{"id":22832,"date":"2025-04-25T23:12:09","date_gmt":"2025-04-25T23:12:09","guid":{"rendered":"https:\/\/www.bloomreach.com\/library\/two-tower-neural-networks-and-personalized-product-recommendations"},"modified":"2026-06-02T19:03:32","modified_gmt":"2026-06-02T19:03:32","slug":"two-tower-neural-networks","status":"publish","type":"library","link":"https:\/\/www.bloomreach.com\/en\/blog\/two-tower-neural-networks","title":{"rendered":"Two-Tower Neural Networks and Personalized Product Recommendations"},"content":{"rendered":"

Product recommendations<\/span><\/a> are a critical use case for modern <\/span>digital commerce<\/span><\/a> companies. Recommendation systems play a significant role in suggesting <\/span>personalized products<\/span><\/a> to customers, resulting in increased sales and customer satisfaction.<\/span><\/p>\n

Collaborative filtering has been a popular technique in ecommerce product recommendations for many years, with the traditional approach involving matrix factorization. However, as the limitations of this more traditional approach have surfaced, more advanced techniques such as neural networks have been utilized.<\/span><\/p>\n

Neural networks<\/span><\/a> are a powerful tool from the world of <\/span>artificial intelligence (AI)<\/span><\/a>. They use interconnected nodes, or neurons, in a layered structure inspired by the human brain to teach computers how to process and interpret data. This type of machine learning, known as deep learning, allows computer systems to adapt and improve over time by learning from their mistakes. By leveraging this technology, AI-powered neural networks can solve complex problems that traditional computing would struggle with.<\/span><\/p>\n

When testing a Two-Tower Neural Network instead of the <\/span>Alternating Least Squares (ALS)<\/span><\/a> matrix factorization algorithm, we saw significant improvement with personalized product recommendations in the recall@10 offline metric. Specifically, we tested this on six customers and saw 147%, 23%, 27%, 25%, 25%, and 90% improvement on each respective customer for an average of just over 56% improvement per customer. Results varied across catalog and interaction volume characteristics \u2014 the spread reflects real differences in how much each customer’s data set benefited from feature-rich representations over pure interaction signals.<\/span><\/p>\n

The Two-Tower Neural Network is a deep learning algorithm that consists of two parts, where one part serves to represent the user and the other the item, and the parts are connected at the end to create the final prediction. The user tower processes behavioral signals \u2014 browsing history, past purchases, session context \u2014 into a dense embedding vector. The item tower independently processes product features (title, category, description, price) into a matching vector for each item in the catalog. Recommendations are generated by computing dot product similarity between the two embeddings, which makes retrieval fast enough to match users against catalogs of millions of products in real time.<\/span><\/p>\n