Data quality and standards in Actualog
Good product data is not only a nicer product page. It affects sales, procurement, engineering, service, compliance, integrations, and AI-assisted discovery.
Actualog treats product data quality as an operational process. Quality is defined through category models, product profiles, validation, review, approval, publication, and audit.
Key idea
A product record can be filled in and still be poor quality.
For example, it may use the wrong unit, duplicate another product, describe a variant inconsistently, contain an outdated certificate, or be complete for one catalog but not ready for another channel.
Product-data quality dimensions
Use these dimensions when reviewing a product, category template, import, or catalog.
Completeness
Are the required attributes, documents, media, names, descriptions, and relationships present?
Validity
Do values follow the expected type, format, unit, allowed range, and controlled list?
Consistency
Are the same concepts represented consistently across products, variants, languages, catalogs, and source systems?
Uniqueness
Does each real product have the right identity, without unnecessary duplicates or merged records that should stay separate?
For products, exact identity and duplicate prevention use the category identity policy and the stored Product Identity Hash; a generated product name is a readable label, not the uniqueness key.
Currency
Is the information still current, or has the source system, document, category template, certification, or product lifecycle changed?
Traceability
Can users see who created or changed the information, when it changed, and what was reviewed or approved?
Comparability
Can products in the same category be filtered and compared through compatible attributes and units?
Channel readiness
Does the product contain the information required for the intended catalog, customer, procurement process, integration, export, or AI interface?
How Actualog supports quality work
Category templates
Categories define reusable expectations:
- attributes;
- data types;
- measures and units;
- controlled values;
- required and optional information;
- product-family and variant rules;
- naming and identity rules;
- readiness requirements.
Product Profiles 360°
Product information is stored as governed data instead of being scattered across PDFs, spreadsheets, and free-text descriptions.
Learn about Product Profile 360°
Validation and quality signals
Actualog can identify missing, invalid, inconsistent, or stale information according to the configured product model and workflow.
Review, approval, and audit
Quality work remains connected to users, roles, decisions, and history. This helps teams understand whether a value is only proposed, reviewed, approved, or published.
AI-assisted quality work
Hex can help classify products, map imported fields, normalize units and values, draft translations, enrich content, and identify gaps. AI suggestions remain subject to human review.
Expert communities
Experts can improve shared categories, terminology, templates, reference values, and definitions that several companies reuse.
Relevant standards and classification systems
Actualog is designed to support standards-oriented product-data work. Alignment with a standard does not automatically mean certification or compliance for every implementation.
ISO 8000 — data quality
The ISO 8000 family addresses data quality concepts, master-data quality, and repeatable quality management.
Related Actualog practices include:
- clear data requirements;
- measurable quality checks;
- stewardship and review;
- traceability;
- continuous improvement.
ISO 22745 — open technical dictionaries
ISO 22745 describes an approach to technical dictionaries and structured item descriptions through classes and properties.
Related Actualog concepts include:
- reusable attribute definitions;
- controlled terminology;
- multilingual terms and definitions;
- units and allowed values;
- categories and templates.
ISO 29002 — exchange of characteristic data
ISO 29002 addresses models and exchange approaches for characteristic data and terminology.
Related Actualog concepts include:
- typed attribute values;
- class-and-property descriptions;
- dictionary-friendly terminology;
- import, export, and integration mapping.
eCl@ss and UNSPSC
Classification systems such as eCl@ss and UNSPSC can provide external reference structures or mappings.
They do not replace the need for an operational category template that fits the organization’s actual products, variants, catalogs, and workflows.
Alignment is not certification
Actualog can support standards-oriented structures and governance. That does not automatically mean:
- the software is formally certified to a specific standard;
- every customer implementation is compliant;
- every export satisfies an industry requirement;
- internal policy, evidence, and review are no longer required.
Formal compliance depends on the exact requirement, configured data model, operating process, evidence, and integration behavior.
Use “aligned with principles” unless formal conformance has been established for the specific claim.
Data quality for AI systems
AI cannot make fragmented product information trustworthy merely by rewriting it.
Reliable AI use requires:
- stable product identity;
- defined attributes;
- consistent units;
- controlled values;
- evidence and source context;
- current documents;
- review and approval;
- clear publication boundaries.
Actualog’s approach is to improve the structure and governance underneath AI, then use AI to accelerate work on that governed foundation.
AI assists. People govern.