Case Study: Segundamano achieves 3× conversions and real-time personalized recommendations with Recombee

Bringing Personalization to Advertising for 160 Million Monthly Views

Segundamano, one of Mexico’s largest online classifieds platforms with about 160 million monthly views, needed to personalize recommendations across a constantly changing catalog of roughly 4 million products per month. To solve this at scale and in real time, Segundamano partnered with Recombee and adopted Recombee’s collaborative‑filtering recommender engine (using the recommendationsToUser approach) to automate personalized suggestions.

Recombee implemented a hybrid solution combining collaborative filtering, deep‑learning image similarity and advanced Spanish text processing, with realtime model updates and sub‑100 ms recommendations applied to the “También te podría interesar” product‑page box. The deployment delivered clear results: Recombee recommendations accounted for 10% of item views, produced 3× more conversions and 4× more interactions from users who engaged with recommendations, and enabled high‑performance, realtime personalization across Segundamano’s marketplace.

Segundamano

Marco Alvarez

Product Manager