July 2026 — AI & Ecommerce news.
What changed, as it happens.
Tracked daily from primary sources. The items below are what we're watching for clients right now.
Google’s own data: people ask AI about buying things ~3x more than they actually shop
Google and DeepMind compared 14.65 million non-work conversations from the Gemini app, AI Mode in Search, and the Gemini API (April 6–19, 2026) against the American Time Use Survey. Consumer-purchase topics show up in AI conversations roughly three times more often than Americans spend time shopping, with far larger gaps for government services (~20x), professional and personal-care services (~7x), and education (~5.8x). The report carries no click data, so it says nothing about whether those conversations sent anyone to a website — but as a demand signal for being present inside AI answers, this is Google measuring its own surface.Source: Search Engine Journal →
Our take: People research purchases in AI at three times the rate they actually shop, and most of it never produces a click for anyone. Being the source the answer quotes is the only distribution that survives that math, which is what all the schema and citability work on this site is for.
Read our full take →Google starts rolling out Connected Apps in AI Mode
Google began rolling out Connected Apps in AI Mode Search for US users, letting it hand off tasks to third-party services like Instacart, Canva, and YouTube Music through MCP-based connections. It extends AI Mode from answering questions toward completing actions across the apps people already use — another sign the AI answer surface is becoming a place transactions start, not just where research ends.Source: Search Engine Journal →
Our take: Each connector makes AI Mode more of a starting point and less of a search page. The near-term play for stores is unchanged: be the citable source. But watch which commerce apps get connectors next; that's the transaction path being assembled in public.
Read our full take →AI Mode can now write to your Google Calendar, not just read it
Google connected Calendar to Personal Intelligence in AI Mode for US users, so it can add invites directly and factor an existing schedule into its answers. It’s the first Personal Intelligence connection that creates entries rather than only referencing data — a step from AI Mode answering questions toward acting on them.Source: Search Engine Journal →
Our take: First write access is the pattern to note: AI Mode is accumulating permission to act, not just answer. As that reaches commerce, the stores agents can act on will be the ones with clean, structured, machine-readable data, the same work that earns citations today.
Read our full take →Lantern launches a platform that scores how AI shopping portals read your catalog
Lantern released a platform that uses AI agents to predict how AI shopping portals interpret a merchant’s product catalog, flag visibility constraints, and apply fixes, with an “agent-readiness score” to gauge how well a catalog surfaces in AI-driven shopping. It’s part of a fast-growing category built around making product data legible to AI buyers.Source: Practical Ecommerce →
Our take: You don't need the tool to use the idea: score your own catalog the way an agent reads it — complete attributes, specs as on-page text, markup that agrees with the page. Third-party scores are fine; the fixes are the point.
Read our full take →Alli AI ships a WordPress plugin that serves AI crawlers a pre-rendered HTML copy
Alli AI launched a WordPress plugin that serves AI crawlers a fully pre-rendered HTML version of each page — while human visitors see the normal site — and auto-deploys structured data site-wide. It’s a bet that AI-crawler-specific rendering is worth the parallel-version maintenance Google’s own team keeps cautioning against.Source: Practical Ecommerce →
Our take: We'd skip this one. Serving crawlers a different page than humans see is cloaking-adjacent maintenance debt, and Google's own team keeps warning against parallel AI versions. Make the one real page machine-readable instead.
Read our full take →BigCommerce/PayPal survey: shoppers want AI to help them discover, not buy for them
A BigCommerce and PayPal survey of ~3,000 consumers across the US, UK, and Australia found about two-thirds interested in agentic shopping tools — but with a strong preference for AI handling product discovery rather than autonomous purchasing. Fraud protection and purchase confirmation topped the list of trust requirements.Source: BigCommerce →
Our take: Discovery-not-checkout is the design brief. Shoppers want AI to find and compare, then want control at purchase. Winning the comparison layer (spec tables, honest differentiators, real answers) pays off before agentic checkout plumbing does.
Read our full take →Square launches ChatGPT and Claude integrations so AI can order from its sellers
Square released a ChatGPT app and a Claude plugin that let AI platforms discover Square sellers and place orders, starting with US food-and-beverage merchants and with no added marketplace commission. Square is also working with Amazon on an Alexa+ integration and sits on the AAIF, W3C, and UCP standards groups — another platform wiring its catalog into AI checkout.Source: Digital Commerce 360 →
Our take: Food-and-beverage first, but the template is the story: platform wires catalog in, AI orders out, no added commission. BigCommerce already names OpenAI and Perplexity as channels in its API. When the switch appears for your store, catalog quality decides how that goes.
Read our full take →Klaviyo moves its marketing and service AI agents to public beta
Klaviyo pushed two AI agents to public beta: Composer, a campaign builder, and Customer Agent, a service-automation tool, with real-time customer data shared between them. Early-access brands include Spanx and Dermalogica.Source: Digital Commerce 360 →
Our take: On Klaviyo? Run the beta on one low-stakes flow, supervised. Hold both agents to the standard you'd hold a junior hire: review everything they send until they've earned autonomy, because AI email at scale amplifies whatever it's given, including mistakes.
Read our full take →
Custom ecommerce software, AI apps, and SEO — the work other agencies quote around, built in-house.
20+ years of BigCommerce engineering, now AI-augmented. Tell us your store, your stack, and your deadline — we quote fixed scope on the first call.