Find fakes before your customers do.

Watching thirty marketplaces

Counterfeit monitoring is a wide, low-volume crawl. Listing pages are light once assets are blocked:

Ten dollars a month covers the monitoring. The reason to use residential here is not price, it is that counterfeit sellers geo-target: a listing that a shopper in Brazil sees may be invisible from an office address in Berlin, and a hosting range often sees a sanitised catalogue.

Capture the evidence in the same session that found it

A takedown request needs the listing, the timestamp and the market it was visible from. Hold one session per market long enough to screenshot what you found:

Python: find and capture in one session

def market_session(country, session_id):
    login = f"login_c_{country}_s_{session_id}_ttl_15m"
    return f"http://{login}:password@proxy.sotaproxy.com:10000"

for i, market in enumerate(markets, start=1):
    proxy = market_session(market, i)
    for listing in scan(marketplace, proxy=proxy):
        if suspicious(listing):
            evidence = screenshot(listing.url, proxy=proxy)
            store(listing, evidence, market=market, seen_at=utcnow())

• Screenshot through the same address that found the listing. A capture taken from another market is not evidence that the listing was served there.
• Record the country and the timestamp with every capture. Marketplaces ask for both, and sellers change listings within hours of a complaint.
• Recheck after a takedown. The same seller usually relists within days under a new account, and only a repeat scan catches it.
• Scan the markets where you actually sell, plus the ones your distributors cover. Counterfeits follow distribution, not headquarters.

Why complaints get rejected