Steam uses a machine learning system to recommend games to players based on what they and millions of other users have actually played — not on genre labels or tags. The system launched as an experiment in July 2019 and now powers multiple features across the Steam store, including the front page, the Discovery Queue, and the "Play Next" shelf.

The recommendation engine uses a neural network trained on billions of play sessions from millions of Steam users. Unlike traditional recommendation systems that rely on metadata like genre or price, this model works purely from behavior: if players with similar playtime patterns enjoy a game, that game gets surfaced to you. The model disregards manually assigned tags or user-written descriptions entirely. Steam's Discovery Queue — powered by this system — has served 18 billion game store page views to 115 million players. Valve reports that personalizing the front page with this approach increased purchase conversion rates by 27% and expanded the number of games showcased to customers by 46%.