What a configuration could hand over
With a larger catalogue, the expensive mistake is a price list maintained twice. The configurator can read product lines, variants and media straight from Pimcore instead of copying them into another system. Which objects and attributes are used follows your data model.
Data objects
Product lines and their variants would be read as data objects of your classes through a Datahub endpoint, limited to the attributes the configurator needs for the choice.
Assets
Surface photos, textures and data sheets can stay as assets in Pimcore, with the configurator linking to them rather than keeping its own copy.
Localized fields
If names and descriptions live in localized fields, the configurator could take the matching language version for each language of the website.
Mutations
In the other direction, a submitted configuration could become a new object in an agreed folder, if the endpoint allows a create mutation. Whether enquiries belong in a PIM at all is a separate decision.
What we need to connect it
A dedicated GraphQL configuration would be set up in Datahub for the configurator, with its own API key sent with every request.
The workspaces of that configuration decide which object and asset folders the configurator can reach and with which rights: read, create or update.
We go through the endpoint schema together in the GraphiQL explorer before mapping, so it is clear which classes and attributes are used.
If you run a test instance of Pimcore, the connection is built against it first.
Worth deciding up front
- We agree which objects are ready for the website. A half-finished variant should not reach the configurator just because it sits in a folder the endpoint can see.
- Prices are only sometimes kept in Pimcore. If they live in the ERP, the configurator would take product data from Pimcore and prices from the other source, which is settled at the start.
- Datahub screens and options differ between Pimcore versions, so the scope is verified against the version you run.
Questions about this connection
Do we have to change our Pimcore data model for the configurator?
Mostly not, the configurator would read your existing classes. What tends to be missing is production limits and combination rules, and together we decide whether you add them to Pimcore or they stay in the configurator.
How would a new variant added in Pimcore show up on the website?
The configurator would pick it up on its next data load. Whether that load is triggered by a change in Pimcore or runs at regular intervals is designed during implementation.
Other integrations
Put Pimcore to work in your sales.
Tell us what you sell and how an enquiry is handled today. On the demo you will see the configurator, and we will go through which data could flow from it into your system.