This isn't about any one platform. It's a genuine shift in how people find products at all, and it's already mainstream, not an early trend worth watching from a distance.
The Numbers Behind the Shift
NIQ (Nielsen IQ), in a report published in August 2026 with World Data Lab, found that 74% of shoppers now use AI for some form of product discovery. That's not an early-adopter statistic. It describes most shoppers, most of the time, which changes the calculation for any business still treating this as optional or experimental.
Visual Search Is Already Enormous, Not Emerging
Google's own VP of Global Advertising stated in March 2026 that Google Lens now handles more than 25 billion visual searches a month, with roughly 20% of those carrying genuine commercial intent, someone trying to identify or buy a product from a photo rather than just satisfying curiosity. At that scale, a shopper pointing a camera at something and asking "where can I get this" is a mainstream behavior, not a novelty feature.
What's Actually Changing: Search-and-Sift Becomes Ask-and-Act
The old pattern was search, browse several sites, compare manually, then decide. The emerging one compresses most of that into a single exchange: describe what's needed, let an assistant narrow the options, evaluate what comes back, act. The research phase that used to take multiple sessions across multiple sites increasingly happens once, inside one conversation. That's a genuine change in how a purchase decision gets made, not just a new interface on top of the same behavior.
What This Means If You Sell Products, Whatever Platform You're On
This isn't specific to any one platform. The specific technical mechanics for Shopify stores, how Google's Shopping Graph actually reads product feed data, are covered in detail in our guide on Shopify SEO: What Actually Moves the Needle. The underlying shift here applies just as much to a business selling through its own website, a marketplace listing, or an in-store catalog with an online presence. The common thread across all of them: an AI system making a recommendation is reading and comparing structured product data, materials, dimensions, real use cases, genuine attributes, not inferring quality from marketing copy the way a human scanning a page might. A product page that reads well to a person but has thin or missing structured data is increasingly invisible to the system actually making the recommendation.
A Practical Way to Think About This, Instead of Panic-Redesigning Everything
None of this means every product business needs to rebuild its site or chase a new platform trend. The more useful exercise is to audit the questions a shopper, or an AI assistant acting on their behalf, would actually need answered before buying: what it's made of, what size or configuration fits their situation, what it's compatible with, what the return policy actually says, whether real buyers found it useful for their specific use case. If that information exists only as a paragraph of marketing copy on a product page, or worse, only in a PDF spec sheet or a sales rep's head, it's effectively invisible to a system that has to compare hundreds of options in seconds. The fix is rarely a redesign. It's usually filling in the structured data fields that already exist on most ecommerce platforms but sit half-empty because they don't visibly affect how a page looks to a human visitor.
Want to know how discoverable your products actually are to AI-driven shopping search?
Get Your Visibility Audit →Frequently Asked Questions
How many shoppers actually use AI for product discovery now?
Nearly three-quarters. NIQ (Nielsen IQ), in a report published in August 2026 with World Data Lab, found that 74% of shoppers now use AI for some form of product discovery, a mainstream behavior at this point, not an early-adopter trend.
How big is visual search for shopping already?
Large and growing quickly. Google's own VP of Global Advertising stated in March 2026 that Google Lens now handles more than 25 billion visual searches a month, with roughly 20% of those carrying commercial intent, someone looking to identify or buy a product from an image, not just satisfy curiosity.
What does "ask and act" mean for how people shop?
It describes the shift from browsing multiple sites and comparing options manually (search and sift) to asking an AI assistant to narrow choices, summarize options and make a recommendation in a single conversation (ask and act). The research phase that used to take multiple sites and multiple sessions increasingly compresses into one exchange.
Does this only apply to Shopify stores?
No. This shift affects any business that sells physical products, regardless of platform. The specific technical mechanics for Shopify stores, including how Google's Shopping Graph reads product feed data, are covered separately in our guide on Shopify SEO; the behavioral shift covered here applies just as much to a business selling through its own website, a marketplace, or in-store with an online catalog.
What should a product business actually do about this?
Prioritize structured, complete and accurate product data, materials, dimensions, use cases, genuine attributes, over keyword-stuffed titles and descriptions. AI systems making a recommendation are reading and comparing actual product data, not inferring quality from marketing copy the way a human browsing a page might.