AWS 24% Growth, AI Chips Surge, Capex to Hit $200B+ (AMZN Q4 2025 Earnings Call)
Export as clean Markdown. Drag & drop into ChatGPT, Claude, or Gemini.
Amazon’s 2025 momentum centers on AWS accelerating to 24% growth and AI-driven demand translating into rapidly monetized compute, while retail and ads reinforce cash-generating scale. The key investor question becomes how quickly that AI-capacity build converts into durable profitability.
AWS accelerates to 24% growth as AI and chips monetize capacity
Amazon reported $213.4 billion in revenue, up 12% year over year excluding foreign exchange, with operating income of $25 billion. Management emphasized that AWS growth “continued to accelerate to 24%,” the fastest in thirteen quarters, and that AWS is now a $142 billion annualized run rate business.
That matters because the financial leverage comes from demand converting into installed capacity, not just spending. The CFO also tied profitability movement to investment timing, noting AWS operating income was $12.5 billion while operating margins “are going to fluctuate over time” based on AI investment and depreciation.
Guidance frames Q1 operating income with FX and LEO cost headwinds
For Q1, management guided net sales to $173.5 billion to $178.5 billion, and said the range includes “a favorable impact of approximately 180 basis points from foreign exchange rates.” They also guided Q1 operating income to a maximum of $21.5 billion (as stated on the call), highlighting how FX and segment costs shape earnings cadence.
On segment-specific pressure, the CFO stated: “Within the North America segment, we do expect a year over year cost increase of approximately $1 billion related to Amazon LEO.” The guidance logic matters because it shows higher near-term expense tied to satellite launches, with a later shift to capitalized costs as manufacturing and launch services move “capitalized” later in the year.
Retail, ads, and fulfillment speed deepen engagement and ad monetization
North America delivered $127.1 billion revenue, up 10% year over year, while International was $50.7 billion up 11% year over year excluding FX. Management linked unit growth and profitability to improved delivery performance and network cost structure, including regionalization and fulfillment network refinements.
In retail, the CEO highlighted everyday essentials and grocery as key drivers, stating everyday essentials “grew nearly twice as fast as all other categories in The US” and represented “one out of three units sold.” In advertising, the CFO said ads revenue grew 22% in Q4 and exceeded $12 billion of incremental revenue in 2025, which matters because it monetizes consumer activity across search, browsing, and streaming signals.
Q&A centers on ROI visibility, AI chip supply, and how agents may reshape retail
In response to investor questions about return on invested capital, the CFO said investments are designed so “we are putting into service with customers all capacity that we are getting,” and he argued demand and backlog turn that capacity into P&L. He also pointed to AWS margin as already reflecting investment-to-monetization progress, citing AWS operating margin “35% operating margin through Q4” (noting it fluctuates with AI investment and depreciation).
On AI supply-demand dynamics, management discussed both backlog and capacity constraints. The CEO stated AWS backlog was “$244 billion,” up 40% year over year, and said every provider faces capacity limitations, so Amazon is “being incredibly scrappy” in adding power and compute. For chips, he emphasized Trainium monetization and supply visibility, stating Tranium three supply is expected to be “nearly all... committed by mid-2026,” and that customers want inference deployed where application data and workloads already run.
Finally, analysts asked whether agentic shopping could compress the retail funnel. The CEO responded with a customer-experience framing, stating customers want “broad selection, low prices, really fast delivery,” and trust, and arguing retailer-specific agents have an advantage because horizontal agents “get a lot of the product details wrong” and “the pricing wrong.”