Key Takeaways
- Amazon's Q1 2026 revenue hit $181.5B, beating estimates. AWS grew 28% — its fastest pace in 15 quarters — with an AI revenue run rate now above $15B
- Rufus monthly active users are up 115% and engagement nearly 400% year-over-year. Buy For Me and Scheduled Actions push Amazon firmly into agentic commerce
- Andy Jassy explicitly stated that advertising will continue to thrive in an agentic world, signaling Amazon's intent to lock the buying decision inside its own AI surface rather than cede it to horizontal agents
Amazon Pairs an AWS Reacceleration With a Sharper Agentic Commerce Story
The e-commerce and cloud giant just released quarterly results with a focus on AI investment, AWS growth acceleration, and agentic commerce strategy.
sherwood.newsOn April 29, 2026, Amazon reported first-quarter results that beat the Street on revenue and operating income. Sales reached $181.5B, ahead of the $177.2B consensus, while operating income of $23.9B cleared the high end of management's prior guide. Yet the most striking part of the call was not the headline beat — it was how directly CEO Andy Jassy tied AWS's reacceleration to Amazon's emerging "agentic commerce" stack.
AWS revenue climbed to $37.6B, up 28% year over year — the fastest growth in 15 quarters and a notable uptick on a very large base. AI services within AWS are now running above a $15B annualized run rate, while Amazon's custom silicon business is growing roughly 40% quarter-over-quarter and has crossed a $20B annual run rate. The story has shifted from "is AWS growth still slowing" to "how durable is this rebound."
Reading the Q1 Print and Q2 Guide
Q1 segment data confirmed how heavily Amazon's mix has tilted toward cloud and advertising. Ads generated $17.2B in the quarter, up 22%, with trailing-twelve-month revenue past $70B. Stores units grew 15% year over year — Jassy's highest reading "since the tail end of covid lockdowns."
Q2 guidance came in mixed: revenue of $194B-$199B (midpoint above the $189.1B consensus) and operating income of $20B-$24B (midpoint slightly below the $22.9B view). Investors are now pricing in strong demand alongside meaningful investment drag, which explains why AMZN whipsawed in postmarket trade despite a 26.3% April rally heading into the print.
The capex backdrop matters: Amazon plans roughly $200B in 2026 capex, well above the $146.1B prior consensus and the largest single-company plan among the hyperscalers' aggregated $630B-$690B. Jassy's defense is that AI capacity is being "monetized as quickly as it is installed," with AWS's $364B backlog (excluding the new $100B+ Anthropic deal) cited as proof of demand.
The Agentic Commerce Stack: Rufus, Buy For Me, and Scheduled Actions
The most strategically important thread of the call was Amazon's agentic commerce push — the idea that AI agents will increasingly research, compare, and execute purchases on a shopper's behalf. Amazon is building this stack across three layers.
Rufus: an in-app shopping agent that is already being used at scale
Rufus, Amazon's generative AI shopping assistant, posted monthly active users up 115% and engagement up nearly 400% year-over-year. It can research products, track prices, and auto-buy items when they hit a target price. Crucially, ad surfaces are baked in: Sponsored Product and Brand prompts already drive roughly 20% of shoppers who tap a brand prompt to keep conversing about that brand, according to Jassy.
Buy For Me: pulling third-party SKUs into Amazon's funnel
Buy For Me lets shoppers complete purchases on third-party retailer sites — Shopify, WooCommerce, Squarespace stores, and others — without leaving Amazon's app. The experience runs on Amazon Bedrock, with Amazon Nova and Anthropic's Claude powering the agentic checkout. In March 2026, Amazon expanded the program by letting merchants actively contribute product feeds via Feedonomics, Salsify, and CEDCommerce. For Amazon, this captures demand for SKUs it does not stock without ceding the front end to a competing agent.
Scheduled Actions: from assistant to autonomous agent
The day before earnings, Amazon launched Scheduled Actions for Rufus. Rufus can now place orders without a fresh shopper prompt — on a calendar, at a price trigger, or when it predicts a household has run out. Because Scheduled Actions integrate with Shop Direct and Buy For Me, Rufus can autonomously buy from third-party merchants when their catalog or price beats Amazon's. This is the moment Rufus moves from assistant to agent.
Will advertising survive an agentic world?
Asked on the call how agentic commerce would affect ads, Jassy was direct: "I actually believe that advertising will do well in a world of agentic commerce." He argued that horizontal third-party agents still struggle with accurate pricing, inventory, and personalization, while Amazon intends to make Rufus "the best shopping assistant anywhere." Multi-turn conversations also create multiple ad surfaces that did not exist in traditional search. Creative Agent, Amazon's agentic ad-creative tool, has now expanded to Canada, the UK, France, Germany, Italy, Spain, and India.
What This Means for Merchants and Marketers
The implications extend well beyond Amazon's P&L. First, being discoverable by Rufus is now a discipline. Sponsored prompts are pay-to-play, but organic surfacing depends on the structured quality of titles, attributes, descriptions, and reviews. Catalogs that are easy for an LLM to cite will win.
Second, Buy For Me and Shop Direct effectively reroute non-Amazon demand back into Amazon's funnel. Merchants must weigh the channel-mix benefit of incremental orders against deepening dependency. Shopify merchants in particular should pair this with Universal Commerce Protocol-based access to Google AI Mode and ChatGPT to avoid single-channel concentration.
Third, Scheduled Actions break last-click attribution. When Rufus places an order weeks after the click that influenced it, ACoS and TACoS at the campaign level start to mislead. Lengthening attribution windows and managing efficiency at the SKU level becomes the practical fix.
Conclusion
Amazon's Q1 2026 print told two stories at once. AWS's return to 28% growth, a $15B+ AI run rate, near sell-out demand for Trainium2 and Trainium3, and a $200B capex plan together argue that AI demand is real and is converting into revenue. At the same time, Rufus's 400% engagement jump, the maturation of Buy For Me, and the launch of Scheduled Actions show Amazon working hard to keep the purchase decision inside its own agent — not someone else's. Agentic commerce has clearly moved past the proof-of-concept phase. For merchants, marketers, and analysts, the question for the rest of 2026 is whether their data, channel mix, and measurement stack are ready for buyers who never see a search results page.
