Amazon's AI shopping assistant is the largest one in production by a wide margin. Under the Rufus name it reached more than 300 million customers in 2025. In May 2026 Amazon merged it with Alexa+ and renamed it Alexa for Shopping, and it now sits in the main search bar for every signed-in US customer. Here are the numbers Amazon has disclosed and the independent data around them.
Headline Numbers
| Metric | Figure | Source |
|---|---|---|
| Customers who used Rufus in 2025 | 300 million+ | Amazon Q4 2025 earnings |
| Incremental annualized sales | Nearly $12 billion (up from a ~$10B run rate disclosed in Q3 2025) | Amazon Q4 2025 earnings |
| Monthly active users, year over year | +115% | Amazon Q1 2026 call, via Modern Retail |
| Engagement, year over year | +400% | Amazon Q1 2026 call, via Modern Retail |
| Purchase likelihood for users who engage | 60% more likely to complete a purchase | Amazon Q1 2026 call, via Modern Retail |
| Share of Amazon app sessions using Rufus | 30% on Nov 1, ~40% on Black Friday, 38% at month end (2025) | Sensor Tower |
| Purchase sessions on Black Friday vs prior 30 days | +100% with Rufus, +20% without | Sensor Tower, via TechCrunch |
Two cautions. The $12 billion is Amazon's own incremental estimate, not audited revenue. And Sensor Tower itself says it cannot separate Rufus's causal effect from the fact that motivated shoppers are more likely to use it.
What Changed With the Rename
Alexa for Shopping combines Rufus's product knowledge with Alexa+'s personal context, so it draws on shopping history, browsing, and earlier conversations across Alexa devices. Amazon lists price history up to a year back, price alerts that can buy automatically at a target price, scheduled and recurring purchases, side-by-side comparisons inside search results, and AI overviews on search and product pages. It needs no Prime membership and no Echo device. Separately, Alexa+ itself is free for Prime members and $19.99 a month otherwise.
Buy for Me: Amazon Shopping the Open Web
The feature most relevant to brands outside Amazon is Buy for Me. Amazon says the assistant can shop "tens of millions of items in other online stores" and complete the purchase on the customer's behalf using their Amazon address and card. That makes Amazon's assistant a buyer on independent brand sites, not only a recommender on its own marketplace. Brands whose sites block automated checkout or hide price and stock behind scripts are the ones that get skipped.
Sponsored Prompts
Amazon also sells placement inside the assistant. On the Q1 2026 call Andy Jassy said nearly 20% of shoppers who interact with a brand's sponsored prompt continue the conversation about that brand, and added that "advertising will do well in a world of agentic commerce." Expect assistant placements to become a standard line in Amazon ad budgets.
Brand Visibility Implications
At Amazon's scale, whether its assistant names your product in an answer affects sales directly. The assistant reads listing content, reviews, and customer questions, so listing copy written for keyword search needs rewriting for questions and comparisons. Off Amazon, Buy for Me means your own site is being read by an Amazon agent deciding whether to buy from you. Treat both surfaces as places where an AI decides, and measure them separately from ChatGPT and Google because they draw on different data.
Methodology
Figures on this page come from primary sources wherever one exists: company announcements, developer documentation, and earnings materials, including OpenAI's commerce documentation, the ACP specification repository, Shopify's Spring '26 developer release, Shopify's agent profile docs, and Amazon's Q4 2025 earnings release. Where only trade press reports a figure, such as retailer conversion comparisons from Modern Retail or Digital Commerce 360, the page says so. Where sources disagree, a range is given. Status as of September 2026.
How Presenc AI Helps
Presenc AI tracks how AI shopping assistants describe and rank your products against competitors on category and comparison prompts, so you can see where listing or site changes move your share of recommendations.