Outside the four walls that scanned it, who does this tag actually belong to?

Every stocktake or cycle count now turns up EPCs that don't match the expected list. Resolving one today means a manual floor search — physically locating the item to work out whether it's genuinely misplaced stock, a unit that belongs to a neighboring store, or something that was never properly encoded at all.

• Replenishment runs on inventory counts that stay wrong until every variance is resolved by hand
• Store staff time goes into floor searches for units that, in many cases, were never theirs to find
• Unresolved variances quietly erode confidence in RFID accuracy itself — a cost the whole industry bears, not just the retailer holding the reader
• No single organization — retailer, brand, or tag issuer — can tell these cases apart on its own

The infrastructure to answer this question at scale did not fully exist until recently.

• GS1 published Tag Data Standard 2.3 in January 2026, giving RFID tags a resolvable web address by default — the addressing layer the industry needed is now standardized.
• GS1's EPCIS and Core Business Vocabulary standards already define how organizations describe and exchange event data about a tagged item — the vocabulary layer already exists.
• What hasn't existed is a cross-organization layer that resolves ownership using that vocabulary — every implementation to date has stayed inside one organization's own systems.
• That gap is open right now because most tag issuers and integrators haven't yet built anything on top of the new addressing standard.

1. 01

   A contributor — a tag issuer, integrator, retailer, or brand — donates anonymized issuance and scan-event records to the Repository.

2. 02

   After a stocktake or cycle count, a requesting party submits every EPC that didn't match its expected stock list — as a batch, not one at a time.

3. 03

   The Repository resolves each EPC against every contributor's records and returns one of three outcomes: yours (misplaced, reconcile it), theirs (often an RF cross-read from a neighboring store, ignore it), or unknown (not GS1-encoded or not yet contributed).

4. 04

   No contributor can see another contributor's raw data. The Repository only ever answers the questions it was asked.

• Held by an entity independent of any single tag manufacturer, retailer, or brand — including the organizations that seed it
• Seeded with anonymized issuance and event records donated by founding contributors, to give the Repository useful coverage from day one
• Access is compartmentalized by design: a contributor's raw records are never visible to another contributor, or to the Repository's own operators, beyond what a single query needs to answer
• Built on GS1 identification and event standards, so participation never requires licensing a competitor's proprietary technology
• Governance decisions — who joins, what a query can ask, how disputes are handled — sit with the independent entity's own board, not with any one contributor

Does contributing data expose our customers or commercial relationships to competitors?

No. Contributions are anonymized at the organization level, and access is compartmentalized — no contributor, including a founding contributor, can see another contributor's raw records.

Who actually controls the Repository, and can that change later?

An independent entity, governed by its own board — not by any single tag manufacturer, retailer, or brand. That structure is the basis for participation, not a phase-one arrangement subject to later revision by one contributor.

What if our EPCs aren't encoded to the GS1 SGTIN standard?

Two paths: share your own encoding logic so the Repository can resolve against it, or those EPCs stay outside what the Repository can answer for. Non-compliant EPCs are never a barrier to using the Repository for everything else.

What does participation cost?

Retail verification lookups are free. Collateral and authentication attestations — the paid tier — are priced per query, and that revenue funds the Repository, not the free lookups that build its coverage.

Why would a tag manufacturer join a repository a competitor helped start?

Because the alternative is a credibility gap that hurts every tag manufacturer's pitch for RFID, not only the founding contributor's. A shared, independently governed layer is a smaller ask than trusting one competitor's data directly — which is exactly why it isn't run by one.

Contactadmin@epcledger.orgRead the whitepaper (PDF)