Amazon renamed Rufus on May 13, 2026. The assistant now lives under a new name, Alexa for Shopping, and the change caught a lot of sellers off guard mid-optimization.

Here is the part of the Amazon Rufus vs A10 debate that actually matters for your listings: the two systems aren’t competing for the same job. A10 decides what shows up when a shopper types a search. Alexa for Shopping decides what gets recommended when a shopper asks a question. A listing built only for one of these is invisible to the other.

What Is Amazon A10?

A10 behavior is well documented by third-party seller tools and agencies who track ranking changes across thousands of listings. The signals that consistently move rank:

• Conversion rate. The most heavily weighted signal most sellers and agencies report. A higher-converting listing tends to outrank a listing with more total sales but a lower conversion rate.
• Click-through rate. How often shoppers click your listing when it appears in search results.
• Sales velocity and organic sales history. Recent, sustained sales carry more weight than a one-time spike.
• External traffic. Amazon rewards traffic from outside the platform, from Google, social, or influencer links, more than it did under A9.
• Seller authority. Account health, return rate, and feedback score now factor into how much Amazon trusts your listing.
• Keyword relevance. Still matters. It is just no longer the only thing that matters.

A10 is a lexical system. It matches the words a shopper types against the words in your title, bullets, and backend search terms. If your listing does not contain the word “hiking” anywhere, A10 will struggle to surface it for “hiking gear,” even if your product is objectively a great fit. Alexa for Shopping was built to solve exactly that limitation.

What Is Amazon Rufus (Now Alexa for Shopping)?

Alexa for Shopping is Amazon’s AI shopping assistant, accessible through a cursive “A” icon in the Amazon app, on Amazon.com, and on Echo Show devices. Shoppers ask it questions in plain language instead of typing keywords. “What’s a good gift for someone who loves hiking?” is a Rufus-style query. “Hiking backpack” is an A10-style query. Same shopper, same intent, completely different input, and your listing needs to answer both.

Under the hood, the assistant uses retrieval-augmented generation (RAG), pulling information from your product listing, customer reviews, community Q&A, and Amazon’s product catalog before generating a response. The relevance-matching layer behind this is called COSMO, a knowledge graph Amazon researchers described in a paper presented at ACM SIGMOD 2024. COSMO does not just match words. It maps products to real-world use cases, so a search for “shoes for a wedding” can surface formal footwear even if your listing never uses the word “wedding,” as long as your listing clearly communicates that the shoes are dressy, low-heeled, and appropriate for formal occasions.

The rename itself is simple in practice. Amazon merged Rufus with Alexa+ to combine product knowledge with shopping history and personalization, and it says Rufus reached over 300 million customers before the change, according to Amazon’s own announcement. CNBC reported that the standalone Rufus chatbot was discontinued the same day Alexa for Shopping launched. The recommendation logic, the data sources, and every optimization tactic below carry over unchanged. We are keeping the word “Rufus” throughout this article because that is still what most sellers search for, and Amazon itself has said the two names refer to the same system.

Amazon Rufus vs A10: The Core Difference

Put side by side in the Amazon Rufus vs A10 comparison, the two systems are not really in competition. They are answering two different questions.

A9, A10, and COSMO all reward some form of relevance, but only Alexa for Shopping actually reads your reviews and Q&A before deciding whether to recommend you. That is a meaningfully different bar to clear, and it is why a listing that ranks well in search can still get skipped in a Rufus conversation. If your listing is vague or your attribute fields are half-empty, the assistant does not guess in your favor. It just skips you and recommends a competitor with a fuller data profile.

Which System Controls Rankings vs. Recommendations

In the Rufus vs A10 debate, think of it as two separate lanes running in parallel, not a relay race where one hands off to the other 

A10 controls the search results page. Type a query and get a ranked grid; click one. That page exists whether or not Alexa for Shopping is involved. Alexa for Shopping controls a different surface entirely, the conversational layer that sits in the search bar, on product pages, and inside the app. When a shopper asks a question instead of typing keywords, Amazon routes them into that conversation, and Alexa for Shopping mentions a short, curated list of products, not a fifty-item results page.

That distinction matters more than most sellers realize. A10 is a ranking exercise: you can sit at position 40 and still get found by a patient shopper scrolling. Alexa for Shopping behaves closer to a binary cut. If your product is not in the short list the assistant chooses to mention, you simply do not exist for that specific conversation. No Amazon-confirmed number exists for exactly how short that list is. That is a different kind of pressure than traditional SEO.

Keyword SEO vs. Generative Engine Optimization (GEO)

Traditional Amazon SEO optimizes for one thing: does your listing contain the words a shopper is likely to type. Generative Engine Optimization, or GEO for short, optimizes for a different question: does your listing give an AI system enough clear, structured, trustworthy information to confidently recommend your product in a conversation?

The two are not opposites. GEO does not replace keyword research; it adds a layer on top of it. A title built purely for keyword SEO might read “Stainless Steel Water Bottle Vacuum Insulated Metal Bottle BPA Free,” packing in every synonym a keyword tool suggested. A title built for GEO on top of that same keyword base reads “32 oz Stainless Steel Insulated Water Bottle with Straw Lid for Hiking, Gym, and Travel.” ” Same core keywords, but now the title also answers “who is this for” and “what is it for,” which is exactly what Alexa for Shopping is trying to extract when it reads your listing.

Practically, GEO shows up in three places on a listing: bullet points written as complete, benefit-and-context sentences instead of keyword fragments, backend attribute fields filled out completely instead of left blank, and Q&A or review content that reinforces the same claims your title and bullets make. Keyword SEO alone gets you into the A10 results page. GEO is what gets you mentioned by name when a shopper asks Alexa a shopping question instead of typing one.

How to Optimize for A10

Traditional Amazon SEO fundamentals have not gone anywhere. If you skip these, no amount of AI-friendly copy will save your ranking.

• Front-load your title. Put your primary keyword and one key differentiator in the first 80 characters, since mobile shoppers see the least of your title before it truncates, and mobile makes up the majority of Amazon’s traffic.
• Use backend search terms for what your visible copy can’t fit. Synonyms, misspellings, and secondary keywords belong here, not stuffed into your bullets. 
• Protect your conversion rate before chasing more traffic. A higher-converting listing outranks a listing with more total sales but a lower conversion rate, so fixing images and copy usually moves rank faster than adding ad spend.
• Make your main image work for click-through, not just compliance. A clean white background and a properly zoomed product photo affect whether shoppers click through from the results page at all, which is a direct A10 input.
• Drive external traffic where it makes sense. Google Ads, social, and influencer links carry ranking weight under A10 that they didn’t under A9.
• Keep your account health clean. Return rate, feedback score, and account tenure now factor into seller authority, one of the more overlooked A10 signals.

How to Optimize for Alexa for Shopping (Rufus)

This is where most listings fall short, not because sellers don’t know it matters, but because it takes more structural work than swapping a few keywords.

• Fill every backend attribute field, not just the required ones. Material, target audience, intended use, compatibility. Alexa for Shopping draws on these directly, and an empty field is a question the assistant simply cannot answer about your product.
• Write bullets as complete sentences, not fragments. “Keeps drinks cold for up to 24 hours during hiking, commuting, and workouts” gives the assistant context a fragment like “Vacuum insulated” never will.
• Match your title, bullets, description, and A+ Content exactly. If your title says “100% cotton” and your description says “cotton blend,” that inconsistency reads as a red flag, not a minor typo, and it can lower how confidently the assistant recommends you.
• Put readable text on your secondary images, not just your infographics. The assistant reads text inside images, so a callout with a real spec or use case gives it something to cite that a lifestyle photo alone can’t.
• Seed your Q&A section proactively. Don’t wait for shoppers to ask. Post the 8 to 10 questions your reviews and support inbox already show people asking, and answer each one with real detail.
• Use your A+ Content as a knowledge base, not just a design gallery. Comparison charts and detailed text blocks give the assistant material to cite when a shopper asks how your product differs from another.

Before and After: One Listing, Rewritten for Both Systems

Theory is easy to nod along to and hard to apply. Here’s what the shift actually looks like on a real listing, a stainless steel water bottle.

Old style, A10-only thinking:

Title: Stainless Steel Water Bottle Vacuum Bottle Metal Bottle BPA Free

Bullets:

• Stainless steel
• Vacuum insulated
• BPA free
• Leakproof
• Durable

This title and these bullets will index for the right keywords. That’s the whole strategy, though, and it stops there. Nothing here tells Alexa for Shopping who the bottle is for or what problem it solves, so it has almost nothing to work with when a shopper asks “what water bottle should I get for hiking in summer.”

Rewritten for both systems:

 32 oz Stainless Steel Insulated Water Bottle with Straw Lid, for Hiking, Gym and Travel

• Keeps drinks cold for up to 24 hours during hiking, commuting, and workouts, so you’re not carrying warm water halfway through a trail. (A10: keeps “insulated” and “cold” indexed. GEO: answers the “how long” question directly.)
• Leakproof lid design prevents spills inside backpacks and gym bags, tested for daily commuting. A10: “leakproof” stays indexed. GEO: names the exact use case a reviewer would describe.
• Fits most standard car and bike cup holders, including narrow mounts. (GEO: answers a compatibility question the old bullets never addressed.)
• BPA-free 18/8 stainless steel construction, the same grade used in most premium insulated bottles. A10: keyword coverage. GEO: gives the assistant a specific material fact to cite.
• Built for hikers, commuters, and gym-goers who need a bottle that survives being dropped, refilled, and thrown in a bag daily. (GEO: directly names the audience, which is what COSMO maps intent against.)

The keywords didn’t disappear. Every word from the old version is still present somewhere in the new one. What changed is that each line now also answers a question a shopper might ask an AI assistant, not just a query they’d type into a search bar.

How Reviews and Q&A Influence Recommendations

Alexa for Shopping reads your reviews before it reads your ad copy’s promises. If enough reviewers describe your product in terms that contradict your listing (a “beige” product repeatedly called “yellow” in reviews, for example), the system can effectively override your own claims. The fix isn’t complicated. Monitor what language customers actually use in their reviews and Q&A, then work that same language into your bullets and A+ Content so the two sources reinforce each other instead of quietly disagreeing.

Common Mistakes When You Optimize for Only One System

The two most common failure patterns look almost identical from the outside, but they come from opposite mistakes.

Over-indexed on A10, invisible to Rufus. The listing ranks fine in search and converts decently but never gets mentioned when a shopper asks Alexa a direct shopping question. Usually the cause is empty backend attributes and bullets written as keyword fragments instead of sentences.

Over-indexed on Rufus, weak on A10. The listing reads well and answers every use case question but doesn’t show up on page one for the core keyword. Usually the cause is a title leading with a benefit statement instead of the primary keyword, or backend search terms left blank because natural language copy was assumed to cover it.

Neither mistake is really about effort. Both come from sellers picking a side instead of treating this as one job with two audiences.

Vendor Central vs. Seller Central: Does This Change Anything?

Not the underlying tactics, but it does change what you can control. Vendor Central accounts often have access to a wider range of A+ Content modules and Amazon-managed advertising placements, which gives more surface area for GEO-style content. Seller Central accounts have more direct control over backend attributes and can update listings faster without waiting on Amazon’s content team. Neither setup is disqualifying for either algorithm. It just changes which levers are fastest to pull.

A Future-Proof, Ongoing Amazon SEO Strategy

Whether you’re weighing Amazon Rufus vs A10 priorities or just trying to keep both systems happy, treat this as a recurring audit, not a one-time fix. Once a quarter, pull your top 10 listings by revenue and check three things: are all backend attributes still filled in, does your bullet copy still match what your most recent reviews say, and has a new competitor entered your category with a more complete GEO profile than yours. Amazon updates both systems continuously, and a listing that was fully optimized in January can quietly fall behind by summer without a single word of your copy changing.

If you want a second pair of eyes on how your top listings perform in the Amazon Rufus vs A10 landscape, our Amazon listing copywriting team can audit them and show you exactly where the gaps are. 

Key info