In February 2024, Amazon announced the launch of its AI-powered conversational shopping assistant, Rufus. Named for the company’s favorite corgi, Rufus lets sellers search for the products they need by typing or speaking full sentences to a chatbot designed to make helpful recommendations. Rufus is currently in closed beta testing, available to certain users through the Amazon app.
With the rollout of Google’s AI overviews, AI search is already changing the game. Marketplaces from Walmart to Alibaba have recently introduced similar conversational shopping assistant offerings.
In the scheme of things, that puts Amazon a bit behind the game and — if it hasn’t been released by the time you read this — we can almost certainly expect a fully-public-facing Rufus this year. It’s not clear whether Rufus will replace ordinary search; the two will likely exist side-by-side for a while.
For sellers, the switch to conversational search means you might have to loosen your grip on keyword-based algorithms,emphasizing natural human speech and writing patterns. Today we’ll explore what makes Rufus different and what it could mean for your growing Amazon business.
What Is AI Search?
Before we dive into the specifics of Rufus, let’s quickly recap AI search.
AI search relies on a host of new technologies, like machine learning and natural language processing, to make sense of complex user queries. Unlike conventional search, which relies on a string of keywords to deliver straightforward results, AI can “talk” to users.
In e-commerce, AI search can use past customer interactions, deep knowledge of a product catalog, and external data (like Google search results) to deliver nuanced interpretations of user intent and context. Then, it can use that understanding to generate in-depth, personalized recommendations.
What Is Rufus?
Rufus is an AI-powered chatbot, or virtual shopping assistant, that uses AI search technology to help users find products more easily on Amazon. Trained on a variety of data sources, Rufus is designed to make intelligent recommendations that take specific user needs and context into account.
For example, you can ask Rufus for nuanced assistance, with things like:
- Recommendations: “What should I get my mom for Mother’s Day?”
- Comparisons: “What’s better, glass or metal water bottles?”
- Options: “What are the top 10 styles of in-ear headphones?”
- Style Advice: “Which dresses would be appropriate for a 40-year-old office worker?”
- Considerations: “What factors should I consider when choosing an elliptical machine?”
- Task Planning: “What items should I buy if I want to start making pastries at home?”
How Does Rufus Work?
Like other generative AI tools, Rufus “learns” from big sets of data, synthesizing information from all these sources in order to make future predictions, recommendations, and decisions.
According to Amazon, Rufus is trained on Amazon’s product catalog, customer reviews, community Q&A’s, and “information from across the web.” That means, when it answers a user query, it might be extrapolating data from any of these sources.
It’s very much worth noting that, though Rufus is built into Amazon, it’s (at least for now) not confined to just being a personal shopping assistant. In other words, you can treat it like ChatGPT or Gemini, asking it whatever questions you want. It won’t engage with you over unsafe or questionable topics, but you may be able to chat with Rufus about history or get it to do your homework.
Amazon app users in the current wave of beta testing for Rufus may see an option in the regular search bar to “search or ask a question.” Users can also trigger it by swiping up from the bottom of their screens. Some Reddit users complain that Rufus shows up as an unwelcome pop-up that apparently cannot be disabled.
What Else Do We Know About Rufus?
One TechCrunch analysis notes that the beta version of Rufus may struggle with more nuanced recommendations, not necessarily taking into account all of the user’s criteria for a product. This same article also notes that Rufus may speak negatively of products in Amazon’s catalog, calling out products that may fit poorly or be “outdated.”
A Yahoo!Finance review found that a beta version of Rufus gave great recommendations for fitness equipment and some travel queries, but somewhat irrelevant recommendations when discussing video games. It also struggled with differentiating between dining plates and weightlifting plates.
When Axios put Rufus through some rigorous questioning, a few concerning results came up: the bot didn’t immediately reject queries for white supremacist or Christian nationalist products.
However, The Wall Street Journal noted Rufus does a better job of anticipating potentially problematic queries and redirecting querants toward more positive uses of their energy compared to other chatbots.
What are the Potential Benefits of Rufus?
With an ever-increasing amount of noise on Amazon, it’s harder than ever for customers to find your products. Add to that the way Amazon prioritizes Sponsored content and other paid resources, and it can feel impossible for a growing brand to get noticed.
AI search could help brands cut through the noise, while simultaneously improving the customer experience. Instead of sifting through dozens of identical-looking product thumbnails, an AI search tool might quickly connect users to the best products to meet their needs.
What are the Downsides to Rufus?
While beta versions of Rufus don’t seem to give preferential treatment to certain products or brands (not even Amazon products), it’s possible the full rollout could prioritize sponsored brands just like the current search interface.
Technical issues or a fundamental inability to understand the user’s question could also lead to search issues. This could be an even bigger problem if Amazon removes the ability to use conventional search as well.
6 Ways to Prepare for Amazon’s Rufus Rollout
Since AI search doesn’t rely on traditional keywords, you may have to upgrade your Amazon strategy, following a new set of rules to get your brand noticed. Here are some ways to make sure you’re prepared for the switch to AI search and how it could spell more sales for your business.
1. Ask the Right Questions
AI search tools like Rufus don’t just scan your listings to find the most relevant keywords. Instead, they use advanced algorithms to actually read your listings, product reviews and customer Q&A, and bolster their findings with information from across the web.
For the best chance at appearing in AI search results, put yourself in your customers’ shoes. Consider all the potential questions or sentence fragments that might lead them back to your products, and make sure your listings answer those questions in natural language.
For example, let’s say you sell a sunrise alarm clock that slowly gets brighter in order to wake up the user. With ordinary SEO, the user might type in “sunrise alarm” or “bright alarm clock” or “quiet alarm clock,” or something else to that effect.
With AI search, they might type or speak in much more general terms, sometimes even seemingly unrelated sentences or fragments, like:
- “Products to regulate my circadian rhythm?”
- “Non-hormonal melatonin alternatives”
- “Alarm that won’t wake up my partner”
- “Ways to wake up slower”
- “How to make my home more serene”
- “Seasonal affective disorder alarm”
When creating your product listings, consider all the potential reasons a user might need your product — not just the most obvious.
2. Use Amazon AI Search Best Practices
While your listings should be written in natural language, they also need to consider how algorithms operate.
For example, AI search and other generative AI technologies tend to prefer text and pages that answer user questions early and in plain language. It also might help to learn about key search tactics like structured data and schema markup to help the bots process your listings.
Always follow Amazon’s latest listings guidelines and update all product pages for optimal retail readiness. As Rufus continues to evolve, Amazon will very likely use it to further prioritize the listings that best meet its standards. Be sure to take advantage of all Amazon listings optimization tools and adhere strictly to any structural guidelines.
3. Keep Tabs On Your Reviews and Q&As
One standout feature of Amazon’s Rufus is that it considers data from customer reviews and user Q&A answers in making AI search product recommendations. Amazon has already deployed similar technology in a more transparent way with its review summaries on product listings pages.
To keep your summaries positive, you’ll want to make sure you get as many (good) product reviews as possible. Among other strategies, you can get more product reviews by:
- Using the “Request a Review” button
- Leveraging Amazon Vine
- Enclosing package inserts with your products
- Requesting reviews on social media
- Reaching out gently to critical reviewers
And of course, you’ll need to make sure your product listings are as accurate as possible, to avoid disappointing customers.
4. Don’t Forget Voice
Rufus can’t talk yet (nor bark, for that matter), but users can speak their questions and it’ll answer them in text format. Given other AI trends, it seems likely that it will eventually become capable of full voice conversations, and could be integrated with (or, gasp, replace) Alexa.
To keep your customer engaged, prepare for situations where Rufus might read your product listing aloud. Consider the flow of your content, keep sentences short, and try not to put your customer to sleep.
5. Stay Flexible
Remember, this is just a beta rollout, and Rufus is bound to undergo many changes as it develops. Rather than putting all effort behind meeting specific guidelines or trying to hack your way to the top, stay focused on delivering a great product and keeping listings up to date with product changes and Amazon’s requirements.
Use all the Amazon tools available, including:
- A+ Content
- A/B Testing
- Amazon Vine
- Automated Pricing
- Amazon Brand Registry
6. Watch Your Seller Ratings
As always, Amazon will almost certainly continue to prioritize sellers and products with fast shipping, few returns, and great customer service.
To stay relevant, keep your Amazon Seller Rating and Seller Feedback Rating in tip top shape by:
- Continuously optimizing your listings to represent your products clearly
- Optimizing your inventory planning to prevent stockouts and know when to push promotions
- Addressing customer issues promptly
- Requesting removal of negative seller feedback
- Getting more customer feedback and putting it into practice
When in doubt, read your customer reviews and compare them with your competitors’. Then, do everything in your power to improve your product and your listings and offer a better experience.
Raise the Rufus
Like a corgi’s bark, Amazon’s Rufus could cut through all the background noise and bring your perfectly-optimized listing right to your customers’ heels. With the right preparation and tools, you’ll have no trouble staying up to date and ahead of any ongoing changes.
To unleash your Amazon potential, subscribe to our SellersFi newsletter and get free access to the latest Amazon tips and resources designed to help you scale your business, your way.