Most DPP readiness assessments are questionnaires. They ask whether you have a sustainability strategy, score you out of five, and produce a chart. That tells you nothing you can act on. The only useful test is run against your actual product records, field by field. It takes an afternoon. Below are the twelve checks we run on a client catalogue. Each one names the data it tests and what a fail looks like.
What DPP readiness actually means
The Ecodesign for Sustainable Products Regulation (Regulation (EU) 2024/1781) entered into force on 18 July 2024. It sets almost no product requirements by itself. The actual rules arrive later, product group by product group, through delegated acts. The Commission’s first working plan (COM(2025) 187) was adopted on 16 April 2025. It gives indicative adoption years: iron and steel in 2026, textiles, tyres and aluminium in 2027, furniture in 2028, mattresses in 2029.
So nobody is failing an audit today. What you are testing is whether your product data could carry a passport when the delegated act for your category lands. That gap is usually two to four years wide, and it is exactly the amount of time a catalogue remediation takes. Our digital product passport work almost always starts here, because the answer changes the size of the programme by an order of magnitude.
One more thing worth saying plainly. This is a data problem before it is a compliance problem. The legal text is short. The work of getting composition, origin and evidence into structured fields against half a million SKUs is not.
Before you start: pull three exports
You cannot run this audit against a slide deck. Pull three things first.
- A full item export from your PIM or ERP for one representative category. Every attribute, not the ones marketing uses. Two to five thousand SKUs is plenty.
- The supplier data as it arrived, raw, for your five largest suppliers in that category. Spreadsheets, XML, PDFs, whatever it was.
- Access to wherever declarations of conformity, test reports and safety data sheets are kept.
Every check below runs against one of those three. If you cannot produce export one in under an hour, that is a finding in itself. Write it down and carry on.
1. A persistent, unique product identifier
The regulation requires the passport to connect through a data carrier to a persistent unique product identifier. Persistent is the load-bearing word.
The check. In your item export, count blank GTINs. Then look for GTIN reuse: any identifier attached to more than one product record over time. That usually happens after a discontinuation and a relaunch.
A fail looks like. More than two per cent of sellable items without a GTIN. Or GTINs recycled onto successor products. Or, most commonly, the GTINs living in a finance system and being re-keyed into the PIM by hand.
2. One agreed level of granularity
This is the check that catches most people, and it is the one nobody expects. The DPP registry implementing regulation (Commission Implementing Regulation (EU) 2026/1778) requires registration at the granularity specified by the applicable law: model, batch or item. Where you register at item level, you also have to link the batch and model identifiers.
The check. Take twenty products. For each one, say out loud whether your master record is a model, a batch or an individual item. Then check whether the record structure agrees with you.
A fail looks like. No batch concept anywhere in the estate. Or a “product” record that is really a model, with size and colour held as free text rather than as structured variants. If you cannot express three levels, you cannot register at the level a delegated act asks for.
3. The legal manufacturer, as a record not a string
ESPR distinguishes the economic operator from the brand. Your data usually does not.
The check. In your export, look at the brand or manufacturer column. Is it a controlled reference to a supplier record, with a legal entity name and an identifier behind it? Or is it typed text?
A fail looks like. “Bosch”, “BOSCH”, “Bosch Power Tools” and “Robert Bosch Ltd” all present in the same column. Own-brand items where the actual manufacturer is not recorded at all, only the retail brand.
4. The facility, not just the country
Delegated acts are expected to ask where a product was made, not just which country it came from. The regulation already provides for unique facility identifiers.
The check. Pick ten SKUs from a single supplier. Can you name the plant each was made in?
A fail looks like. Country of origin only, held for customs purposes, and out of date. Or the plant is known to procurement, verbally, and recorded nowhere.
5. Composition as numbers, not prose
Every draft DPP dataset we have seen wants material composition expressed numerically. The JRC’s preparatory report for textiles was published in 2026 and has not been adopted. It proposes fibre composition, recycled content and organic content as structured values, with chain of custody evidence behind them.
The check. Search your long descriptions for percentage signs and material names. Count how many products carry composition data that only exists in the marketing copy.
A fail looks like. “Made from 80% recycled polyester and 20% elastane” sitting in a description field, with no corresponding attributes. That sentence is not data. It cannot be validated, aggregated or published to a passport.
6. Substances of concern with location and concentration
This is the hardest field group in the whole exercise, and the one with the longest lead time on supplier engagement. The textiles proposal asks for substances of concern identified, located within the article, and quantified by concentration. Identification alone is not enough.
The check. For twenty SKUs, try to answer: which substance, in which component, at what concentration.
A fail looks like. A safety data sheet PDF attached at supplier level rather than SKU level. Or a yes/no compliance flag with no underlying detail. Almost everyone fails this one. The useful output is knowing how far you are from the answer, not pretending you are close.
7. Units of measure that survive an export
The check. Filter your numeric attributes and look for anything alphabetic. Then export to CSV and look again, because unit handling often breaks on the way out.
A fail looks like. “1.5m”, “1500mm” and “1.5” in the same column. Weights in kilograms for one supplier and pounds for another, with no unit reference field. If you have already done the work on product attributes properly this check passes in five minutes. If it does not pass, treat it as a prerequisite rather than a DPP task.
8. Commodity codes that agree with your taxonomy
The registry stores commodity codes alongside identifiers, and customs authorities will check against them. That makes classification a compliance field, not a logistics field.
The check. Join your PIM export to your ERP commodity codes on SKU. Count the mismatches and the blanks.
A fail looks like. Commodity codes maintained only in the ERP, disagreeing with the PIM taxonomy node in over five per cent of cases. We see this in almost every distributor catalogue. Classification and taxonomy and attribution work is normally the first thing that has to be fixed, because everything downstream inherits from it.
9. Supplier coverage of the fields you do not own
You will not author composition, footprint or facility data yourself. Your suppliers hold it. The question is what proportion they currently send.
The check. Take your five largest suppliers in the category and list the DPP-relevant fields from checks 4, 5, 6 and 10. Count how many of those fields arrive in the supplier’s standard feed, as data, in any format.
A fail looks like. Under thirty per cent coverage, which is typical. Or coverage that technically exists but arrives as PDF datasheets that a person has to read. That is not a feed. Fixing it is a supplier data onboarding programme with a twelve to eighteen month runway. That is why this check matters more than the other eleven.
10. Evidence documents attached to the SKU
The check. Pick ten SKUs. For each, find the declaration of conformity and the most recent test report. Time yourself.
A fail looks like. More than two minutes per SKU. Documents filed by supplier and date on a shared drive rather than linked to the product record. No version or expiry tracking, so nobody knows whether the certificate on file is current.
11. A change history you can query by date
The passport has to stay available for at least the expected lifetime of the product. That means the record has to be reconstructable, not just current.
The check. Ask your PIM administrator to show you what a given SKU record said eighteen months ago.
A fail looks like. Silence. Most catalogues overwrite in place. Attribute-level change history with a timestamp and a user is a configuration decision. It is far cheaper to switch on now than to backfill later.
12. Audience-specific publication
ESPR sets different access rights for different actors. The public sees one subset. Recyclers, repairers and market surveillance authorities see more.
The check. Ask whether you can publish three field subsets of one product record to three endpoints, from one source, without a manual export.
A fail looks like. One flat feed, all fields or nothing, built for the website. If your only publication route is a nightly full export to ecommerce, you have a channel problem as well as a data problem. This is where PIM solutions for DPP earn their place. The access layer is genuinely hard to bolt on afterwards.
Scoring your DPP readiness
Score each check pass, partial or fail. Do not weight them. Then read the pattern rather than the total.
- Nine or more passes. Your data model is sound. The remaining work is field population, which is scoping and effort rather than redesign.
- Five to eight passes. Normal for a mid-market distributor. The gaps are usually granularity, facility and substances of concern. Budget twelve to eighteen months.
- Four or fewer. The DPP is not your first problem. Fix identifiers, units and classification first, because none of the rest can be built on top of them.
One deliberate asymmetry: checks 2, 9 and 12 are structural. Failing those three costs more than failing the other nine combined. They change the shape of the system, not the contents of a field.
What DPP readiness does not require yet
Being honest about scope keeps the programme fundable. As of August 2026, no ESPR delegated act for textiles, furniture, tyres, aluminium or mattresses has been adopted. Product-specific data lists are proposals, not law. You do not need to pick a DPP platform this year. Be wary of anyone selling you one against a deadline nobody has set.
You do need identifiers, granularity, classification and supplier coverage, because those take years and are useful regardless. Everything in this audit is work you would want done anyway. That is the test of whether a compliance programme is honest. If the remediation would be wasted should the regulation slip, it is the wrong remediation. Our view on this sits in more detail in the DPP fundamentals material.
Key takeaways
- ESPR is in force, but the product requirements come later through delegated acts, with indicative adoption from 2026 to 2029 depending on category.
- Run the audit against three real exports: item data, raw supplier data, and the document store.
- Granularity (model, batch, item) is the check that most catalogues fail and the one that costs most to fix.
- Substances of concern with location and concentration is the longest lead time item, because the data sits with your suppliers.
- Structural failures on granularity, supplier coverage and audience-specific publication matter more than any individual missing field.
- Nothing in this audit is wasted effort if the timetable moves.
If you want a second pair of eyes, we run this audit as a fixed-scope exercise against a live catalogue. You get the scored findings back with a remediation sequence. A thirty minute call is enough to tell whether it is worth doing. Get in touch, or start with the digital product passport overview if you are still working out what is in scope for your categories.