Throughout the history of ecommerce, machine learning has helped online retailers often know what their customers want before they want it. In fact, one article estimated that 35% of the revenue of one of the world's largest ecommerce platforms came from their product recommendations.1 That's a staggering statistic when you consider that these were purchases that people made, without intentionally setting out to! It's the digital equivalent to making that "impulse buy" when you're checking out at your local grocery store, except on a much larger scale. The reason behind shoppers' willingness to buy recommended products is personalization. By using machine learning to analyze the buying patterns of millions of users, common themes tend to show up that group people who often buy the same products.