Brand protection · Counterfeit listings
Counterfeit listing detection on marketplaces and social commerce, per market
Counterfeit detection is a funnel: sweep marketplaces and social shops in each market for your marks, score the listings that look wrong, and turn the ones you confirm into a notice the platform can act on. Every step runs on public pages, from an exit in the market the listing targets.
Pay per GB · No subscription · Prices in EUR, excl. VAT
- €2.00
- per GB on ARC Fast, excl. VAT
- −44%
- at 1 TB (€1.12 / GB)
- 185+
- countries, city targeting
- Never
- expires — no subscription
Written for
- Brand protection and IP legal teams
- Anti-counterfeiting analysts
- Brand enforcement agencies
- Marketplace trust and safety liaisons
The short answer
- Detection and enforcement are different jobs: sweep broadly and cheaply, then capture only the listings you will report.
- Score listings on signals you can store: price far below market, wording that avoids your mark, mismatched photos, new sellers with stock in volume, and shipping from unexpected locations.
- Under the EU Digital Services Act, applicable to all platforms since 17 February 2024, a notice should include the reasons, the exact electronic location such as the URL, your name and email, and a good-faith statement.
- The OECD and EUIPO estimate global trade in counterfeit goods at about USD 467 billion in 2021, 2.3% of world imports.
The problem
What goes wrong without the right IPs
Listings target one market
Counterfeit sellers often list for a specific country, with local language, currency and shipping. A sweep from your head office sees a different set of results, or none.
Keyword search misses the obvious
Sellers avoid the brand name in titles, using misspellings, "style" or "inspired" wording, or only the product type with your photos. Searching for the mark alone finds the careless ones.
Notices bounce for missing details
A notice without the exact listing URL, a clear reason or the notifier's details is easy for a platform to set aside. Reconstructing them weeks later is slow and sometimes impossible.
Listings vanish and reappear
Removed listings return under a new seller name or a new listing ID. Without a fingerprint of what was removed — photos, wording, price — each comeback is treated as new.
Images are necessary and expensive
Unlike price monitoring, detection often needs the photos. On per-GB billing, loading every image of every search result is the largest line in the budget.
What to collect
What you can collect
- Listing URL and listing ID
- The exact electronic location. A notice under the DSA is built around it.
- Platform, market and exit country
- Where the listing was found and from which market it was viewed.
- Title and description text
- As shown. Mark misspellings and evasive wording are detection signals.
- Price and currency
- A price far below your lowest authorised price is one of the strongest single signals.
- Image hashes
- Perceptual hashes of listing photos, to spot your own product shots reused and to match comebacks.
- Seller identity and trader details
- Display name plus the contact and register details EU marketplaces must show for traders.
- Ships-from location and stock claims
- Large stated quantities and unexpected shipping origins raise the score.
- Risk score and reasons
- Your scoring output, with the signals that triggered it, so a reviewer can confirm or reject quickly.
- Notice status
- Not reported, reported (date, reference), actioned, reappeared.
How it runs
How a collection run works
- 01
Sweep search results per market
Search each marketplace and public social shop for your marks, common misspellings and generic product terms, from a residential exit in the market. Collect result pages with images blocked; you need titles, prices, sellers and thumbnails' URLs, not the pictures yet.
Rotating sweep, images fetched later and only for candidates13 linesimport httpxGATEWAY = "proxy.arcproxies.io:8000"PASSWORD = "PASSWORD"def sweep(country: str, url: str) -> str: user = f"USERNAME_pool-arcfast_country-{country}" proxy = f"http://{user}:{PASSWORD}@{GATEWAY}" with httpx.Client(proxy=proxy, timeout=30) as c: return c.get(url, headers={"Accept": "text/html"}).textresults = sweep("it", "https://marketplace.example/it/search?q=brand+bag")# parse titles, prices, sellers and image URLs; score; queue candidates - 02
Score, then fetch images for candidates only
Score each result on price, wording, seller age and shipping signals. Fetch photos only for listings above a threshold, hash them, and compare them against your own product shots and previously reported listings.
- 03
Review before reporting
A person confirms each candidate. Automated scoring finds listings; it does not decide what is counterfeit, and a wrong notice against a genuine reseller is a problem of its own.
- 04
Capture the listing as evidence
For each confirmed listing, take one capture in a sticky session: exit check, full page, screenshot, HTML hash and UTC time. The parent brand protection page sets out the capture record.
- 05
File the notice and track the outcome
Submit through the platform's notice mechanism with the fields it asks for. Record the reference, recheck the URL after the platform's response, and keep the image hashes to catch the listing coming back.
Setup
Recommended setup
Detection is about seeing what consumers in each market see, so residential is the default. Mobile is for social commerce viewed on phones; evidence captures go on the pool where failure is least acceptable.
| Plan | Role | Session | Targeting | Price |
|---|---|---|---|---|
ARC Fast Rotating residential | Default for sweeps and candidate pages Residential exits in each target market, at the steepest volume ladder we sell. Sweeps are the bulk of the bandwidth. | Rotating for sweeps; sticky per capture | Country + city | €2.00 / GB €1.12 at 1 TB (−44%) |
ARC Ultra Premium residential | For evidence captures that must not fail Premium residential for the confirmed listings that become notices, where a failed capture means waiting for the next sweep. | Sticky, 10 minutes per capture | Country + city | €3.50 / GB €2.17 at 1 TB (−38%) |
ARC Mobile Mobile 4G/5G | For social commerce on carrier networks Genuine 4G/5G exits for public social-shop pages where the mobile view is what you need. Country only, four markets. | Sticky, 5–60 minutes | Country (DE, GB, US, JP) | €3.50 / GB €3.15 at 1 TB (−10%) |
Prices are list prices in EUR, excl. VAT. Volume discounts apply automatically at checkout — full ladder on the pricing page.
Connect
One connection string per market
The country, city and sticky session are tokens in the username. Change the token, change the market; keep the session handle, keep the IP.
http://USERNAME_pool-arcfast_country-de_session-k3f9a1x2_sesstime-10:PASSWORD@proxy.arcproxies.io:8000Swap _country-de for any market you quote. The full grammar is on the residential proxies page.
Illustration
What daily counterfeit sweep in 6 markets costs
An illustration with round numbers — your pages, payloads and retry rate will differ. The arithmetic is the real price ladder.
Assumptions
- Search queries per market
- 150
- Result pages per query
- 3 at 350 KB, images blocked
- Candidates with images fetched
- 200 a day at 2 MB
- Markets
- 6
- Days
- 30
- Bandwidth per month
- ≈ 41 GB
On ARC Fast
€72.16per month, excl. VAT
41 GB bought in one purchase at €1.76 / GB (−12% volume rung, applied automatically).
The same 41 GB on ARC Ultra: €127.72 at €3.12 / GB. Route only the jobs that need it through the pricier pool.
Candidate image fetches are about 12 GB of this; loading every result's images instead would multiply the total several times. Evidence captures add a few MB each.
Bandwidth never expires, so a bigger single purchase is only ever cheaper per GB. See the ladder.
In depth
Counterfeit listings, in depth
Signals that separate counterfeit listings from genuine ones
No single signal proves a listing is counterfeit. Several together make it worth a reviewer's minute, and a stored set of reasons makes that minute short.
| Signal | How to measure it | Weight |
|---|---|---|
| Price far below market | Listing price vs your lowest authorised price in that market | High |
| Your own product photos | Perceptual hash match against your image library | High when combined with low price |
| Evasive wording | Misspelt marks, "style", "inspired", mark only in images | Medium |
| New seller, high stock | Seller first-seen date vs stated quantity | Medium |
| Unexpected shipping origin | Ships-from vs your distribution footprint | Medium |
| Comeback of a removed listing | Image hash or text match with a reported listing | High |
Genuine but unauthorised sellers trip some of these signals too. The sibling page on unauthorised resellers covers that case, which is handled differently.
Notice and action under the EU Digital Services Act
The Digital Services Act, Regulation (EU) 2022/2065, has applied to all online intermediaries in the EU since 17 February 2024. Its Article 16 requires hosting services, which includes marketplaces, to provide easy-to-access electronic mechanisms for anyone to notify them of specific items they consider illegal content, which covers listings of illegal goods.
The mechanism must make it possible to submit a notice containing a sufficiently substantiated explanation of why the item is illegal, a clear indication of its exact electronic location such as the exact URL, the notifier's name and email address, and a statement of good-faith belief that the notice is accurate and complete. A notice that lets a diligent provider identify the illegality without a detailed legal examination gives the platform actual knowledge of it.
Article 22 adds trusted flaggers: entities with particular expertise, independent of any platform, awarded the status by the Digital Services Coordinator of their Member State. Platforms must process their notices with priority. Article 30 obliges marketplaces to collect and verify traders' identity details before letting them sell, which is what makes the seller fields in your record fillable. How these apply to a specific listing is a question for your legal team.
From listing to notice: the record to keep
Build the record so the notice writes itself. Everything Article 16 asks for, and everything a reviewer at the platform will want, should be in one row with its files.
- Exact URL and listing ID — the electronic location, as loaded.
- Reasons — the confirmed signals, in words a non-specialist can follow, and the registration of the mark concerned.
- Capture — screenshot, HTML hash, exit IP and country, UTC time, session handle.
- Seller details — as displayed on the listing and the seller profile.
- Notifier details and good-faith statement — from your team, not from the pipeline.
- Outcome — reference, response date, decision, and any reappearance.
How large the problem is
The OECD and EUIPO's 2025 study of trade in fakes estimated global trade in counterfeit goods at about USD 467 billion in 2021, equal to 2.3% of world imports. It is an estimate of trade in physical goods, not a count of online listings, but it gives the scale behind the listings a detection pipeline finds.
For a monitoring budget the useful conclusion is narrower: sweeps are cheap per market on per-GB billing, images are not, and the review step is the real constraint. Size the sweep to what your team can review each week.
Works with anything that takes a proxy URL
- Python requests / httpx
Playwright- Puppeteer
Crawlee- n8n
- Docker
FAQ
Questions
Something not covered? Write to contact@arcproxies.io.
How do I report a counterfeit listing in the EU?
Through the platform's notice mechanism, which the Digital Services Act requires. Include why the listing is illegal, its exact URL, your name and email, and a statement that the notice is accurate and complete in good faith. Keep a capture of the listing with the exit, time and HTML hash. This is a description, not legal advice.
Why use residential IPs for counterfeit detection?
Because counterfeit listings often target one market's consumers, and some pages differ by the kind of address asking. A residential exit in the target market gets the results a shopper there gets. Data-centre ranges can be served a reduced page or nothing.
Do I need to download every listing image?
No. Sweep result pages with images blocked, score listings on text, price and seller signals, and fetch photos only for candidates above your threshold. That keeps image bandwidth to a fraction of the total.
Can I monitor social-commerce shops?
Public, logged-out shop and product pages, yes. ARC Mobile adds carrier IPs in Germany, the UK, the US and Japan where the mobile-network view matters. Private groups, messages and anything behind a login are out of scope.
What is a trusted flagger?
Under Article 22 of the Digital Services Act, an entity with particular expertise in detecting illegal content, independent of any platform, that has been awarded the status by its Member State's Digital Services Coordinator. Platforms must handle its notices with priority.
How much bandwidth does daily counterfeit detection use?
In the example above, 150 queries in each of six markets, three result pages each, plus 200 candidates a day with images, is about 40 GB a month. Loading images for every result instead of candidates only would multiply it.
Sources & review
Last reviewed by the ARC Proxies team. Prices are read from the live catalog at build time.
- [1]EUR-Lex: Regulation (EU) 2022/2065 (Digital Services Act), Articles 16, 22 and 30
- [2]European Commission: Digital Services Act starts applying to all online platforms in the EU
- [3]European Commission: The impact of the Digital Services Act on digital platforms
- [4]OECD: Global trade in fake goods reached USD 467 billion, posing risks to consumer safety and compromising intellectual property (May 2025)
Brand protection
More on brand protection
Get started
Buy a gigabyte and try it on your own targets.
No subscription, no sales call. Code WELCOME15 takes 15% off your first purchase, and whatever you do not use is still there next quarter.