KNOWLEDGE BASE
How do you model product and category data in e-commerce?
Table of Contents — English
What is the difference between a product and a variant?
A product is a commercial concept that shares a common description and attributes. A variant is the concrete unit sold according to color, size, or another option. A SKU is the code the business uses to track stock. The scope and uniqueness rule of the SKU must be clearly defined; it should not be assumed that every external system will use the same identifier.
For example, "linen shirt" can be the product and "blue, size M" the variant. The description can be kept at product level, and stock and barcode at variant level. Some attributes are managed at different levels depending on the business; the model should be tested with real catalog examples.
What steps are used to build the data model?
1. Select representative product examples. Cover cases such as a single-option product, a multi-variant product, and a discontinued product. Examining only the tidiest records hides the exceptions.
2. Determine the level and the owner of each field. Write down which system is the source for product name, variant code, description, size, unit, price, and image.
3. Establish persistent internal identifiers. A URL slug or an editable product name should not be used in place of the primary identifier. External system identifiers can be kept in separate mapping fields.
4. Define the required-field and validation rules. An empty title, a duplicated SKU, an invalid unit of measure, or an unrelated variant should be caught before publishing.
5. Separate the publication statuses. Draft, published, and archived records should behave differently. The links of an archived product to past orders must be preserved.
How should categories and attributes be separated?
A category provides navigation and classification; an attribute specifies a property of the product. "Shoes" can be a category, "42" a size, and "leather" a material. Turning every attribute value into a category can make management harder. Associating a product with more than one category and choosing a primary category are separate decisions.
How is catalog data prepared for GEO and SEO?
Title, description, size, and usage conditions should not contradict each other. Product codes and brand spelling should be used consistently across all channels. Google's guide on e-commerce site structure explains that the links between categories and products support discovery. Do not turn the catalog into records that can only be reached through on-site search.
If Product structured data is to be used on the product page, the information should reflect the actual product and the visible content. Google explains that product data can be provided through page markup and Merchant Center. This technical option does not mean that every product will receive a special search appearance.
How do you test that the model works correctly?
Change the name of a variant and check that the external system mapping does not break. Try to create a second record with the same SKU. Verify that the expected warning is produced when a value without a unit is entered in a measurement field. Finally, review how an archived product's description appears in an old order; the historical commercial record should not become ambiguous because of a new catalog change.
What are the frequently asked questions?
What is the next step?
Prepare your representative product and variant examples; let's clarify the catalog fields and their source systems with PROJX Digital.
