ECLASS at scale: How WAGO classifies its product portfolio with AI

WAGO at a glance

WAGO GmbH & Co. KG, headquartered in Minden, is a family-owned company and has been one of the world's leading providers of electrical interconnection and automation technology since 1951. Its products are used globally across industrial, energy, process, and building technology as well as in mechanical and plant engineering, making WAGO the world market leader in spring clamp technology.

Today, the WAGO Group employs around 9,000 people - approximately 4,400 in Germany - and operates production and sales subsidiaries in more than 30 countries.

Before DQC

WAGO manages an extensive product portfolio with thousands of active items across connection technology, automation, and electronics (100,000+ products). As demand from industry and trade for standardized product data grew, ECLASS as an international classification standard moved increasingly into focus - not least because customers and trading partners are now routinely expecting structured data handoffs. Properly classified products mean more than mere compliance: they improve discoverability in catalogs and tenders, secure orders from major customers, and make product data fit for digital business models.

The challenge: ECLASS classification is complex. For every product, the right class must be identified, the relevant attributes populated, and consistency ensured across the entire portfolio. At WAGO's scale, doing this manually simply isn't viable - too time-consuming, too error-prone, too dependent on individual expert knowledge.

What was needed was a scalable approach that largely automates ECLASS classification, systematically leverages existing product knowledge, and still keeps domain experts in control. That's exactly where WAGO partnered with DQC.

Deep-dive

To handle ECLASS classification efficiently at this scale, WAGO turned to the DQC Platform - not as a replacement for domain expertise, but as an intelligent amplifier of it.

The Approach

• The entire product portfolio is reviewed - regardless of whether an ECLASS assignment already exists or not

• Based on the relevant attributes, the platform automatically suggests the appropriate ECLASS class, including a rationale and confidence score

• For products that cannot be unambiguously assigned, classification options are presented for WAGO's domain experts to make the final call

The ECLASS node is built as an agentic workflow: from the configured source columns, the agent generates a factual product description and searches the ECLASS taxonomy for matching classes via semantic similarity. Candidates are evaluated in hierarchical context. The output is the ECLASS class, a confidence score, and a rationale.

What sets this apart from a traditional rules engine: the system improves with every approval. It learns from WAGO's actual product reality - not from static rules alone. Over time, this creates a consistent classification logic tailored specifically to WAGO's portfolio.

Step 4: Automated ECLASS suggestions with confidence score

„For us, ECLASS is more than a data standard — it's a key to making our product portfolio internationally interoperable, digitally usable, and future-proof. With the DQC Platform, we bring speed, quality, and scalability into alignment.“

Simone Brinkmann-Tewes, Head of Digital Engineering Experience

With DQC

Using the DQC Platform, WAGO classifies its entire product portfolio efficiently and at scale - what would have taken roughly 12 months and €500,000 manually, WAGO now accomplishes in a matter of days at a fraction of the cost.

The approach is designed to last:

• Classify the full portfolio of 100,000+ products

• Stay current with ECLASS updates - automatically, no manual work required

• Classify new products as they're added, with no backlogs or gaps

Outlook

With the DQC Platform, WAGO has transformed ECLASS classification from a persistent manual burden into a scalable, continuous process - and stays automatically up to date with every new ECLASS version.

Data quality is not a one-time initiative, but a continuous process. With the right platform, it can be managed efficiently, consistently, and at scale.