Exact-match rules see clean vendors. The network is what connects them.
What looked clean
Two suppliers. Separate legal entities. Different bank accounts, tax IDs, and phone numbers. Every exact-match duplicate rule passes.
Then you look at how they behave together: they always bid on the same tenders, rotate who wins, share directors and postal addresses, and submit quotations minutes apart. That is not two competitors. That is one economic circle with two badges— and rule-based analytics never draw the edge between them.
Never share. Yet they…
What rules check
Bank account
Tax ID
Phone
What the network shows
Always bid together
Alternate winning tenders
Share directors
Share postal addresses
Submit quotations within minutes
Why people should care
Sophisticated collusion avoids the fields your duplicate rules scan.
Shared bank accounts and identical tax IDs are still high-value signals— but many rings deliberately keep those clean. What they cannot hide as easily is relationship structure: co-bidding, win rotation, people, addresses, and bid timing. If your analytics only fire on exact equals, those rings stay invisible.
Five relationship edges a graph catches
Treat vendors (and related people or addresses) as nodes. Treat shared behaviour and attributes as edges. Dense clusters—especially with bid-rotation patterns—are the investigation queue.
-
Always bid together
Vendors that co-appear on the same tenders far more often than chance or category peers.
Business impact: Surfaces pairs that move as a pack, not as independent competitors.
-
Alternate winning tenders
Win/loss sequences that rotate between the same small set of suppliers over time.
Business impact: Flags managed competition that keeps prices high while looking fair.
-
Share directors / beneficial owners
Link vendor master and registry data where people sit on more than one “independent” bidder.
Business impact: Proves common control even when bank and tax IDs differ.
-
Share postal addresses
Same building, suite, or normalized address across supposedly unrelated suppliers.
Business impact: Connects entities that never share payment identifiers.
-
Quotations within minutes
Near-simultaneous bid submissions—or near-identical file metadata—across “separate” vendors.
Business impact: Exposes coordinated cover bidding that timing rules alone rarely catch.
Why rule-based analytics miss this
Exact-match rules
Same bank? Same tax ID? Same phone? If not, the vendors look independent.
Relationship graphs
Co-bidding, rotation, people, places, and timing— edges across tenders and master data, not one field at a time.
Clean identifiers do not mean independent suppliers. Build a graph instead of rule-based analytics.
How foretale.ai helps
foretale.ai builds vendor relationship networks across tender history, awards, vendor master, and related attributes—co-bidding, win rotation, shared people and addresses, and bid-timing clusters—with explainable evidence for every edge and community.
Procurement and audit teams investigate dense clusters—not a spreadsheet of exact-match duplicates.
How many “independent” bidders are actually one network?
Most companies don’t know—until they stop checking fields in isolation and start mapping relationships. Continuous AI analytics can surface vendor networks before the next award.
Request a demo