your catalog knowledge is the advantage. A reviewer praising "fast shipping" on a product that ships in three weeks is a signal only someone who knows the product would catch.