Intergamma wanted to provide products with related products such as variants, alternatives, products in the same family, accessories and upsell options on its websites. Establishing these product relationships manually is a time-consuming process. We have made it possible to automate this process.


We carried out this project in which the following product relationships were detected by an algorithm in the products in the product categories window decoration, switch material, hardware, bathroom accessories from GAMMA:
Base products / variants: based on a base product, the products from the same supplier but different on a property such as color or size
Alternatives: products with the same properties, but from a different supplier (and possibly a better margin)
Family: products in the same line / family, eg power strip in the same style as dimmer


The software has a user interface in which a user can type the found product relationship per product and view and approve the related products. The algorithm models the relationships based on smart Natural Language Processing (NLP), whereby the distinguishing properties of the product (such as color, width and length) are automatically derived from the product description. Fuzzy matching is also used to detect the similarities between values ​​such as gray / brown and brown / gray.
If available, the products are provided with product images so that assessment can take place easily.
The software can be configured with filters so that the right product relationships can be found.

The approved relationships can be exported in an Excel file where reference is made per base product to the relationships found, including the relationship type (eg variants) so that the completeness of the relationships can be assessed.