As Amazon ramps up its focus on artificial intelligence-powered shopping, sellers on the company’s sprawling marketplace are exploring how best to optimize their content while navigating the uncertainties that come with emerging technology.
Like many other tech giants, Amazon has been pouring money into AI, ramping up its spending on the data centers, chips and real estate needed to power artificial intelligence. The technology has surged in popularity since OpenAI’s ChatGPT burst onto the scene in late 2022. Amazon has released an array of AI-powered tools for merchants and customers alike, including a consumer-facing chatbot assistant Rufus, which the company rolled out to all U.S. users in July, ahead of its crucial Prime Day sale.
Since then, AI-driven search has become increasingly implemented across Amazon’s site. Rufus, for instance, now appears in Amazon’s main search bar, in addition to a widget at the bottom of the screen, which opens the bot’s chat interface. Users also encounter AI-generated recommendations on product pages, as well as AI summaries of customer reviews. Amazon has even started populating Rufus’s chatbot interface with ads, per Adweek.
In January, Doug Herrington, CEO of Amazon Worldwide Stores, said during a keynote presentation at NRF that AI is playing a crucial role in how Amazon operates. “AI is becoming transformative for our business, and we really haven’t had a technology revolution as large as this since the start of the internet,” Herrington said.
For customers, the rise of AI on Amazon’s marketplace represents a departure from traditional searches that rely on Boolean operators, keywords or product names to find specific items. Now, customers can find what they’re looking for with conversational phrases like, “Is this jacket machine washable?” and “What should I consider when buying a new laptop?” For Amazon sellers, the question remains how precisely to optimize for AI and the extent to which consumers will even adapt to this new form of e-commerce search.
“The way search works now is fundamentally different from traditional SEO,” said Max Sinclair, a former Amazon employee and founder of Ecomtent, an agency that helps e-commerce brands adapt to AI-driven search. “Large language models don’t operate on keywords alone; they interpret context and intent. That means brands can’t just stuff listings with long-tail keywords and expect the same results. They need to focus on structuring product information in ways AI understands.”
Even though AI is still nascent, Sinclair said interest from brands has been growing. Ecomtent has been growing topline revenue at 35% every month for the last couple of months, which Sinclair attributed to demand from brands and sellers who want to optimize for AI on Amazon.
At the end of last year, Amazon shared that customers have asked Rufus over half a billion questions so far, an Amazon spokesperson told Modern Retail in an email.
Using industry benchmarks, Ecomtent estimated that Rufus accounted for nearly 14% of Amazon’s overall searches in October. By the end of 2025, this figure could reach 35% as more consumers turn to AI, Sinclair said.
For some, optimizing for AI has led to a sales boost. Pattern, an e-commerce accelerator, has been tracking the impact of AI-driven search optimizations, and brands that revise their product listings to align with AI-driven search parameters see a median revenue lift of 15-20%, Pattern found.
A scholarly paper about Amazon’s AI search technology, authored by a team of scientists employed by Amazon’s science division and published in early 2024, found that AI-driven personalization resulted in a 0.7% increase in conversion rates — seemingly small but worth hundreds of millions of dollars in additional annual revenue, given Amazon’s scale, the authors wrote at the time.
However, AI optimization alone won’t fix a poorly performing product. “If it’s a bad product, it doesn’t help,” said Jacob Miller, vp of data science at Pattern. “You can customize content and make the best possible listing, but if customers ultimately don’t find value in the product, AI won’t sustain it in search rankings.”
As AI-powered search continues to evolve, sellers are watching closely to see how consumer adoption unfolds. Miller said the broader trend of AI-assisted product discovery extends beyond Amazon, with platforms like OpenAI’s ChatGPT and Google’s AI tools also surfacing product recommendations. Indeed, some brands are seeing sharp traffic spikes driven by Google Gemini and ChatGPT search recommendations, Modern Retail previously reported.
Optimizing for search is no longer just about using the right keywords. “Amazon’s AI can now factor in contextual elements about a shopper’s background or past behavior, even if they don’t explicitly search for those details,” Miller said. “If a shopper mentions in a query that they’re looking for a birthday gift for their mother, AI can pull from listings that have signals indicating they are good for gifting.”
As an example, Miller tested Rufus by searching for triathlon gear and found that Amazon recognized his prior searches for gear that could withstand chlorinated water, likely because of previous triathlon-related queries.
For sellers, the challenge lies in adapting to this new paradigm, especially as the technology evolves. Will Haire, co-founder and CEO of BellaVix, a marketing agency that helps brands grow on Amazon and Walmart marketplaces, has been running experiments with Rufus to better understand its impact.
“What we do know is that the AI looks at the product detail page, it looks at the question and answer section, and it looks at reviews,” Haire said.
To better align with AI search behavior, Haire’s team has been experimenting with incorporating natural language patterns into product content and tracking long-tail keyword performance. However, Haire said BellaVix hasn’t observed a correlation between AI optimization and increased sales. That’s because there are too many variables at play to isolate AI search as a significant driver of growth.
In November, Rajiv Mehta, a vp at Amazon who works on conversational AI shopping, including Rufus, told Modern Retail that one of the most important factors for sellers, when it comes to search optimization, is ensuring product listings contain accurate, up-to-date and factual information.
“One of the things we know when customers are shopping on Amazon is that they’re really trying to make informed decisions and get good information about the products they’re trying to choose and shop for,” Mehta said in an interview at the time. “With Rufus, you can just ask those questions, and Rufus is able to predict what sorts of questions that customer may have and actually surface those questions proactively to a customer.”
What I’m reading
The FTC has asked a Seattle judge to delay its case against Amazon Prime for its allegedly deceptive practice, citing staffing and budget shortfalls
Chinese authorities called in Walmart executives this week following complaints that the retailer was urging some Chinese suppliers to lower prices to offset U.S. tariffs, according to The Wall Street Journal.
TikTok’s ad revenue could top $32.4 billion in 2025, a 24.5% jump from 2024, if the U.S. does not proceed with its plans to ban the app, per Marketing Dive.
What we’ve covered