In a January blog post, Amazon detailed the expansion of its AI-driven search and discovery tools designed to interpret customer intent rather than rely solely on keyword matching. The system now factors in review sentiment, price sensitivity, delivery speed expectations and return-rate patterns, drawing on prior browsing behavior across hundreds of millions of items in more than 30 categories. By surfacing a concise set of product options and ranking them according to inferred needs, Amazon aims to reduce the average time spent per purchase decision by up to 40% in complex categories such as electronics and home appliances.

Amazon has begun rolling out new agent-like interfaces, including the shopping assistant ‘Rufus,’ which can track price fluctuations, summarize thousands of reviews and suggest alternative products. Visual search through Amazon Lens has achieved a 25% higher match rate for user-submitted images compared with legacy keyword methods. The ‘Buy for Me’ service, currently in beta with 50,000 users, autonomously completes purchases—even for third-party sellers—based on user-defined preferences, handling tasks from checkout to delivery scheduling. Iterative testing with customer cohorts has driven a 15% lift in conversion rates for participants.

Despite a modest 5% share-price gain in 2025, Amazon reported 13% year-over-year revenue growth in its third quarter, with net income rising 38% over the same period. The advertising segment posted 24% growth to reach $17.7 billion, representing nearly 10% of total net sales, while AWS returned to a 20% growth trajectory driven by AI workloads. Its in-house Trainium2 AI chips saw sequential revenue increase of 150%, now accounting for a multibillion-dollar annualized run rate. At a forward P/E multiple in the low 30s, analysts argue the company’s diversified growth engines and durable operating margins position it for a re-rating in 2026.