In short: The Seller Network automatically groups infringing accounts, sellers and websites that share contact details into clusters, so you can investigate and act on a whole network rather than one account at a time.
Understanding the Seller Network
Infringers rarely work through a single account. The Seller Network automatically groups connected infringing accounts, sellers and websites into unified clusters by linking fragmented digital identities through shared contact information.
This lets you identify and investigate entire networks of connected entities, rather than addressing individual accounts in isolation. Clusters include all accounts found across Corsearch's entire database, based on the linked contact information, so you get the full picture.
1. The Seller Network feed
Use the feed to see which networks are the biggest risk and where to start.
In Seller Network view, clusters are ranked by the number of potentially infringing posts, with the highest listed first. The view shows:
Search Bar & Filters: narrow results by date, geography, label and more.
Export: export all filtered clusters in Excel format.
Rank: the overall severity of the cluster, based on the number of posts associated with it.
Description: a brief description of the cluster and its infringing elements.
Posts: the total number of moderated and unmoderated posts in the cluster.
Infringement %: the percentage of infringing posts out of all posts in the cluster.
Moderation %: the percentage of moderated items in the cluster.
Contact Info: the number of contact information elements (e.g. emails, phone numbers).
Learn more: Contact Info Filter · How to Export Data
2. Viewing a cluster
Open a cluster to see every account, post, image and contact detail that connects the network.
Clicking a cluster opens it in a new window, showing:
ID/Name & Creation Date/Last Update: edit the name by clicking the pencil icon.
Seller Network Details: the number of posts and accounts/sellers in the cluster, the percentage of infringing posts, the percentage of moderated posts, and categories (e.g. apparel, accessories).
Tags: all tags applied to the cluster.
Images: all images in the cluster, including infringing ones.
Associated Accounts: all accounts/sellers that appear in the cluster.
Related Posts: all posts in the cluster, including infringing ones.
Connection Reasons: the contact information that links the cluster together.
Comments: type @username plus your message, click the arrow to submit, and use the paperclip icon to attach files or reference images.
From a cluster you can also Export all its Associated Accounts and Manually Add accounts to it.
Learn more: Accounts · How to Create/Add a Tag
3. How clustering works
Knowing how links are made helps you trust, and explain, why accounts sit in the same cluster.
Clustering uses a deterministic, rules-based mathematical model, not Machine Learning (ML) or probabilistic guessing. Every link between entities is based on concrete, shared contact data.
The system evaluates contact details (phone numbers, emails, social handles, WeChat/WhatsApp) scraped from account bios, post descriptions and images. It draws a direct link between any two accounts that share at least one identical contact value, regardless of platform; for example, a Facebook account and a TikTok account sharing the same phone number.
Note: Clusters are fully recalculated every Saturday, rescanning all historical and newly uploaded accounts.
The three-phase process
Discovering connections: the system evaluates all known sellers ("posters") and websites. If any two entities share identical, normalised contact information, a direct link is established.
Grouping the network: the system applies a transitive ("friend-of-a-friend") rule. For example, if Seller A is linked to Website B, and Website B is linked to Seller C, all three are grouped into one cluster.
Aggregation and profiling: once the cluster is formed, it's saved and aggregate statistics (total posts, moderation percentages) are calculated.
Data normalisation
To ensure accurate matching, contact data is standardised before comparison:
Phone numbers: must be longer than 5 digits (after excluding leading zeros) and can't be all zeros.
Emails: must conform to standard email formatting rules.
Social handles: must pass basic validity checks (e.g. can't be all zeros).
Cluster formation logic
Two entities are always linked by a shared, normalised contact point. What happens next depends on whether they're already in a cluster:
Entity A status | Entity B status | System action |
Not clustered | Not clustered | Creates a new cluster containing both Entity A and Entity B. |
Already clustered (Cluster X) | Not clustered | Adds the unclustered entity to the existing Cluster X. |
Already clustered (Cluster X) | Already clustered (Cluster Y) | Merges Cluster X and Cluster Y into one cluster (subject to the safety limits below). |
Learn more: Contact Info Filter
4. Safety limits
Safety limits stop unrelated sellers being grouped together by mistake.
Linking is an exact-match process, with no per-link confidence scores. Because the transitive rule could lead to "over-clustering" (linking unrelated entities through a generic customer service number or spam email), fixed safety limits are built in:
Shared contacts: if a single phone number or email is shared by more than 100 distinct entities, the system ignores it.
Cluster size: a single cluster can't exceed 100 members in total.
Contacts per seller or website: no more than 100 contact points are evaluated for a single seller or website.
Allowlisting: known, legitimate brand entities and websites are strictly excluded from clustering.
Learn more: Accounts
Quick reference
If you want to know… | Look at… |
Which networks to tackle first | 1. The Seller Network feed |
Who and what is in a cluster | 2. Viewing a cluster |
Why two accounts are linked | 3. How clustering works |
When clusters are updated | 3. How clustering works (every Saturday) |
Why a contact or cluster isn't growing | 4. Safety limits |