ETIM Classification Workflow for Distributors

Raw supplier feeds rarely arrive with correct ETIM class codes. A typical distributor onboarding a new industrial range receives a spreadsheet with free-text descriptions, a mix of proprietary attribute names, and units that may or may not match what the ETIM feature expects. The result: products stuck at the group level, mandatory features left blank, and a BMEcat export that a trading partner rejects on the first pass. Claro resolves product identity across those incoming feeds, enriches missing attributes from datasheets and structured sources, proposes ETIM classes with confidence scores, and writes clean EC codes and feature values back into your PIM or ERP — so the workflow below executes against trusted data instead of raw supplier chaos.

This playbook gives you a repeatable ETIM classification workflow that takes raw supplier products and produces correctly classed, attribute-complete records ready for BMEcat or ETIM xCHANGE export. Run it whenever you onboard a new supplier range, adopt a newer ETIM release, or remediate a catalog where class codes are missing, wrong, or stuck at the group level.

Before you start

You will need a product list with at least a description, manufacturer, and MPN; access to the ETIM release your trading partners expect (ETIM 9 or ETIM 10 are most common as of 2026); and a target schema for your PIM or data pool. Decide upfront whether ETIM is your only classification standard or one of several — if you also publish to retail or procurement channels, see the ETIM vs UNSPSC vs eClass comparison before you begin.

Before and after

The workflow

1. 1
   Lock the ETIM release and scope your classes

   Confirm which ETIM release each trading partner accepts, since class codes and feature sets change between versions. Pull the list of ETIM classes that apply to your range rather than the full dictionary. A distributor of fasteners will touch a few dozen classes; a broadline industrial supplier may touch several hundred. Export the candidate class list with their mandatory and optional features before you classify a single product — this is your target spec, not the full ETIM dictionary.

2. 2
   Resolve product identity across supplier feeds

   Before classification, confirm you are working on deduplicated, identity-resolved records. If a product appears in three supplier feeds with three slightly different descriptions, classify once and propagate — not three times independently. Claro’s entity resolution layer handles this step automatically, flagging records that refer to the same real-world product so you classify the entity, not the feed row.

3. 3
   Group products and assign candidate classes

   Cluster products by description, category, and MPN family so similar items get classed together. Assign a candidate ETIM class to each cluster — for example, a contactor range to EC000141 or a miniature circuit breaker range to EC000042. Classify the cluster, then spot-check outliers. This is where the majority of time savings come from — avoid item-by-item classification from scratch.

4. 4
   Map source attributes to ETIM features

   For each class, map your incoming supplier columns to the ETIM feature codes (EF codes) and value lists. Pole count, rated current, IP rating, and material each map to a specific feature with a defined data type — numeric, logical, range, or a controlled value (EV) list. Record where each value came from so you can trace it later. See the schema mapping guide for the mapping discipline that makes this durable across supplier updates.

5. 5
   Classify and normalize values to ETIM value lists

   Apply the classification and convert raw values into ETIM-valid forms: numerics into the feature’s expected unit, free-text into the correct EV value, and booleans into logical features. Units deserve special attention — a cable cross-section in AWG must become mm² where the feature demands it, and a torque value must match the declared unit. Use the AWG to mm² converter for cable work. Reject any value that has no valid target rather than forcing it through.

6. 6
   Validate completeness and correctness

   Check that every product carries a valid leaf-level EC code, that all mandatory features for that class are populated, and that values sit inside their allowed lists or ranges. Flag products that fall back to a parent group — these need human review. Treat unit mismatches and out-of-list values as hard failures. The ETIM classification checker runs this check before you generate the export.

7. 7
   Export, run a partner check, and write back

   Generate the BMEcat or ETIM xCHANGE export and run it through a structural validator before sending. See how to validate an ETIM XML export for the exact sequence. Resolve any errors, then write the approved EC codes and EF values back into your PIM or master record with their provenance. The next release cycle then starts from clean data — not raw supplier files.

Common pitfalls

Other recurring problems: mixing ETIM releases inside one export so feature codes do not resolve; ignoring controlled value lists and shipping free-text where an EV code is required; and dropping units, which silently corrupts numeric features. A distributor sending cable dimensions in inches against a feature defined in millimetres will pass schema checks but ship wrong data. Finally, classifying without recording data provenance means every future re-classification cycle restarts from zero — capture the source of each value the first time.

Related

FAQ