How Bots Scrape Sold Listings to Price Counterfeit Goods
Sold prices are the target, not list prices
A list price is a hope. A sold price is a fact, and facts are what counterfeiters need. Scrape a marketplace's completed sales and you learn what a handbag, a sneaker, or a watch actually changes hands for, in which condition, and in which size. That is the whole pricing model for a fake, delivered free, updated daily, and sorted by the marketplace itself.
The scrape is cheap because the data is public. Sold-item pages sit behind no login on most marketplaces, and a bot that walks them in order can harvest months of price history in an afternoon. The counterfeiter is not guessing at a price. They are reading the market's own ledger and picking a number just under the real one.
The pricing trick that makes a fake look real
Price a fake at half the real sold price and buyers get suspicious. Price it five or ten percent under the real sold price and it reads as a motivated seller, a quick sale, a bargain. The fake rides the real product's price history the way a remora rides a shark, close enough to look like part of the same animal.
This is why the fakes cluster just under the genuine sold price instead of far below it. The scraper's job is to find the exact line where a buyer stops asking questions. Every sold listing the bot reads sharpens that line a little more, and every seller whose history gets scraped is donating the data that prices the fake competing with them.
What sellers can actually see
You cannot see the scraper, but you can see its footprints. Traffic spikes on listings that stopped selling months ago. The same ranges of addresses viewing sold-item pages in sequence. Your own product photos turning up on new listings from sellers you have never heard of. None of these is proof on its own, and marketplaces are slow to act on any one of them, because scraping and shopping look the same from the server side.
The honest answer is that the defense sits with the platform, not the seller. Rate limits and bot detection can slow a harvest, and takedown programs like eBay VeRO can remove the fakes once they are posted. What a seller can do is narrower: watermark photos so a stolen image is traceable, report the fakes that appear, and stop treating sold listings as harmless public history. They are the price guide for the people copying you.
Keep reading: Fake Listings, Real Chargebacks: How Marketplaces Spot Counterfeit Sellers, Reseller Account Takeover: How One Login Drains a Marketplace Payout.
Frequently asked questions
Why do counterfeiters care about sold listings instead of active ones?
Active listings show asking prices, and asking prices are aspirational. Sold listings show completed transactions, the price a buyer actually paid. That number tells a counterfeiter the ceiling a buyer will accept, so the fake can be priced just under the real sold price and still look like a plausible deal.
How can I tell if bots are scraping my sold listings?
Watch for traffic spikes on old listings that no longer sell, repeated views of sold-item pages from the same ranges of addresses, and your product photos appearing on new listings from sellers you do not know. None of these proves scraping on its own, but together they are the pattern: someone is harvesting your price history, not shopping.
Does reporting a counterfeit listing remove the scraper?
Reporting the listing can remove that one fake, but it does nothing to the scraper that priced it. The scraper will price the next fake from the next sold listing it harvests. Stopping the scrape means making the data harder to harvest at scale, which is a platform job: rate limits, bot detection, and blocking the ranges that hammer sold-item pages.
What does eBay VeRO actually do for sellers?
VeRO is eBay's Verified Rights Owner program. A rights owner who joins can report listings that infringe their intellectual property, and eBay removes them. It is a takedown tool, not a scraping defense: it acts on the listing after the fake is already priced and posted, and it covers intellectual property, not price data.