Can you maximise recommendations generated by the Related Products model without relying on content-based filtering?
The only way to optimise the Related Products model to generate more recommendations without using content-based filtering is by sending more click and conversion events.
Related Products is designed as a hybrid model that combines content-based and collaborative filtering. If you choose to rely solely on collaborative filtering, you must compensate for the absence of content-based recommendations by increasing event data.
The benefit of including content-based filtering is that you can get recommendations for the products that don’t have enough event data and can further refine the recommendations for the ones that do.
The hybrid model works by combining, for each product, the recommendations based on user interactions (collaborative filtering) with those based on the attributes you choose (content-based filtering). This means some products will have a set of recommendations from only one model, whereas some will have a combination of the two.