For most of the internet’s history, product discovery followed a familiar pattern.
A shopper searched for something. Google, Amazon, or another marketplace returned a list of results. Brands competed to appear near the top, and the shopper clicked through several options before deciding what to buy.
That system still matters. But in 2026, another layer is rapidly forming between the shopper and the product.
Instead of searching “best running shoes for women,” a customer can ask an AI assistant, “I run three times a week, have slightly wide feet, and want something lightweight under $150. What should I buy?”
The difference looks small. It isn’t.
Traditional search helps shoppers find options. AI increasingly helps them narrow those options down.
For brands, that changes the challenge from simply being found to being understood well enough to be recommended.
Search Isn’t Disappearing. Discovery Is Changing
There is a temptation whenever a new technology arrives to declare the old one dead. That would be a mistake here.
Google Search still drives enormous amounts of commercial traffic. Marketplaces remain major product discovery engines. SEO, paid search, social platforms, creators, and retail media are not disappearing because consumers have started asking ChatGPT or Gemini what to buy.
What is changing is how those channels fit together.
Google has described AI Mode shopping journeys as a move from keywords toward natural conversations. Its Shopping Graph contains more than 50 billion product listings, with information including prices, reviews, colors, and availability. AI Mode can use that data to help shoppers narrow down products based on much more specific requests.
OpenAI is moving in a similar direction. In March 2026, it expanded product discovery capabilities for merchants, allowing brands to provide structured product feeds so their catalogs can be represented more accurately when shoppers explore and compare products in ChatGPT.
The shopping journey is becoming conversational.
From Keywords to Questions
Traditional ecommerce search has largely trained brands to think in keywords.
“Best moisturizer for dry skin.”
“Wireless headphones.”
“Carry-on suitcase.”
Those searches communicate what someone wants, but very little about why they want it.
Conversational AI allows shoppers to provide significantly more context.
Someone can ask for “a moisturizer for sensitive skin that feels lightweight, works under makeup, and doesn’t contain fragrance.” Another shopper can ask for “a carry-on that fits European airline limits, has a laptop compartment, and is easy to lift into an overhead bin.”
Those aren't simply longer keywords. They contain constraints, preferences, situations, and intent.
For an AI system trying to recommend products, knowing that a suitcase exists is no longer enough. It needs enough reliable information to determine whether that suitcase satisfies the specific conditions the shopper has described.
That is where product discovery starts to look different from traditional SEO.
Ranking and Recommendation Are Two Different Problems
SEO has traditionally focused on a fundamental question: can a search engine find and rank your page for the right query?
AI product discovery introduces another question: does the system understand your product well enough to confidently include it in an answer?
Consider two skincare products.
Both might be optimized for “face moisturizer.” Both could rank for similar keywords. But one product page clearly explains that the formula is fragrance-free, designed for sensitive skin, lightweight, non-comedogenic, and suitable under makeup.
The other simply says it provides “premium hydration for healthier-looking skin.”
To a human shopper, the first description is more useful. To an AI system trying to answer a highly specific shopping question, it also provides substantially more information to work with.
That does not mean brands should start stuffing pages with awkward phrases designed for AI.
It means specificity is becoming more valuable.
Product Pages Are Becoming Data Sources
For years, ecommerce brands have treated product pages primarily as destinations.
Get someone onto the page, then convince them to buy.
Increasingly, those pages also function as sources of information for systems helping customers decide what to buy before they ever arrive.
Product titles, descriptions, structured attributes, prices, availability, reviews, FAQs, imagery, merchant feeds, and information elsewhere on the web can all contribute to the broader digital picture surrounding a product.
This is why consistency matters more than it used to.
If a product is described one way on the brand website, another way on a marketplace, and differently again by retailers or other sources, understanding the product becomes harder. Clear attributes and consistent positioning make the product easier for both people and machines to interpret.
The same principle applies to creative.
A beautiful lifestyle image may create desire. An image that also clearly communicates size, use case, material, compatibility, or another meaningful attribute can help remove uncertainty.
The goal isn't to turn every image into an infographic. It is to make sure the complete product experience communicates enough useful information to support a buying decision.
AI-Referred Shopping Traffic Is Still Small, But Growing Fast
The scale of this change deserves some perspective.
AI referrals have not overtaken search, social, email, or paid advertising. For most ecommerce businesses, those channels will remain much larger sources of traffic today.
But the growth rate is difficult to ignore.
Adobe reported that traffic from generative AI sources to U.S. retail websites increased 393% year over year during the first three months of 2026. During the 2025 holiday season, AI-driven retail traffic had grown even faster.
The more interesting question may not be how much traffic AI sends, but what happens before that traffic arrives.
A customer coming from traditional search may still be exploring ten possible products. A customer who has spent several minutes discussing their requirements with an AI assistant may arrive after much of that comparison has already happened.
The consideration process is moving upstream.
That means brands increasingly have to influence the decision before the click, not only after it.
What Brands Should Do Differently in 2026
The wrong response to this shift would be to abandon SEO and chase the latest acronym.
The better response is to make products easier to understand everywhere they appear.
Product pages should contain specific attributes instead of vague marketing language. FAQs should answer the questions customers actually ask. Product feeds should be complete and current. Claims should remain consistent across websites, marketplaces, retailers, and other channels.
Brands should also pay more attention to the information surrounding them outside their own websites. Reviews, credible editorial coverage, community discussions, creator content, and other third-party sources can contribute to the wider context customers and AI systems encounter when researching a product.
Most importantly, brands should test what AI assistants actually say about them.
Ask ChatGPT, Gemini, Google AI Mode, and marketplace assistants the same questions your customers would ask. See which brands appear. Look at how your product is described. Identify important attributes that are misunderstood or missing.
That won't reveal a secret ranking formula.
It will reveal whether your digital presence clearly communicates what you sell, who it is for, and why someone should choose it.
The Bigger Shift
For twenty years, digital marketing has been built around earning the click.
Rank higher. Increase impressions. Improve click-through rates. Drive more traffic.
Those metrics still matter.
But AI-mediated discovery introduces another layer of competition before the click ever happens.
A shopper may begin with twenty possible products and ask an AI system to narrow them to three. If your product never makes that shortlist, the quality of your landing page may not matter because the shopper never reaches it.
That is the fundamental change brands need to understand.
The old internet rewarded brands that were easy to find.
The next phase of product discovery may increasingly reward brands that are easy to understand, compare, and confidently recommend.
SEO isn't disappearing.
But being searchable is no longer the whole game.
The bottom line
Product discovery is no longer just about ranking for the right keywords. As AI becomes part of the shopping journey, brands also need to make their products easy to understand, compare, and recommend.
SEO still matters. What is changing is what happens before the click. AI can help shoppers narrow dozens of options to a shortlist before they ever visit a product page.
The brands that adapt will optimize not just to be found, but to be understood.
If your product creative needs to keep up with how discovery is changing, Dobby Ads works with ecommerce brands on product images, video, 3D, marketplace content, and creative strategy.
