How do bots flood marketplace question sections with fake buying advice?

The last honest looking space on the page

Reviews are now suspected by default, so shoppers moved their trust to the question and answer section, where answers appear to come from past buyers with no reason to lie. That trust is exactly the asset being farmed. A planted answer costs almost nothing and sits under the buy button influencing decisions for months.

Operators answer the questions shoppers actually ask before purchase: does it fit, does it work with this model, is the seller reliable. Each planted answer is written to close the sale in a chosen direction, either toward a listing the operator controls or away from one they compete with.

What planted answers try to achieve

The positive play steers compatibility and quality questions toward a copycat or inferior variant, often by claiming personal use the account never had. The negative play is quieter: sowing doubt about a rival seller delivery, authenticity, or warranty, in language careful enough to read as a helpful warning rather than an attack.

A third play buries truth. When real buyers post answers warning about a defect or a scam variant, bot answers flood in afterward, pushing the warning below the visible fold and diluting it with cheerful contradictions until a scanning shopper sees disagreement instead of a signal.

How the accounts behind answers are built

Answering accounts rarely start answering on day one. They are warmed with purchases, reviews, and innocuous answers across many categories, so the profile that eventually praises a copycat looks like a busy, ordinary buyer. Volume is spread across hundreds of such accounts, each answering rarely enough to look human.

The text itself is generated to vary: different lengths, small errors, personal details rotated from templates. Moderation tuned to catch identical spam misses a conversation where every voice is synthetic but unique, which is the point of spending on generation rather than copy and paste.

Signals a seller or shopper can use

Check the answerer history. Planted accounts answer across unrelated categories with the same confident tone, praise products with no purchase pattern that fits, and appear in clusters on the same listings within short windows. Genuine buyers answer narrowly, about what they own, and often admit what they do not know.

Content tells too. Planted answers push a decision: buy this variant, avoid that seller, ignore the warning. Honest answers describe an experience, including the inconvenient parts. An answer section where nobody is ever uncertain has usually been curated by someone selling something.

Defenses that protect the section without closing it

Marketplaces should tie answer weight to verified purchase of the exact product, label unverified answers plainly, and rank answers by helpfulness votes from verified buyers rather than recency, which is the ranking flood attacks exploit. Sudden answer velocity on one listing deserves the same alerting as sudden review velocity.

Sellers can monitor their question sections weekly, answer factual questions officially before bots do, and report planted clusters as one documented pattern. A question section with a fast, factual seller presence is both better for shoppers and far harder to farm.

Keep reading: How do bots fake wishlist and watchlist demand on marketplaces?, How do bots run fake flash-sale storefronts that copy marketplace listings?, and How do bots exploit marketplace price-matching guarantees? cover the nearby cases in this series.

Frequently asked questions