Complementary products

Collab recommender demo



What is it?

When viewing a particular item, this recommender suggests other types of products based on which ones other customers have bought together.

For example, customers who bought a travel bag have also often bought a toiletry bag. This is similar to Amazon's book recommendations: "People who like x also like y".

image info

Demo link

Login: "freitag" / "freitag-DEMO"


  • For users: Truely relevant suggestions mean less search effort.
  • For Freitag: Better sales of complementary items such as accessories that customers might not know exist.


  • About 3 person days


  • Export of sales history in online shop
  • Freitag's products API

Used Technologies

Collaborative filtering recommender with pandas and opencv

Further vision

The recommendations could be further refined, and the bulk of the effort would be to integrate this recommender into Freitag's online shop.