Alexa for Shopping: Brand Catalog Guide

Fact: Amazon retired Rufus — used by more than 300 million customers in 2025 — and replaced it with Alexa for Shopping, a full AI agent that can make purchases across the web on a customer’s behalf.
Impact: Traditional keyword-based ranking and sponsored search are no longer the only levers for brand visibility; catalog completeness and AI-surface readiness now determine whether your product appears in shopping conversations at all.
Surprise: Alexa for Shopping can buy products from other retailers’ websites without the seller’s consent — turning Amazon into a universal shopping interface and sparking immediate backlash from merchants who say they never opted in.

Three years after Amazon launched Rufus as a cautious experiment in conversational product discovery, the company has pulled the plug on it. Not upgraded it — retired it. What replaced it is a different proposition entirely: Alexa for Shopping is not a chatbot bolted onto a search bar. It is an agent. One that tracks prices, schedules recurring orders, compares products side by side, and, most consequentially, completes purchases on other retailers’ websites without the customer ever leaving Amazon’s interface.

More than 300 million customers used Rufus in 2025. Every single one of them is now being migrated to a system designed to do far more. That is the number brand managers need to sit with before deciding whether this warrants attention.

What Alexa for Shopping Actually Does — and Why It Is Not Just an Upgrade

The surface-level summary: Amazon merged Rufus and Alexa+ into a single assistant, making it free for all U.S. customers on the Amazon app, Amazon.com, and Echo Show displays — no Prime subscription required. But the capabilities assembled under that banner represent a genuine shift in how purchase decisions get made.

Alexa for Shopping evaluates multiple products simultaneously, comparing features, prices, and review data in a single response. It draws on a user’s full purchase history to build a level of personalization that keyword-based search never achieved. Price tracking and recurring order scheduling move buying behavior away from active search sessions entirely — meaning a customer hunting for pet food or household essentials may never run a keyword search again.

Then there is the “Buy for Me” capability. When a product is unavailable or uncompetitive on Amazon, Alexa navigates to another retailer’s website and completes the purchase using the customer’s saved payment and shipping details. Several merchants have already reported that they never agreed to this arrangement. CNBC confirmed the feature is live and that Amazon views it as expanding choice rather than cannibalizing third-party seller relationships. Not everyone finds that framing persuasive.

300 Million Users and the Advertising System That Has to Adapt

Amazon’s advertising revenue is one of the fastest-growing segments of its business precisely because it sits between intent and purchase. When a customer types “wireless earbuds under €50,” the auction determines who gets seen first. Alexa for Shopping disrupts that model in one critical way: the agent interprets intent rather than matching keywords, which means relevance is no longer purely a function of what you bid.

Sponsored products will still appear inside Alexa conversations, Amazon has confirmed — but only when they “enhance” the shopping experience. Placement will depend on how well a product matches the AI’s interpretation of a user’s specific, personalized request. That interpretation is driven by structured catalog data: complete attributes, accurate variant information, A+ content that gives the model rich context about what a product actually does and for whom.

What’s striking about this move is the competitive timing. The agentic commerce layer is assembling simultaneously across every major platform. Google launched AI Mode this week with BNPL integrations through Affirm and Klarna. Etsy debuted a native app inside ChatGPT. Amazon’s decision to retire Rufus — not iterate on it — signals that the company views agentic shopping as the primary interface going forward, not a supplement to keyword search.

Is your Amazon catalog ready for AI-driven discovery? See how Epinium’s catalog tools prepare your listings for the agent era →

The Catalog Gap Is Now the Visibility Gap

Here is the uncomfortable reality for brands that have relied on advertising spend to maintain Amazon visibility: an AI agent cannot recommend a product it cannot understand. When Alexa processes a request for “a lightweight running jacket for cold mornings under €120,” it is not running a keyword match. It is interpreting meaning — and pulling from structured product data to identify candidates worth surfacing.

Epinium data

Across catalog audits of more than 8,000 Amazon product listings managed by Epinium, roughly 3 in 5 carry at least one critical attribute gap — missing variant data, incomplete bullet points, or outdated backend keywords — that AI recommendation surfaces systematically deprioritize. With keyword search, advertising spend could offset these gaps. With Alexa for Shopping, it cannot.

Brands that prepared their catalogs for Rufus now have a structural advantage. Those that treated AI readiness as a future concern face a more urgent problem. Rufus grew 115% in its first year of agentic deployment while a majority of brands still carried critical catalog gaps. Alexa for Shopping has a larger user base, higher purchase intent integration, and agentic capabilities Rufus never had.

What we’re seeing at Epinium is a compression of the timeline brands assumed they had. The shift from search-box ranking to AI-surface readiness was always predictable — most brands simply expected more time. The catalog gap is now the visibility gap. They are the same problem.

Frequently Asked Questions

Does Alexa for Shopping affect all Amazon sellers or only Prime-enabled listings?

Alexa for Shopping is available to all U.S. customers regardless of Prime status, and Amazon says it is designed to “expose even more products” rather than narrow results. In practice, the AI’s ability to recommend a product accurately depends on how well the listing’s structured data — attributes, variant information, bullet points — matches the user’s request. Sellers with incomplete catalog data face lower recommendation rates regardless of fulfillment tier or Prime eligibility.

Will existing Sponsored Products campaigns still perform with Alexa for Shopping in place?

Amazon confirmed sponsored products will appear inside Alexa conversations when they “enhance” the experience. However, placement is no longer purely auction-based: AI relevance plays the primary role. Brands with low-quality catalog data may find their ad spend generates fewer impressions despite active campaigns, because the model deprioritizes poorly structured listings before any bid is ever evaluated.

This is the most contentious element of the launch. The “Buy for Me” feature allows Alexa to navigate to third-party retailer sites and complete purchases using a customer’s saved credentials. Multiple retailers have said they never opted in. Brands should audit their DTC checkout flows and terms of service — and assess whether being purchasable through Amazon’s AI agent is a strategic asset or a margin liability, depending on their pricing architecture.

When should a brand prioritize catalog optimization over ad spend in response to this?

Immediately, for any product with incomplete attributes or generic bullet points. Advertising can still drive traffic, but if Alexa encounters a poorly structured listing during a shopping conversation it will move to the next candidate — and no bid level changes that outcome. Catalog optimization is the foundation; advertising amplifies reach once the foundation is solid. Brands spending heavily on Amazon ads with incomplete listings are effectively paying to deliver shoppers to a page the AI will skip.

What happens to content strategies built specifically for Rufus?

Rufus answered specific product questions only when customers opted into it. Alexa for Shopping is embedded in the main search bar and operates proactively throughout the shopping session. Content optimized for Rufus — rich Q&A sections, detailed use-case descriptions, product FAQs — remains valuable and transfers well. The additional requirement is structured backend attribute completeness, which powers Alexa’s comparison and filtering features in ways Rufus never needed.

The retirement of Rufus is not a reset — it is an acceleration. Amazon spent two years studying how customers interact with conversational product discovery. Those learnings are now built into Alexa for Shopping, which carries a mandate to become the default way hundreds of millions of customers make purchasing decisions. Brands still waiting to get serious about AI-surface readiness have arrived at the point where waiting has a measurable cost.

Ready to close the catalog gap before Alexa for Shopping reshapes your brand’s visibility? Epinium’s catalog management platform audits, structures, and continuously optimizes your Amazon listings for AI-driven discovery — so your products appear in the conversations that convert. Discover how Epinium prepares your catalog for the agent era →