What Is PIM? Managing E-Commerce Product Data from a Single Source

What is PIM, and which fields belong where between the ERP and e-commerce? Product data model, attribute and variant design, data quality rules and an import checklist.
Technical summary
Technology
PIM, ERP, WooCommerce REST API, schema.org Product, GTIN
Topic
Modeling product data with PIM and publishing it to channels
Level
Intermediate
Prerequisites
  • Knowledge of the ERP product record structure
  • WooCommerce attribute and variation concepts
In this article8 sections

If the same product’s description says one thing in the ERP, another in the Excel catalog file and a third in the WooCommerce dashboard, the problem isn’t the software but data ownership. Product data isn’t just price and stock: technical attributes, variants, images, datasheets, translations, channel-specific descriptions. In a setup where this data is updated by hand in three separate places, inconsistency is inevitable. Below, we explain where a PIM (Product Information Management) system sits between the ERP and e-commerce, when it’s really needed and how to build the product data model.

The division of roles between PIM, ERP and e-commerce

The short answer to “what is PIM” is: the single source of truth for the rich data that describes “what the product is”. The ERP manages the commercial and operational side of the product (stock, cost, price, tax, procurement); the e-commerce platform presents the data to the customer and collects orders. The PIM takes ownership of content and attributes between the two.

Most of the problems we see in the field come from a field being editable in more than one system. The first job is to designate a single owner system for each field:

Data field Owner system Note
SKU, basic product record ERP Created only in the ERP
Stock and warehouse ERP / WMS Web only reads
Price and discount ERP Not moved to the PIM
Technical attributes PIM Typed fields with units
Description, title, translation PIM Can be channel-specific
Image, datasheet, certificate PIM / DAM Versioned file
SEO fields, URL E-commerce Channel-specific

This table can vary according to the organization’s processes; what mustn’t vary is that the decision is in writing and the integration follows this rule. For example, if the description field in WooCommerce is fed from the PIM, a manual fix made in the dashboard is overwritten at the next transfer. If the editor doesn’t know this, they make the same fix every week. Tell editors about this behavior clearly and, if possible, make the relevant fields read-only on the web side. The technical patterns for moving data between systems are covered separately in the ERP and web integration patterns article.

Signs that you need a PIM

Not every e-commerce project needs a PIM. If several of the following appear together, the cost of managing product data in a separate layer starts to fall below the cost of the mess:

  • Product information goes to more than one channel: the website, a B2B portal, marketplaces, a printed catalog, dealer lists.
  • Technical attributes are used for filtering and comparison and vary by product family.
  • The catalog is published in two or more languages.
  • Putting a new product on sale requires contributions from several teams (product management, marketing, translation, e-commerce), and who has completed what can’t be tracked.
  • Supplier data arrives in different formats and is edited by hand every time.
  • Custom fields have been opened in the ERP to add marketing copy and images, and the ERP can’t carry this load.

Conversely, for a single-channel, single-language catalog with a simple attribute structure, a well-defined transfer from the ERP to WooCommerce and a disciplined attribute structure within WooCommerce may be enough. PIM is first a data governance decision, then a software purchase. A PIM set up without defining the process becomes a third place where data sits scattered.

The product data model: attributes, variants and categories

Whichever PIM is chosen, the real work of the project is designing the product data model. The model has three basic components.

Attributes and attribute sets

Every attribute should have a data type: text, number and unit, selection list, multiple selection, yes/no, date or file. Keeping values like “Power: 2.2 kW” as free text makes filtering and unit conversion impossible. In a numeric attribute, the value and unit should be kept separately, and the unit should be chosen from a fixed list. Text can’t be filtered.

Attribute sets (also called families or templates) determine which fields are mandatory depending on the product type. A cable’s attributes aren’t the same as a server’s; defining sets by product family instead of a single giant form makes both data entry and quality control easier.

Variant axis

Variants are used only for axes the customer chooses between: color, size, capacity and the like. In WooCommerce, variable products are generated from attributes marked as variation axes, and each variation can have its own SKU, price and stock. The mistake we see most often in the field is setting up technically different products (models sold under different product codes, with different documents) as variations of one product; in that case images, documents and SEO content can’t be separated.

Category tree and classification

The category tree in the store menu is a navigation structure and changes with marketing needs. The product’s technical classification, however, determines the attribute set and is more permanent. Keep the two separate so the data model doesn’t break when the menu is reorganized. For aligning the category structure with the site’s information architecture, see the corporate website information architecture article.

The example below shows the structure of a normalized product record transferred from the PIM to a channel:

{
  "sku": "KBL-C6A-305",
  "family": "network_cable",
  "classification": "cable/copper/f-utp",
  "attributes": {
    "category_standard": "Cat6A",
    "conductor_gauge": { "value": 23, "unit": "AWG" },
    "jacket_material": "LSZH",
    "length": { "value": 305, "unit": "m" }
  },
  "locales": {
    "tr_TR": { "name": "Cat6A F/UTP LSZH Kablo 305 m" },
    "en_US": { "name": "Cat6A F/UTP LSZH Cable 305 m" }
  },
  "assets": ["kbl-c6a-305-front.jpg", "kbl-c6a-305-datasheet.pdf"],
  "status": { "completeness": { "web_tr": 100, "web_en": 80 } }
}

There’s no price or stock in the record. Deliberately; these go directly from the ERP to the channel. The key that matches the product across systems is the SKU, and it should never be reused in any system. Giving a discontinued product’s SKU to a new product mixes up past orders with the new catalog.

Data quality rules and the enrichment workflow

The value of a PIM comes from making it measurable when data is “ready to publish”. For this, completeness rules are defined per channel and language: if the Turkish name, short description, at least one image and the family’s mandatory technical attributes are filled in for the web, the product is complete for that channel. A product isn’t sent to a channel until completeness is achieved.

Rule type Example
Mandatory field The family’s technical attributes can’t be empty
Format The EAN/GTIN check digit is validated
Value list Color is chosen only from the defined list
Media The main image is at least the specified resolution
Consistency Variant axes are filled in for all variants

The enrichment workflow usually follows this order: the product record is opened in the ERP, and the SKU and basic information land in the PIM; product management completes the technical attributes; marketing adds the title, description and images; the translation team completes the other languages; after final approval, the product is published to the channel. The owner of each step and the pending work list should be visible. This is what turns a PIM from a data store into a workflow tool.

Publishing product data to channels

There are two principles on the publishing side: send only the data that has changed (delta) and define a separate mapping for each channel. Attribute names in the PIM and global attributes in WooCommerce (stored in WooCommerce as taxonomies with the pa_ prefix) don’t have to be exactly the same; the mapping table is kept in the integration.

The WooCommerce REST API offers the /wp-json/wc/v3/products/batch endpoint for bulk updates. The update request needs the product ID; so the integration must first find the SKU-to-ID mapping and cache it. Sending a SKU query on every update stretches a transfer of thousands of products over hours.

# Finding the product ID by SKU
curl -s -u "$WC_KEY:$WC_SECRET" \
  "https://store.example.com/wp-json/wc/v3/products?sku=KBL-C6A-305"

# Bulk update (by default at most 100 objects per request)
curl -s -u "$WC_KEY:$WC_SECRET" -X POST \
  -H "Content-Type: application/json" \
  "https://store.example.com/wp-json/wc/v3/products/batch" \
  -d '{"update":[{"id":1543,"short_description":"Cat6A F/UTP, LSZH jacket"}]}'

Images shouldn’t be re-uploaded on every transfer; the file’s hash should be stored and the media updated only when it changes. Otherwise the media library swells with copies of the same image and it’s noticed when the disk fills up. On the web side, product data is also converted into schema.org Product structured data and, if needed, product feeds; if the GTIN, brand and attributes are clean in the PIM, these outputs are consistent too. In stores using caching, the relevant product pages must be purged after a bulk update; the pitfalls here are explained in the WooCommerce cache errors article.

Product data import checklist

Whether setting up a PIM or moving an existing catalog to a new platform, check the following before a bulk import:

  • SKUs are unique and contain no spaces or invisible characters; a case rule has been set.
  • Each field’s owner system is in writing; the import writes only that system’s fields.
  • Attribute values are normalized: different spellings of the same value are merged, units are in a separate field.
  • Variant axes and the parent–variant relationship are defined; there are no orphan variants.
  • The category mapping table is ready; unmapped products are reported.
  • Image file names are linked to the SKU; a list of missing images has been produced.
  • Character encoding is UTF-8; Turkish characters and the decimal separator have been tested.
  • A trial was first run in the test environment with a small sample set, and the result was checked on screen.
  • Existing product URLs are preserved or a redirect list is ready.
  • A database backup was taken before the import and the rollback path has been defined.

The item usually skipped on the checklist is invisible characters: a trailing space on a SKU coming from Excel silently breaks the matching. For catalog migrations involving URL changes, also add the preventing SEO loss when migrating a website steps to the plan to preserve search visibility.

Mistakes encountered in practice

  • Moving price and stock into the PIM: Managing prices in two systems starts a debate about which one is right. Commercial data stays in the ERP.
  • Using the category tree as the data model: A menu change breaks the attribute sets.
  • Setup without governance: In a PIM set up without defining roles and approval steps, products pile up in “draft” status.
  • Integration that isn’t one-way: Trying to write fixes made in the web dashboard back to the PIM produces conflicts; the content flow should be designed to be one-way.

In projects where the product data model and the ERP–WooCommerce transfer need to be designed together, we plan the WooCommerce e-commerce and ERP and API integration work from the same data ownership table.

Frequently asked questions

Above how many products is a PIM needed?

What decides it isn’t the number of products but the complexity of the data. A catalog of a hundred products that is multilingual, multichannel and heavy on technical attributes can benefit from a PIM; a catalog of thousands of products that is single-channel and simple can be managed with a well-designed ERP transfer.

Which fields should be kept where between the PIM and the ERP?

Stock, cost, price, tax and accounting codes stay in the ERP. Marketing descriptions, technical attributes, images, documents and channel-specific content are kept in the PIM. The SKU must be the same in both systems and created in only one of them.

Can WooCommerce be used as a PIM?

In small, single-channel catalogs, WooCommerce can effectively be the source of product data. But since it doesn’t offer PIM capabilities such as workflow, field-level completeness rules, multichannel publishing and version history, this approach gets stuck as the number of catalogs or channels grows.

What’s the most common mistake in product data migration?

Transferring attribute values without normalizing them. Different spellings of the same value, units embedded in text and free-text fields used as filters break filtering and comparison functions after the migration.

Support

Contact us for support or a discovery session on this topic.

Contact form
[email protected]

Eretek Web

We build and maintain corporate websites, WordPress and CMS platforms, intranet and SharePoint portals, document management, and form and approval workflows, and keep them secure and up to date.

Knowledge Center

Other articles