Before the Passport: What Is Product Identity Actually Worth?

Reconciling the same product already costs time

A GS1 Global Data Synchronisation Network pilot at Geisinger Health, with Johnson & Johnson, Abbott and Teleflex supplying data, shows the cost clearly. Geisinger's previous process had taken 52 hours over six months to map and load 36 items. In the pilot, it could extract data from several suppliers through one source, match the items and assign GTINs in less than an hour.

That does not mean every business loses 52 hours on 36 products. It is one documented healthcare pilot, and the improvement came from the synchronisation network and process as well as the identifiers. It does show that translating the same product between organisations and systems is a real, measurable operating cost.

An identifier field is not the same as product identity

eMAG, a large marketplace, shows the opposite failure. After collecting GTINs without strong enough validation, it found 80,000 products carrying the same bogus GTIN and a further 200,000 identifier conflicts. That is one marketplace's case, not a measure of how common the problem is.

The lesson is not to add more barcodes. Identity only works when the identifier is correctly assigned, scoped to the right product level, validated and connected to the right product record.

Stable identity matters even when it is not a GTIN

Google Merchant Center tells merchants that if they change a product's ID, the product is treated like a new product. It has to be approved again, which can take up to three business days, and its past performance is no longer associated with it.

That applies to the ID a merchant assigns inside Google, not only to GTINs. GTIN may be the right external identifier for many products and channels, but the wider point is continuity: a stable reference keeps a product's history attached to it.

The costs of weak identity are spread across the business

Weak product identity creates work in many places at once:

1. retailer and marketplace onboarding;
2. mapping Shopify, ERP, PIM, retailer and supplier IDs;
3. getting model, variant and SKU levels wrong;
4. answering the same evidence request more than once;
5. updating one fact in several systems;
6. joining sales, media and product records;
7. recalls and corrections;
8. feeds for machine-mediated commerce;
9. repair, resale and other after-sale services;
10. Digital Product Passport preparation.

These costs overlap, so they should not be added into a single return figure. The business case has several variables, and each needs its own evidence. For what a passport programme itself costs in effort, see what a textile passport programme costs. For deciding which individual attributes repay the work, see which product facts actually change a business decision.

Identity plus workflow is where the large outcomes come from

The largest published results also show why the identifier should not get the credit alone. Across six NHS trusts in the Scan4Safety demonstrator programme, point-of-care identification and scanning released 140,000 hours of clinical time, produced recurrent inventory savings worth nearly £5 million and non-recurrent inventory reductions of £9 million. At Leeds Teaching Hospitals, the average time to recall a product fell from 8.33 days to less than 35 minutes.

Those results came from identification together with scanning, governed data, connected systems and changed ways of working. That caveat makes the argument stronger, not weaker. The proposition is not to buy a number. It is to make the product reliably identifiable, then connect the work that already happens around it.

Item-level history can also change what a product is worth later, but only under specific conditions. Whether a product's history can change its value owns that question.

What this means for a smaller business

One business may be well served by Shopify. Another may run Akeneo, an ERP or a custom stack. Another may keep a structured master sheet and an evidence folder.

The question is not whether the business owns a PIM. It is whether:

• each sellable product is identified at the right level;
• internal IDs can be mapped to each other;
• current facts are governed;
• evidence can be found;
• external systems can recognise the same product;
• future passport outputs can reuse those facts.

That is a smaller and more practical problem than replacing the stack.

How to think about the return

The value of connected identity is not a single saving from getting a code. A more honest model is:

connected identity net value = repeat work removed + avoidable loss reduced + value actually realised + future work reused − cost to connect − cost to keep connected

Each term needs its own evidence and rules against double counting. For a simple business with one channel and a small range, the early result can be negative. The case strengthens as channels, partners, product levels and after-sale services multiply.

The next question

If the value is real, do not start with the QR code. Start with the object: what exactly are you trying to identify?

Continue to your first product identifier.

Sources

Sources checked as at 17 September 2026.

• GS1, How the Global Data Synchronisation Network is helping hospitals improve patient care: the Geisinger pilot figures. One healthcare pilot; the result combines GDSN, process and GTIN assignment.
• GS1 UK, eMAG: managing marketplace growth with trusted data: the duplicate and conflicting GTIN figures. One marketplace case, not prevalence.
• Google Merchant Center Help, Tips to help your products stay approved: the effect of changing a product ID on approval and performance history. Google-specific.
• GS1 UK, Scan4Safety evidence report press release: the six-trust and Leeds recall figures. Whole-programme outcomes, not the effect of an identifier alone.