What is a Product Identifier?
Executive Summary
Every Digital Product Passport begins with a question that sounds trivial and almost never is: which product is this?
Organisations tend to arrive at passport programmes thinking about the visible parts. They discuss QR codes, resolvers, hosting, and which fields must be published. Those discussions matter, but they all sit on top of an assumption that is rarely tested: that the organisation can name any individual product, unambiguously, and agree with its suppliers, its customers and its regulators about what that name refers to. When that assumption fails, and it fails often, the passport programme stalls before a single record is published.
A product identifier is the mechanism that resolves the question. It is a value assigned to a product so that it can be referred to consistently, by more than one party, over time. It is not a description. It says nothing about what the product is made of or what it does. Its entire job is to point, precisely and durably, at one thing.
This article treats the product identifier as a general concept rather than as any particular scheme. A GTIN is a product identifier. So is a vehicle identification number, a medical device UDI, a serial number stamped into a pump housing, or an internal part number in an engineering system. They differ in scope, governance and recognition, but they answer the same question and obey the same underlying principles. Understanding those principles is more useful than memorising any one scheme, because different industries have arrived at different answers and a passport programme usually has to work with several at once.
The framework introduced below, the Product Identity Hierarchy, sets out seven layers that connect a physical object to a published passport. Its argument is simple and load bearing: every Digital Product Passport depends on a trusted product identity established before any digital information is exchanged.
- A product identifier is a value that points at a product. It is identity, not description. - Identity is the concept; the identifier is the implementation. GTIN is one implementation among many. - Identifiers must be unique within a defined scope, and that scope must be stated explicitly. - Internal identifiers are cheap and private; global identifiers cost governance and buy recognition. - Different identifiers answer different questions: which product, which production run, which individual unit. - A Digital Product Passport cannot be more reliable than the identifier it is attached to.
This is the fifth article in the Standards & Technology pillar and the fifteenth in the tieback Knowledge learning path. It sits beneath What is GS1? and What is GS1 Digital Link?, both of which assume the concept explained here.
Orders identity from product model down to batch and individual item, and shows what each level can prove.
Table of Contents
- Definition
- Why Every Product Needs an Identity
- The Product Identity Hierarchy
- Identity vs Description
- Permanent vs Dynamic Identification
- Internal vs Global Identifiers
- Common Types of Product Identifiers
- How Product Identifiers Support Digital Product Passports
- Benefits for Manufacturers
- Benefits for Supply Chains
- Benefits for Regulators
- Benefits for Consumers
- Product Identifier vs GTIN
- Product Identifier vs SKU
- Product Identifier vs Serial Number
- Product Identifier vs Batch Number
- Common Misconceptions
- Frequently Asked Questions
- Key Takeaways
- Related Articles
- Related Glossary Terms
- References
- About This Article
Definition
The fact of a product being a distinct, recognisable thing that can be referred to consistently by more than one party. Identity is a concept, not a value. It exists whether or not anyone has written it down.
A value assigned to a product in order to express its identity in systems, on labels and in documents. It is the practical implementation of identity: a string of characters governed by rules about how it is allocated, what it refers to, and how long it remains valid.
A published set of rules describing how identifiers are structured, allocated and interpreted, so that organisations who have never met can read each other’s identifiers without prior arrangement. GS1 identification keys and ISO identification schemes are examples.
The granularity at which identity is asserted: a product model, a production batch, or an individual unit. Choosing the wrong level is the most consequential and most commonly rushed decision in a passport programme.
A structured, maintained and accessible record of product information, reached through a data carrier linked to a unique product identifier. The identifier is what the passport hangs from.
Why Every Product Needs an Identity
Identity is what allows two parties to talk about the same thing without being in the same room.
Consider what happens without it. A manufacturer refers to a product by its engineering description. A distributor refers to it by a catalogue line. A retailer refers to it by a shelf label. A recycler refers to it by what it looks like on a sorting belt. A regulator refers to it by a model name printed in a declaration of conformity. Everyone is describing the same object and nobody can prove it. When something goes wrong, and safety recalls are the sharpest illustration, the cost of that ambiguity is measured in units that cannot be found and units that are withdrawn unnecessarily.
An identifier collapses that ambiguity into a single agreed value. It gives every party a shared anchor: this value, and only this value, refers to this product. Once that anchor exists, everything else becomes possible. Records can be joined across organisations. Compliance evidence can be tied to the thing it certifies. A scan can return information about a specific object rather than a generic page.
Three properties make an identifier fit for that job.
Uniqueness within a stated scope. Two different products must not share an identifier inside the boundary where it is used. A part number that is unique inside one factory is not unique across an industry, and pretending otherwise causes silent data collisions later.
Stability over time. An identifier that is reassigned to a different product, or changed when packaging changes, breaks every historical record that referred to it. Identifiers should be allocated on the assumption that they will outlive the systems that store them.
Interpretability by others. An identifier only creates value where the receiving party can recognise it. A value that is meaningful only inside one organisation cannot support a cross border regulatory obligation without translation.
Almost every organisation believes it identifies products reliably, because internally it does. The failure appears at the boundary: the ERP number differs from the catalogue code, which differs from the label, which differs from the supplier’s reference, and no single system holds the mapping. Passport programmes routinely discover this in month three, after the technology has been chosen.
The Product Identity Hierarchy
The framework below sets out the seven layers between a physical object and a published Digital Product Passport. Each layer depends entirely on the one beneath it. Weakness low in the hierarchy cannot be compensated for higher up, which is why passport programmes that begin at layer six or seven so often have to be restarted.
The object that exists in the world: a jacket, a pacemaker, a compressor, a car. It has no identifier of its own and no opinion about how it is described. Every layer above exists to make it referenceable.
The decision about what counts as one product: a model, a variant, a production batch, or an individual unit. This is an editorial and commercial judgement, taken before any value is issued, and it determines everything downstream.
The concrete value that expresses that identity: a number or code allocated under stated rules, printed, stored and exchanged. This is the layer people usually mean when they say “identifier”, and it is meaningless without layer two.
The published rules that make the value interpretable by strangers: how it is structured, who may allocate it, what it refers to and how long it is valid. Without a standard, an identifier is a private convention that stops at the organisation’s boundary.
The operational practices that keep identity true: allocation on product creation, marking in production, capture at goods receipt, retirement when a product is discontinued. An identifier that is not maintained decays into noise within a few product cycles.
The systems that use identity to move information: catalogues, resolvers, traceability platforms, marketplaces, customs declarations and repair networks. They join records across organisations by matching identifiers, and nothing else.
The maintained, accessible record of product information, reached through a data carrier bound to the identifier. It is the visible result of six layers of work, and it inherits every weakness in them.
The hierarchy’s argument: every Digital Product Passport depends on a trusted product identity established before any digital information is exchanged. Layers one to four are decided rarely and are expensive to revisit. Layer five is continuous work. Layers six and seven are expected to change repeatedly over a product’s life. Programmes fail when they treat layer seven as the starting point.
Used well, this framework is a question list rather than a diagram. Can we state what counts as one product? Can we produce the identifier for any unit on request? Can an outside party interpret it? Does a process exist that keeps it true? Only then does the passport conversation begin.
Identity vs Description
The most persistent confusion in product data is between identifying a product and describing it.
A description tells you what something is: a 60 centimetre stainless steel dishwasher, 45 decibels, manufactured in Slovakia, energy class B. Descriptions are useful, human readable and negotiable. They are also unstable. Every one of those attributes can change while the product remains, in commercial and regulatory terms, the same product. Descriptions are also not unique: thousands of distinct products share the description “black cotton t-shirt, medium”.
An identifier tells you which one. It carries no meaning and needs none. Its value comes precisely from its emptiness: because it says nothing, nothing about it needs to change when the product’s attributes change.
It is tempting to encode information in the identifier itself: a factory code, a material class, a year of manufacture. It feels efficient and it fails predictably. The factory closes, the material changes, the year rolls over, and now the identifier is either wrong or has to be reissued, which breaks history. Keep meaning in attributes, keep the identifier meaningless.
A cordless drill is described by the manufacturer as “18V brushless combi drill”, by the retailer as “Cordless Drill 18V with 2 batteries”, and by a repair network as “combi drill, model series 4”. All three descriptions are legitimate. None can be used to reliably join records. The identifier allocated to that model is the only value all three parties can agree refers to the same product.
Permanent vs Dynamic Identification
Not all identifiers are meant to last. Distinguishing the permanent from the temporary avoids a class of design error that is difficult to unwind.
Permanent identifiers are allocated once and are intended to remain valid for the life of the product, and usually beyond it. A model identifier, a vehicle identification number and a unit serial number are permanent in this sense. They are safe to print, stamp, engrave and cite in regulatory documents, because nothing in normal business operation should cause them to change.
Dynamic identifiers exist to describe a state rather than a thing. An order line reference, a shipment number, a warehouse location code and an inventory lot allocation are all identifiers, and all of them are expected to be superseded. They are operational conveniences.
Between the two sit qualifiers: values that narrow a permanent identifier to something more specific, such as a batch number or a serial number attached to a model identifier. A qualifier is permanent within its scope but meaningless without the identifier it qualifies.
Anything physically marked on a product should be permanent, because it cannot be recalled and corrected. Anything operational should live in systems where it can be superseded cheaply. Programmes get into trouble when they mark dynamic values on products, or when they treat a permanent identifier as reissuable because a system made it easy.
Internal vs Global Identifiers
Every organisation already has internal identifiers. The question is not whether to have them but where their authority ends.
Internal identifiers are allocated by an organisation for its own use: part numbers, SKUs, material codes, engineering references. They are free, immediately available, and shaped exactly to internal needs. Their weakness is boundary: they carry no guarantee of uniqueness outside the organisation and no obligation on anyone else to interpret them.
Global identifiers are allocated under a shared governance framework so that they are unique across all participating organisations and interpretable without prior arrangement. That property is not free: it requires membership of, or registration with, an issuing framework, and it imposes rules about when a new identifier must be allocated. What it buys is recognition by parties the organisation has never dealt with, which is precisely what cross border regulation and open supply chains require.
Neither type is superior. Mature organisations run both and maintain an explicit mapping between them: internal identifiers for engineering precision, global identifiers for anything crossing an organisational boundary. The failure mode is having both and not knowing which maps to which.
In most passport programmes, the hardest deliverable is not the passport. It is a trustworthy mapping between internal product records and the globally recognised identifiers published to the outside world. Build and own that mapping early; everything else depends on it.
Common Types of Product Identifiers
Different industries have arrived at different schemes, shaped by what their regulators and trading partners needed. The principles are constant; the implementations are not.
A consumer product. A pair of trainers is identified at variant level by a trade item number, so that colour and size are distinguishable at the point of sale. Individual pairs are usually not serialised, because the commercial and regulatory questions are answered at variant level.
A medical device. An infusion pump carries a unique device identifier combining a device identifier for the model with production identifiers for lot, serial and expiry. Regulation requires that an individual device can be traced to a patient, so identity must reach the unit.
An industrial machine. A compressor carries a manufacturer serial number engraved on a plate, because service history, spare part compatibility and safety inspections all attach to that one machine over a working life measured in decades.
A vehicle. A car carries a vehicle identification number structured under ISO 3779, unique to that single vehicle, used by manufacturers, registration authorities, insurers, recall systems and scrappage facilities alike.
Four industries, four schemes, one principle: identity is asserted at the level where the important questions are asked.
How Product Identifiers Support Digital Product Passports
A Digital Product Passport is a record about a product. To be a record about a product it must be bound to something that names that product unambiguously. The identifier is that binding, and it does four distinct jobs.
It is the key the passport is stored against. Whatever the underlying system, the passport record is retrieved by identifier. If the identifier is ambiguous, the retrieval is ambiguous.
It is what the data carrier carries. The QR code or other carrier on the product encodes the identifier, usually expressed as a structured web address under a standard such as GS1 Digital Link. The carrier is replaceable; the identifier inside it should not be.
It is what joins records across organisations. A supplier’s material declaration, a test laboratory’s certificate and a recycler’s treatment record become one coherent view only when all three cite the same identifier.
It is what determines granularity. Whether a passport describes a model, a batch or an individual unit is decided entirely by the identification level, not by the passport technology.
Regulation reflects this. Under Regulation (EU) 2024/1781, the ecodesign framework requires a data carrier linked to a unique product identifier, with the applicable delegated act specifying the identification requirements for each product group. The battery regulation takes the same approach for battery passports. In both cases the identifier is treated as a precondition, not an implementation detail.
Teams frequently build a passport at model level because it is easy, then discover that the delegated act, a recall obligation or a repair workflow requires batch or unit level identity. Retrofitting granularity means reissuing identifiers, remarking products and reconciling published records. Decide the level first, in writing, per product family.
Benefits for Manufacturers
Reliable identification changes what a manufacturer can say about its own products with confidence.
Product data stops being duplicated across systems, because there is one key to join on. Recalls narrow from “every unit of this model” to “the affected batches”, which is the difference between a manageable corrective action and a commercial event. Compliance evidence attaches to the thing it certifies, so a test certificate can be produced on demand rather than reconstructed. Warranty and returns become verifiable rather than argued. Counterfeit detection becomes possible at all, because a claimed identity can be checked against an authoritative record.
Benefits for Supply Chains
Supply chains are held together by matching, and matching is only as good as the identifiers being matched.
When trading partners share identifiers, goods receipt reconciles automatically, catalogue data synchronises without manual mapping, and customs declarations reference values that authorities recognise. Traceability becomes achievable across multiple tiers, because each participant can pass on an identity that the next participant can read. Where identifiers are private conventions, every organisational boundary requires a translation table, and every translation table is a place where data quietly goes wrong.
Benefits for Regulators
For market surveillance authorities, the identifier is what makes enforcement tractable.
It allows an authority to establish exactly which products a declaration of conformity covers, to scope a corrective action precisely, and to trace a non compliant product back to the responsible economic operator. It also makes cross border cooperation possible: an authority in one Member State can act on a finding from another because both are referring to the same value. Without shared identity, enforcement depends on model names and photographs, which is how unsafe products remain on sale in one market after being withdrawn in another.
Benefits for Consumers
Consumers rarely see identifiers and benefit from them constantly.
A scan that returns information about the actual product in hand, rather than a generic category page, is only possible because the carrier encodes a precise identity. Recall notices reach the people who own affected units. Repairability information, spare part availability and end of life instructions are specific rather than approximate. Authenticity can be checked. In each case the consumer experiences accuracy, and accuracy is a downstream property of identification.
Product Identifier vs GTIN
A product identifier is the general concept. A GTIN, the Global Trade Item Number, is one specific implementation of that concept, defined and governed by GS1. The relationship is category to instance: every GTIN is a product identifier, but not every product identifier is a GTIN.
GTIN is appropriate wherever products move through open trade and need to be recognised by parties with no prior relationship. Sectors with their own established schemes, such as automotive VINs or medical device UDI systems, are not deficient for using them; they are following the identification framework their regulators and partners specify. The correct question is never “is this a GTIN?” but “is this identifier unique, stable and interpretable by the parties who must read it?”.
Product Identifier vs SKU
A stock keeping unit is an internal operational code. It exists to manage inventory in one organisation’s systems and it is shaped entirely by that organisation’s needs. Two retailers selling identical products will have entirely different SKUs for them, and the same retailer may change a SKU during a system migration without anything about the product changing.
SKUs are entirely legitimate and should not be replaced. They should simply never be the value published in a passport or encoded in a public data carrier, because no external party can interpret them and nothing guarantees they will still exist next year.
Product Identifier vs Serial Number
A serial number identifies one individual unit. Used alone it is usually not sufficient, because serial numbers are typically unique only within a manufacturer’s own series; two manufacturers may issue the same serial number for entirely different machines. Combined with a product identifier, the pair becomes globally unambiguous: this model, this unit.
Serialisation is not automatically better. It is more expensive to produce and generates far more data to maintain. It is warranted when the important questions attach to individual units: high value assets, safety critical equipment, products with long service lives, or wherever regulation requires it.
Product Identifier vs Batch Number
A batch or lot number identifies a production run: a group of units made under the same conditions. It sits between model level and unit level identification and is often the most economical answer to recall and traceability obligations, because it narrows scope dramatically without requiring unit level marking.
Batch level identification is the norm in food, chemicals and pharmaceuticals, where the conditions and date of production determine safety. It is increasingly relevant to passports generally, because sustainability attributes such as material origin or recycled content frequently vary by production run rather than by model.
Common Misconceptions
“A barcode is a product identifier.” A barcode is a data carrier: a way of printing characters so a machine can read them. The identifier is the value inside it. The same identifier can appear in a linear barcode, a QR code or an RFID tag without changing.
“Our SKU is our product identifier.” It is an identifier, and it identifies a stock line inside one organisation. It is not recognised outside, and it is not stable enough to publish.
“We can generate identifiers ourselves.” Any organisation can generate values. What cannot be self generated is the guarantee that nobody else has used the same value, and the ability of a stranger’s system to interpret it. That guarantee is the entire product of an identification framework.
“Identifiers can be changed if we get them wrong.” They can be reissued, at the cost of breaking every record, label, certificate and published passport that referenced the old value. Treat allocation as irreversible.
“One identifier per product is enough.” Most products need several: an internal engineering reference, a globally recognised trade identifier, and often a batch or serial qualifier. The requirement is not fewer identifiers but a governed mapping between them.
“Identification is an IT project.” Allocation rules, granularity decisions and retirement policy are business governance. Systems implement those decisions; they cannot make them.
Frequently Asked Questions
What is the difference between product identity and a product identifier?
Identity is the concept of a product being a distinct, referenceable thing. The identifier is the value used to express that identity in systems and on labels. Identity is decided; identifiers are allocated. Getting the identity decision wrong cannot be fixed by choosing a better identifier scheme.
Do we need a GTIN for a Digital Product Passport?
Legislation specifies outcomes rather than vendors: a data carrier linked to a unique product identifier, built on open standards and interoperable. GS1 identification is a widely used and well documented route to that outcome, and in some sectors other schemes such as UDI or VIN are the established answer. The applicable delegated act for your product group is the document that decides.
At what level should we identify: model, batch or unit?
At the level where the important questions are asked about your products. If recalls, service history or authentication attach to individual units, identify at unit level. If manufacturing conditions matter, batch level is usually sufficient. If neither applies, model or variant level is proportionate. Decide per product family and record the reasoning.
Can a product have more than one identifier?
Yes, and most do. An internal part number, a trade identifier, a batch number and a serial number can all apply to the same object simultaneously. What matters is that the organisation maintains an authoritative mapping and knows which value is published externally.
What happens when a product changes?
It depends on whether the change alters what the product is. Corrections to attributes do not require a new identifier. Changes that make the product commercially or functionally different usually do. Identification frameworks publish allocation rules for exactly this question, and following them consistently is more important than the specific threshold chosen.
Can an identifier be reused after a product is discontinued?
It should not be. Historic records, certificates, published passports and archived transactions continue to reference it indefinitely. Reuse silently corrupts that history. Retirement should mean permanent retirement.
Where should a team start?
Establish what counts as one product for each product family, confirm that a value exists for it which is unique and stable, and determine which of your existing identifiers can be interpreted by outside parties. Only then consider carriers, resolvers and passport platforms.
Key Takeaways
- A product identifier is a value that points at a product. Identity is the concept; the identifier is the implementation. - Identifiers answer “which one”; descriptions answer “what is it like”. Never build description into an identifier. - Uniqueness is always relative to a stated scope. An identifier that is unique internally is not automatically unique externally. - Internal identifiers are free and private. Global identifiers cost governance and buy recognition by strangers. - Choose the identification level, model, batch or unit, before choosing any technology. It is the hardest decision to reverse. - GTIN, SKU, serial number and batch number are different tools answering different questions; most organisations need several and a governed mapping between them. - Every Digital Product Passport depends on a trusted product identity established before any digital information is exchanged. - Identification is business governance, not an IT deliverable. Systems implement the decision; they cannot make it.
Related Articles
- What is a GTIN?
- QR Codes vs GS1 Digital Link: What’s the Difference?
- What is GS1 Digital Link?
- What is GS1?
- How Product Data Moves Through the Supply Chain
- How Will Digital Product Passports Change Product Compliance?
- What is a Data Carrier?
- What is EPCIS?
Related Glossary Terms
Definitions of record for the terms used above live in the glossary.
- Product Identifier
- GS1
- GS1 Digital Link
- Data Carrier
- QR Code
- Digital Product Passport
- Product Data
- Product Traceability
- Product Lifecycle
- Economic Operator
- Market Surveillance
- Conformity Assessment
- Delegated Act
- ESPR
- Sustainability Data
- Circular Economy
References
- GS1 identification keys, the primary reference for identification concepts and key types: https://www.gs1.org/standards/id-keys
- GS1 GTIN, definition and allocation of the Global Trade Item Number: https://www.gs1.org/standards/id-keys/gtin
- GS1 GTIN management standard, the allocation rules governing when a new identifier is required: https://www.gs1.org/1/gtinrules/en/
- GS1 General Specifications, defining keys, attributes, carriers and application identifiers: https://www.gs1.org/standards/barcodes-epcrfid-id-keys/gs1-general-specifications
- GS1 Digital Link standard, expressing identifiers as structured web addresses: https://www.gs1.org/standards/gs1-digital-link
- GS1, the global standards organisation: https://www.gs1.org
- International Organization for Standardization, ISO 3779 on vehicle identification numbers: https://www.iso.org/standard/52200.html
- International Organization for Standardization, ISO/IEC 15459 on unique identification: https://www.iso.org/standard/54779.html
- International Organization for Standardization: https://www.iso.org
- Regulation (EU) 2024/1781 establishing a framework for the setting of ecodesign requirements for sustainable products, including the Digital Product Passport, unique product identifier and data carrier provisions, Official Journal of the European Union: https://eur-lex.europa.eu/eli/reg/2024/1781/oj
- Regulation (EU) 2023/1542 concerning batteries and waste batteries, including battery passport identifier provisions, Official Journal of the European Union: https://eur-lex.europa.eu/eli/reg/2023/1542/oj
- Regulation (EU) 2017/745 on medical devices, including the unique device identification system, Official Journal of the European Union: https://eur-lex.europa.eu/eli/reg/2017/745/oj
- European Commission, Ecodesign for Sustainable Products Regulation: https://commission.europa.eu/energy-climate-change-environment/standards-tools-and-labels/products-labelling-rules-and-requirements/ecodesign-sustainable-products-regulation_en
- EUR-Lex, official portal for European Union law: https://eur-lex.europa.eu
About This Article
tieback Knowledge is a continuously maintained reference library covering Digital Product Passports, product traceability, product compliance and related regulations. Articles are reviewed regularly as legislation, standards and implementation guidance evolve.
Related Docs
- Standards & Technology
- What is GS1?
- What is GS1 Digital Link?
- QR Codes vs GS1 Digital Link: What’s the Difference?
- What is a Digital Product Passport?
- How Does a Digital Product Passport Work?
- What is Product Master Data?
- How to Build a Digital Product Passport Implementation Roadmap
- Who Is Legally Responsible for a Digital Product Passport?