New in Endora Commerce: a built-in CRM with sales opportunities See it →
Endora Commerce

Pimcore PIM integration

Pimcore does not impose a ready-made product model: you build it yourself from classes, variants and attributes. Endora Commerce receives the records Pimcore sends and shows them to buyers without retyping between systems.

Pimcore in a company with a large technical catalog

Pimcore is a data management platform where PIM sits next to MDM and DAM. A product there is a data object, and its structure is set by a class definition designed by the implementation team. That is why Pimcore is common among manufacturers and distributors whose product page carries hundreds of parameters and whose every product group has a different set of properties.

For a shop integration this has one important consequence: there is no “standard Pimcore product”. Two installations can describe the same bearing with entirely different classes and fields. An integration with Endora Commerce therefore starts with a review of your data model, not with flipping a switch.

What moves from Pimcore to the shop

Data flows in one direction: Pimcore stays the source of truth about the product and the shop catalog mirrors it. Pimcore sends the data and the shop receives it. Endora Commerce does not query Datahub, does not browse Pimcore folders and does not sync objects by ID: every record arrives complete, already mapped to the catalog structure, and signed with a shared HMAC secret over HTTPS.

  • Content and parameters: products with their descriptions and parameters, categories, attributes and attribute sets.
  • Product media: images and documents, such as data sheets, attach to the product in the shop. The shop downloads them only from the public Pimcore address and from extra HTTPS addresses you allow, such as a CDN. A file that cannot be downloaded is noted in the report, and the product still reaches the catalog.
  • Everything at once: each record carries the full product dataset, rather than content and media separately.

To start, Pimcore sends a full delivery, from the configuration object toolbar or with a console command. Until it finishes, live records are refused, unless you deliberately skip it because the catalog is already aligned. After that, every save in Pimcore sends a live record once the change is committed, so work in Pimcore never waits on the shop. Retrying the same send does not write the data twice.

How Pimcore concepts translate into go-live preparation

Mapping is set up in Pimcore, in the Endora integration package. The package uses the shop’s signed lookups: attributes, attribute sets, categories, languages, sales channels and price lists. The shop admin has no class or attribute mapping screens.

In PimcoreWhat to settle before go-live
Product class and its fieldsWhich fields become the name, description and attributes in the shop
Variants with inheritanceA variant inherits values from its parent object: which fields are shared and which tell variants apart
Classification StoreCategory-specific attributes added without changing the class: which groups of properties the buyer should see
Localized fieldsWhich languages go to the shop and what to do when a translation is missing
AssetsWhich images form the gallery and which files are downloadable attachments

What to prepare on the Pimcore side

The integration does not use Datahub: you do not enter a GraphQL address or a Datahub API key in the shop. Before going live, prepare:

  1. The Endora integration package installed in the Pimcore project with Composer. That is where you configure the mapping, the product folder and the category tree root.
  2. The Endora push URL, copied from the connection screen in the shop admin and pasted into the package configuration. The package appends the probe, full delivery and live record addresses to it by itself.
  3. A shared HMAC secret, the same on both sides. The shop stores it encrypted and never shows it again.
  4. A default, active sales channel in the shop, required before the connection can be enabled.
  5. A Pimcore test environment with a representative sample of the catalog, so the mapping can be checked before production.

Things worth asking about early

  • Name and languages. The package requires an SKU and a localized name field. Each Pimcore language goes to the shop language with the same code, and the name must be filled in for the shop’s default language.
  • Visibility of new products. By default, products arrive inactive and do not go on sale until you switch them on. You can instead have them always active, or have their shop status follow the publish state in Pimcore. Withdrawing or unpublishing in Pimcore deactivates the product in the shop but does not delete it.
  • Prices. Prices with no price list mapped are left out, unless you map them as ordinary attributes.
  • SKU quality. Products are identified by SKU and the numeric Pimcore ID, so the numbers have to be unique and stable.
  • Local edits in the shop. The shared PIM connector layer in the platform defines one consistent way of protecting fields, that is, excluding a locally refined value from being overwritten on import. It covers the name, description, attributes (also per language and sales channel), SEO metadata, gallery positions, attachments and category assignment. Ask which of these fields you want to edit in the shop.

One active PIM per installation

The platform lets you install several PIM connectors, but only one can be active. On a fresh install none is switched on: the operator picks one in the admin panel, and an attempt to enable a second is refused with the name of the connector already running. Changing PIM does not delete data, and products imported earlier stay in the catalog.

It is also worth knowing that the connector does not import stock levels, and the shop does not fetch data from Pimcore on a schedule. Changes arrive when Pimcore sends them. The delivery history in the admin panel shows every run with its outcome, counters and per-record issues, and administrators are notified when a run fails.

When Pimcore, and when another PIM

Pimcore is the natural choice if you already work in it, or if you need your own data model and asset management in one tool. If you are only starting to put product data in order and want a lighter start, see UnoPim. A ready-made model of families and attributes comes with Akeneo, and Ergonode is a PIM developed in Poland.

Want to see how your class model would translate into a shop catalog? Get in touch and show us a sample product page.

Let’s find out what we can connect at your end

Book a free consultation. We will walk through your ERP, warehouse and sales processes, and show you what the integration path looks like step by step.

Michał Zabielski, founder and CEO of Endora

Michał Zabielski

Founder and CEO, Endora

We reply within one business day.