Skip to content
All terms

Glossary

Seller Clustering

Seller Clustering: Linking Fake Sellers to One Network

Seller clustering is the technique of grouping seller accounts, online shops and listings that share common traits, such as contact details, payment data, images, wording or infrastructure, into clusters that likely belong to the same operator. In brand protection it helps brands move from removing single listings to dismantling whole counterfeit networks.

Counterfeiters rarely work through one account. When a listing is removed, the same people open new accounts and continue. Clustering shows that many of those accounts are one business.

What Is Seller Clustering?

Each seller leaves traces: the name and address given at registration, the email domain, the phone number, the way listings are written, the photos used, the shipping origin, the payment provider, even the layout of a shop website. Alone, these traces mean little. Several of them together suggest a connection.

Clustering software compares these traces across many sellers and platforms and groups together the ones that look related. The result is a map: this cluster contains 40 accounts on three marketplaces and 12 websites, all selling the same type of fake products.

Seller clustering is different from related ideas:

  • Infringement detection finds individual suspect items. Clustering links the sellers behind them.
  • Risk scoring ranks items by priority. A cluster can raise the score of every member, since one confirmed counterfeiter casts doubt on its linked accounts.
  • Threat intelligence is broader and covers the actors, methods and trends, while clustering applies the idea to specific sellers.

How It Works

Typical Linking Signals

  • Identity data: names, addresses, email addresses or phone numbers reused across accounts (where publicly visible).
  • Content similarity: the same product photos, near-identical titles and descriptions, copied from official images.
  • Shop structure: the same website template, checkout flow or text across several fake shops.
  • Technical traces: shared hosting, name servers or registrant details across lookalike domains.
  • Logistics: the same shipping origin, return address or tracking patterns.
  • Behaviour: accounts created at the same time, listing the same products, in the same order, at similar prices.

Example

A cosmetics company removes a counterfeit lotion listing, and a few days later a near-identical listing appears under another seller name. Clustering shows that 25 seller accounts use the same photos, identical product titles with the same typo, and a shared return address. Instead of 25 separate cases, the brand handles one operation, and can report it as a pattern to marketplaces and, where suitable, to authorities.

Why It Matters for Brands

  • Stops the whack-a-mole effect. Removing one listing at a time is slow. Acting on the network reduces re-listing.
  • Finds hidden volume. A network of small accounts can add up to large sales that look minor individually.
  • Stronger cases. Showing platforms a repeated pattern supports action under their repeat infringer policies and helps with account suspension.
  • Better use of resources. Evidence and reviewing effort are shared across the cluster.
  • Support for escalation. A documented network is a stronger basis for a cease and desist letter or a referral to law enforcement or customs.

Limits and Care

Clusters are hypotheses. Shared details can be coincidence: two unrelated sellers may use the same fulfilment warehouse or the same stock photo from a supplier. Wrongly linking an innocent seller to a counterfeit network would be a serious error, so each cluster should be reviewed and each member should have its own basis for action. Privacy rules also matter, so only lawfully available data should be used.

How to Act on a Cluster

  1. Confirm at least one member with a clear infringement, for example through a test purchase.
  2. Review other members individually before treating them as linked.
  3. Record the links and evidence in one case file.
  4. Report all confirmed members to each platform, and mention the connection.
  5. Monitor for new accounts that match the cluster pattern.

How Remove.tech Helps

Remove.tech detects infringing listings, shops and accounts across marketplaces, social media, websites and app stores, and looks at the connections between sellers so repeat offenders are not treated as unrelated cases. Confirmed infringements are reported through notices we send for you. See marketplace protection or get started.

Get started

Join Remove.tech,like other leading brands.

Or contact us by email: hello@remove.tech