Case Study: Pepper achieves 21% increase in click-outs to affiliate links with Recombee

Increasing Click-Outs by 21% for Pepper - The World's Largest Shopping Community

Pepper, the world’s largest social shopping community with over 25 million shoppers, faced the challenge of improving product recommendations across iOS, Android and web while replacing an advanced in‑house engine, managing rapidly changing inventory across 10 countries, and increasing click-outs to affiliate links without hurting user satisfaction. To solve this, Pepper engaged Recombee and adopted Recombee’s AI‑driven product recommendations, including a personalized “For you” home feed and Infinite Scroll for real‑time cross‑device personalization.

Recombee implemented an ensemble of incrementally trained models (collaborative filtering, popularity boosters, reinforcement learning/contextual bandits) with centralized configuration and real‑time infinite scrolling. The deployment delivered measurable impact: Recombee’s solution boosted click-outs to affiliate links by 21% on the website and 6% on mobile apps, improved user experience, and provided real‑time personalization and one place to manage recommendations across all Pepper platforms.

Pepper

Heike Guertler

Head of Product