AI-referred traffic to U.S. retail sites grew 269% year-over-year in March 2026. That number came from Adobe’s own data, announced at Adobe Summit. It’s the kind of stat that sounds impressive until you sit with what it actually means: a meaningful and growing share of visitors arriving at Commerce stores didn’t search for anything. An AI agent made a judgment call about whether to recommend the store, the product, or the brand – and sent them there.
That judgment call happened somewhere outside your analytics. You didn’t influence it with a bid. Your meta descriptions didn’t factor in.
This is the problem Adobe is now building against, and the Commerce-specific announcements from Summit are worth paying attention to – not because they’re finished products, but because they signal where the platform is heading and what that demands from the underlying store.
The Gap Adobe Is Trying to Name
Most Adobe Commerce stores have spent years optimizing for search. Clean URLs, structured data, fast PDPs, descriptive product attributes. Good SEO practice. The problem is that LLMs don’t read stores the way crawlers do.
A crawler indexes. An LLM reasons. When a customer asks an AI assistant “which B2B supplier has the best lead times for industrial fasteners,” the AI isn’t pulling a ranked list from an index. It’s drawing on whatever it has been trained on or can retrieve – and making a judgment about which brands it can confidently recommend without hallucinating specs.
Adobe’s new Sense layer, built across AEM and Commerce, attempts to surface exactly this kind of gap: where AI systems misinterpret your products, where your catalog is invisible to LLM-powered discovery, and where structured content exists on the page but doesn’t communicate what an agent needs to act on it.
The framing Adobe used at Summit – “brands need to manage context, not just content” – is a reasonable way to put it. The practical implication is harder: most Commerce stores aren’t built with this in mind, because until recently there was no reason to be.
Conversational Checkout Is the Logical Endpoint of B2B Self-Serve
The Brand Concierge update is the announcement that will age well.
The premise: a customer interacts with an AI interface, asks a product question, gets a contextually accurate answer drawn from live catalog data, and completes checkout inside that conversation. No redirect. No new tab. The transaction happens in the AI surface.
For B2C this is interesting. For B2B it’s more significant. Gartner has been tracking the shift toward rep-free buying for years – 61% of B2B buyers prefer a buying experience without a sales rep involved. Conversational AI is the infrastructure that makes that preference scalable. The buyer gets a knowledgeable “rep” available at any hour who can pull accurate pricing, check availability, and process an order. The merchant doesn’t staff that interaction.
The catch – and it’s a real one – is that this only works if the underlying product data is trustworthy enough for an AI to answer questions from it without inventing details. An LLM fed vague product descriptions, missing attributes, or outdated pricing doesn’t become a capable sales agent. It becomes a liability.
The merchants who will absorb Brand Concierge quickly are the ones who already have clean, enriched catalogs. The ones who don’t will need to do that work first, and it’s not a small project.
What Adobe Announced Is Only as Good as Your Implementation
None of these Adobe Commerce features are plug-and-play on a store that hasn’t been maintained properly.
Catalog enrichment, structured data quality, PDP performance, and content accuracy have always mattered for SEO and conversion. They now also determine whether your store is legible to AI systems at all. A Commerce store running on a solid technical foundation – clean architecture, current version, well-maintained extensions, rich product attributes – will absorb these capabilities without significant friction. A store that’s been patched over multiple Magento versions, with legacy customizations and gaps in product data, won’t.
This is where the gap between merchants will widen. Adobe is building features that reward stores that are already well-built. The merchants investing in technical quality now are positioning themselves for AI visibility that their less-maintained competitors won’t be able to match quickly.
If you’re not sure where your store stands, a technical or performance audit is the honest starting point. If you’re planning to build or extend for these capabilities, the architecture decisions you make now will determine how much of Adobe’s AI roadmap you can actually use.
Atwix has been working in Adobe Commerce for over a decade, holding the #1 global Magento code contributor ranking from 2018 through 2025 as an Adobe Gold Solution Partner.
If you want to understand where your store stands technically, our audit services are the right starting point. If you’re building or extending your Adobe Commerce implementation, let’s talk about what that looks like.