Supply Chain
DPP Data Quality Management
Missing, wrong or stale data is the biggest risk in a passport. Quality dimensions, validation rules and how to mark data quality inside the passport.
The moment a Digital Product Passport is published it becomes a public declaration. A wrong value is riskier than a missing one: missing data is non-compliance, wrong data is a misleading statement.
Six quality dimensions
| Dimension | Question | Check |
|---|---|---|
| Completeness | Are mandatory fields filled? | Field-level requirement |
| Accuracy | Does the value reflect reality? | Compare with documents and measurements |
| Consistency | Do fields contradict each other? | Do shares total 100%? Are dates ordered? |
| Currency | How old is the data? | Age thresholds and renewal alerts |
| Traceability | Where did it come from? | Source reference |
| Evidence | Is there documentation? | Certificate and report links |
Automated pre-publication checks
This is the cheapest quality gain available. Typical rules:
- Material shares total 100% (with tolerance).
- Recycled content ≤ total material quantity.
- Certificate validity date > publication date.
- Country code is a valid ISO 3166 value.
- Carbon value within a plausible range for the category.
- Mandatory fields populated per required language.
Written once, these run on every publication. See passport lifecycle management.
How to mark "unknown"
Leaving a field empty is not the same as declaring it unknown. Define three states:
- Unknown: not yet collected; the supplier process is running.
- Not applicable: the field is meaningless for this product.
- Not yet required: the underlying obligation has not started (some battery carbon fields, for instance).
In an audit, this distinction is the most concrete evidence of good faith.
Provenance
Record beside every critical field: source type (supplier declaration, measurement, database), document number, date and owner. It makes both audits and data refresh far easier. See primary vs secondary data.
The refresh cycle
Data goes stale. Set simple rules:
- Certificates: alert 60 days before expiry.
- Carbon data: annual review.
- Supplier declarations: at contract renewal.
- Material composition: at every formulation change.
Frequently asked questions
Who should do quality control?
The person entering data and the person approving it should be different. Even in small teams, that separation measurably reduces errors. See DPP project roles.
I published wrong data.
Publish a new version and do not delete the old one — when the correction happened matters in an audit. See audit trails and versioning.
How do I verify supplier data?
Document checks plus consistency checks are enough in most cases. Add independent testing or site audits for high-risk items.
