Skip to content
P2Lab AI

Extract measures

Extract Measures reads the description (and any other fields you include) and fills the physical-property fields Shopware uses for shipping and packaging: weight, width, height, length, the scale unit, purchase and reference units, and pack unit labels.

Output fields

Section titled “Output fields”

The model populates these product fields:

  • weight — numeric, in the shop’s default mass unit
  • width, height, length — numeric, in the shop’s default length unit
  • unitId — Shopware unit reference (resolved by name, e.g. “piece”, “kg”)
  • purchaseUnit, referenceUnit — numeric pack sizing
  • packUnit, packUnitPlural — text labels shown in the storefront (“bottle” / “bottles”)

How to use

Section titled “How to use”
  1. Open a product → Specifications tab (Measures & packaging card).
  2. Click the AI assistant button in the card header. A modal titled AI assistant — Measures & packaging opens with these options:
    • Profile — pick a saved inline profile (optional)
    • AI connection — the model that will run the request
    • Read measures from — choose which fields and related data the AI reads. The description is used by default; you can add the name and related data such as categories, category path, tags, manufacturer and existing properties.
    • Fields to generate — which measure fields the AI should fill; all are selected by default
    • Custom instructions — free-text instructions (optional), e.g. “Weights given in lb should be converted to kg”
  3. Click Analyze. The modal expands with a grid of proposed values: weight, width, height, length, unit, purchase unit, reference unit, pack unit, pack unit (plural), plus the model’s reasoning. Anything the model couldn’t infer is left blank.
  4. Click Show sources to see which sentence in the description triggered each value.
  5. Edit any field you want, then click Apply to form. Save the product as usual.

If the product has variants, turn on Apply to variants to run the analysis across the selected variants as a background job and review the result per variant. Auto-apply matched measures writes the proposed measures automatically during that background run instead of waiting for review. See Working with variants.

In bulk

Section titled “In bulk”

extract_measures is also a job type. From AI Assistant → Jobs → New job, pick Extract measures, filter the products and run it. Each product gets one suggestion you can review and apply on the jobs detail page.

Configuration

Section titled “Configuration”
  • Custom instructions — pass extra hints like “Weights given in lb should be converted to kg” or “Ignore packaging size and only use net values”.

Tips & gotchas

Section titled “Tips & gotchas”
Section titled “Related”