QR code, NFC or RFID: which data carrier for the digital product passport?
The EU's digital product passport needs a physical link between garment and data. A practical comparison of QR codes, Data Matrix, NFC and RFID, and what the ESPR actually requires.
KEY TAKEAWAYS Summary by the editors
- Under the ESPR, a digital product passport must be connected through a data carrier to a persistent unique product identifier, and the carrier must be physically present on the product, its packaging or accompanying documentation.
- The ESPR recitals name the QR code and watermarks as examples of data carriers, and GS1 considers the QR code highly likely to be approved for its DPP application standard because smartphones read it almost universally.
- RAIN (UHF) RFID is useful for supply chain and store operations, but smartphones do not natively read it, so it is unlikely to serve as the only consumer-facing carrier.
- NFC is natively read by many smartphones and lets a consumer isolate one tag, but GS1 does not currently list it as an approved carrier.
- Textile-specific rules will come in a delegated act that, according to the Commission's working plan as summarised by Carbonfact, is planned for 2027, so brands should keep data carrier choices flexible.
For most fashion brands, a QR code on the care label or a sewn-in label is the most likely primary data carrier for the EU digital product passport (DPP), because the ESPR names it as an example and almost every smartphone can read it. RFID tags are better suited to logistics and stores, and NFC can add a convenient tap experience. The final choice for textiles will depend on the delegated act, which is not yet adopted.
What does the ESPR require from a data carrier?
The Ecodesign for Sustainable Products Regulation, Regulation (EU) 2024/1781, entered into force on 18 July 2024. As GS1 summarises Article 10, a product passport must be connected through a data carrier to a persistent unique product identifier; the carrier must be physically present on the product, its packaging or accompanying documentation, as set out in the product-specific delegated act; carrier and identifier must comply with listed or equivalent standards; and passport data must be based on open standards and transferable without vendor lock-in.
The recitals add context. They mention a watermark or a quick response (QR) code as examples of data carriers, state that carriers and identifiers should follow internationally recognised standards, and envisage a Commission registry of unique identifiers to which customs authorities have access. GS1's paper notes that Annex III refers to the ISO/IEC 15459 series for unique identifiers.
GS1's own conclusion is that all ISO-compliant, internationally standardised data carriers should qualify, provided the Article 10 conditions are met, while sector delegated acts may add specific requirements.
How do QR, Data Matrix, NFC and RFID compare?
| Carrier | Consumer access | Main strengths | Main limitations |
|---|---|---|---|
| QR code | Read by smartphone cameras almost universally | Low cost, printable on labels, cited in the ESPR recitals, can carry a GS1 Digital Link web address | Needs space and must survive washing; one code per label |
| Data Matrix | Uneven smartphone camera support | Smaller than a QR code at the same error correction, suits small items | Consumer readability not yet assured |
| NFC | Read natively by many smartphones with a tap | Consumer can isolate the intended tag; can carry a web address | Not currently an approved GS1 carrier; higher unit cost than print |
| RAIN (UHF) RFID | Not read natively by smartphones | Bulk reading in warehouses and stores; inventory and loss prevention | Not practical as the only consumer-facing carrier |
GS1 estimates that the DPP data to be encoded is around 40 to 70 characters, essentially an identifier and a link, since the passport content itself sits online rather than on the tag.
Why is the QR code the likely default for fashion?
Three reasons stand out. First, consumer reach: GS1 notes that smartphone cameras support QR codes almost universally. Second, regulatory signal: the ESPR recitals use the QR code as an example. Third, cost and process fit: a QR code can be printed on existing care or brand labels without new hardware in factories. Carbonfact's guidance for textiles also expects the carrier to sit on the physical label and the passport to be managed at product level.
The practical challenges are durability and design. A printed code must remain readable after repeated washing, and the label it sits on must stay attached for the life of the garment if the passport is meant to support repair, resale and recycling.
Where do RFID and NFC fit?
Many fashion retailers already use RAIN RFID for inventory accuracy. GS1 explains that RFID can carry a Digital Link address in binary form, but phones cannot read it natively and it is unclear how a phone would isolate one tag among many. That makes RFID a strong complement for logistics, stores and sorting facilities, but a weak sole consumer channel.
NFC is the opposite. Many smartphones read it natively and the user can target a single item. GS1 says it is not currently an approved GS1 carrier and would be assessed only if requested by its standards community or required by regulation. Some premium brands may use NFC for authentication alongside a printed code.
How can AI support DPP data carriers?
AI does not choose the carrier, but it can make the data behind it usable. Typical applications include:
- Extracting composition, origin and certificate data from supplier documents into structured passport fields.
- Checking passport records for missing or inconsistent values before codes are printed.
- Generating consumer-friendly summaries and translations of passport content from structured data.
- Detecting duplicate or reused identifiers across products.
All of these depend on clean product master data. A well-designed carrier linked to an incomplete record does not meet the purpose of the regulation. Brands should also decide early at which level they identify products, for example per style, per colour and size, or per individual item, because that choice affects label production, costs and what the passport can later show about an individual garment's history.
When will the textile requirements be known?
According to Carbonfact's summary of the Commission's 2025 to 2030 working plan, apparel and textiles are a top-priority product group, with the delegated act planned for adoption in 2027 and a transition period of at least 18 months set in the act itself. Brands therefore have time, but not much, to test carriers in production.
What should brands do now?
- Decide on a unique identifier scheme compliant with ISO/IEC 15459, such as a GS1 identifier with Digital Link.
- Pilot a QR code on care labels and test readability after repeated washing.
- Map where RFID is already used and whether the same identifier can be encoded.
- Assess NFC only for categories where authentication or premium experience justifies the cost.
- Build the passport data pipeline from suppliers, since the carrier is only a pointer to that data.
Frequently asked questions
Does the digital product passport have to be a QR code?
Not necessarily. The ESPR requires a data carrier linked to a unique product identifier and following recognised standards, and it mentions the QR code as an example. The specific carrier for textiles will be set in the textile delegated act.
Can RFID tags be used for the DPP?
RAIN RFID can carry the identifier and is widely used in fashion logistics, but smartphones cannot read it natively. It is therefore best seen as a complement to a consumer-readable carrier such as a QR code.
Where should the data carrier be placed on a garment?
The ESPR requires the carrier to be physically present on the product, its packaging or accompanying documentation, as defined in the delegated act. For textiles, guidance so far points to the garment label, which must stay readable over the product's life.
When do fashion brands need a digital product passport?
The textile delegated act is planned for 2027 under the Commission's working plan, with at least 18 months for implementation, according to Carbonfact. The exact dates will be fixed in the act.
One edition every weekday morning. Read in five minutes. Free for industry professionals.
