Can I use SDM modules in any order?

Yes, SDM modules are highly flexible and can be reused at various stages of your workflow. For example, you can perform light data verification at the start (to ensure key fields like EAN and SKU are present) and then apply more detailed normalization at the end to ensure data quality before export.

However, some sequences are more effective than others. For instance, running an extraction module before setting up families is usually impractical. A typical workflow might start with mapping source attributes and conclude with normalization, allowing for final adjustments before data is exported.

Mapping

Can SDM automatically create new attributes if they don’t exist in the PIM?

No, the mapping module relies on the PIM’s data model. If you need a new attribute, it must first be created in the PIM and synced to SDM before it can be used.

Does SDM support localizable attributes?

Yes, SDM supports localizable attributes (e.g., description_fr-FR and description_en-US). For example, if your spreadsheet includes product descriptions in both French and English, SDM can map the French text to description_fr-FR and the English text to description_en-US. Models in SDM can also handle multiple languages seamlessly.

Value Mapping

Can SDM automatically add new values to lists during value mapping? 

No, to maintain clear governance between the PIM and SDM, value mapping operates as a one-way flow. This means that any new values must be created in the PIM before they can be used in SDM.

Classification

Can SDM automatically create new categories when importing data?

When you import files, new categories are not added automatically. You need to add them to your SDM configuration first. This can be done in one of two ways: 

1. If the categories exist in your PIM, they will be automatically retrieved and added through PIM sync.
2. If you’re not connected to the Akeneo Connector, you can ask your support contact to add the categories manually.

This process helps maintain a consistent and controlled data model between systems and prevents two-way updates.

Extraction

What happens during an Extraction step if an attribute is already filled with a value?

The extraction step can be configured in two ways:

1. The model only fills empty attributes and does not overwrite the existing values.
2. The model runs for specific or all attributes, but it won’t notify you if its output differs from the supplier-provided values. 

Note that for the second option, the extraction step can take a long time to process all attributes. We recommend defining specific attributes.

Can SDM extract data from images?

Not yet! However, you may contact our Pre-Sales or Tech Partners team to explore complementary solutions.

Normalization

Can I set up my own rules at the normalization step?

Currently, rule creation from the interface is not available because it requires backend configuration.