Sunday, October 18, 2026 Next update in 8h 23m 13s

— A weekly publication —

The Agentic Commerce Report

A weekly read of everything that moved in agentic commerce — protocols, payment rails, retailer pilots, regulation. Summarised, sourced, and stitched to what came before.

AEO & Answer-Engine Optimization

32 events tracked

AI-surface product discovery has crystallised across a handful of platforms in late 2025 and early 2026. Tracked weekly here: platform launches, ranking signals, and feed-spec changes that shape which merchants agents surface.

launch

Google Universal Cart Launches on Search and Gemini in the US

Google launched Universal Cart in the US on Search and Gemini, enabling checkout at six named retail partners.

Universal Cart converts Google Search and Gemini into a direct commerce channel for seven retail partners — Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, and Shopify merchants. The feature draws on the Universal Commerce Protocol (UCP, Google's open checkout standard linking retailer inventory to AI-powered commerce surfaces), which Google extended to cart transfer and enhanced checkout flow testing in September (2026-w38-aeo-google-holiday-agentic-updates). Purchases complete within Google's interface using Google Wallet for payment and identity, bypassing the retailer's checkout page. The 60-billion-product Shopping Graph supplies the product catalogue. For retailers already submitting structured feed data to Merchant Center, Universal Cart converts those AEO signals into a direct path to purchase. Stripe, as a UCP partner, handles payment processing for Shopify merchants participating in the program through its UCP partnership.

  1. Google Blog
Issue 39
research

Mastercard Report Predicts That One in 10 People Will Routinely Use AI Agents to Shop and Pay by 2030

Mastercard found 79% of European teens used AI for product research last year in a 13-market, 26,000-person survey.

The survey measured adoption across 26,000 respondents in 13 European markets, segmenting by age to isolate teen behavior from the broader adult consumer population. The 62% monthly-use figure implies AI product research is a routine behavior for the cohort, not a trial. The 61% openness to a comprehensive AI shopping assistant — one that handles discovery, comparison, and checkout — aligns with the capability set Mastercard's Agent Suite for Merchants deploys commercially (2026-w38-pilots-mastercard-tripcom-agentic). The report also projects that one in 10 online shoppers will routinely delegate purchases to AI agents by 2030 — the first named adoption target from a major card network's research division. The 26,000-person sample size makes this the largest published consumer AI shopping survey from a payments network to date. AEO teams at European retailers are the primary audience for this research.

  1. Mastercard
Issue 39
research

The economics of AI monetization

Stripe found a 24x price spread when pricing is hidden from AI agents, in an analysis of 7,600 AI responses.

Stripe's analysis covered 7,600 AI responses generated by agents browsing product and pricing pages. The 24x price spread measures the range between the lowest and highest prices an AI agent reported when pricing was not presented in a machine-readable format — a direct measure of the commercial risk of opaque pricing for AI-served storefronts. Two-thirds of Forbes AI 50 companies use usage-based pricing models, which require real-time consumption data that current AI agents cannot access from static web pages. The finding reinforces the AEO case for structured, machine-readable pricing pages: agents completing purchases without verified pricing data expose buyers to unintended charges. This is the first published throughput figure for AI agent price hallucination at the product-page level and connects directly to the Universal Cart checkout flow launched the same week (2026-w42-aeo-google-universal-cart).

  1. Stripe Blog
Issue 39
launch

Google Cloud introduces the Gemini agent for work

Google Cloud's Gemini agent automates enterprise workflows by connecting to business systems from a single prompt.

The Google Cloud Gemini agent selects the best model for each task, maintains cost controls, and logs all actions in an audit trail. Announced at the Gemini at Work 2026 event on October 8, the agent operates across Google Workspace and third-party tools including Slack and Microsoft 365, returning completed work to documents, inboxes, and developer environments. Agent spending limits pause the agent when exceeded, pending manual resumption. The governance features — audit trails, spending caps, and administrator controls — replicate the oversight primitives that agentic payment systems require before agents can act autonomously on purchasing tasks. For agentic commerce, the Gemini agent's architecture illustrates how enterprise AI agents will receive purchase authority: scoped to explicit limits and logged at the action level. Google disclosed that the agent supports both Gemini and Claude models. It is in private preview.

  1. Google Blog
Issue 41
launch

Introducing Claude Haiku 5.5

Anthropic launched Claude Haiku 5.5 at 75% lower cost than Haiku 4.5, adding computer use and browser use SDK support.

Claude Haiku 5.5 lowers the cost floor for high-volume agentic tasks by 75%. On OSWorld 2.1 computer use, it scored 72.4% — compared to GPT-6 Luna at 48.9% and Haiku 4.5 at 15.7%. Asana reported a 30% latency reduction and up to 2.5x faster inference per agent turn. Anthropic simultaneously added computer use and browser use support to the Claude Python and TypeScript SDKs in beta; Haiku 5.5 runs both tasks. For agentic commerce, the browser use SDK directly enables AI shopping agents to navigate merchant pages without custom integration. Anthropic also halved Sonnet 5.5 cache read prices, cutting costs on most agentic tasks by roughly 20%. The combination — cheaper subagent model plus cheaper cache reads on the reasoning model — reduces the per-transaction cost of multi-step agentic checkout to its lowest reported level.

  1. Anthropic News
Issue 41
launch

More ways to measure ChatGPT Ads

OpenAI added 20 measurement partners, OAIQ pixel, and Conversions API to ChatGPT Ads on October 5.

ChatGPT Ads' measurement infrastructure expanded on October 5 with 20 third-party partners covering attribution (AppsFlyer, Adjust, Kochava), incrementality testing (Haus, Measured, WorkMagic), data connectivity (Hightouch, Tealium, LiveRamp), and brand safety pilots with DoubleVerify and Integral Ad Science. Advanced conversion optimisation now learns from view-through as well as click-through conversions. For answer engine optimization, this matters structurally: it is the first time a large language model conversation surface has connected to a full third-party measurement ecosystem. Advertisers can now measure ChatGPT Ads on the same attribution infrastructure they use for Google and Meta. That comparability will accelerate spend allocation toward AI conversation surfaces, which are increasingly the discovery layer before agentic checkout.

  1. OpenAI Blog
Issue 41
launch

Gemini 4 Argon: our next era of frontier intelligence

Google launched Gemini 4 Argon with a 1-million-token output limit and 77.9% on DeepSWE v1.1, first released to cyber defenders via the Fairwind Program.

Gemini 4 Argon scores 77.9% on DeepSWE v1.1 at launch, ahead of Claude Opus 5.5 at 74.2% and GPT-6 Astra at 74.1%. Its 1 million output token limit — up from 64,000 — extends the range of multi-step agent workflows completable in a single inference call. The Fairwind Program (Google's restricted pre-release distribution channel for government authorities and critical infrastructure) appeared in this report previously for Gemini 3.8 Flash Cyber in 2026-w36. Paid API customers and Google AI Ultra subscribers follow the initial cyber defender cohort. Gemini 3.8 Live, which powers UCP-enabled checkout and Business Agent in Search Live, runs on a different model version; no announced timeline connects Argon to commerce surfaces. Introductory pricing is $2 per million input tokens and $10 per million output tokens.

  1. Google Blog
Issue 40
launch

Let skills in Gemini tackle your most repetitive tasks

Google launched Skills in Gemini, reusable instruction sets replacing Gems, rolling out globally in Gemini and Workspace through November 2026.

Skills let users save a named set of instructions and invoke them with a forward-slash command in the Gemini prompt bar, replacing Gems as Gemini's customization mechanism. Multiple skills stack: a purchasing-policy skill combined with a brand-guidelines skill governs both budget scope and tone in a single interaction. Workspace rollout begins October 5, completing by mid-November; Gemini app rollout begins October 13. Google added 14 Connected Apps to Gemini in 2026-w39, including Squarespace, Webflow, and SeatGeek. Skills define repeatable agent behaviors inside those connected contexts — a merchant can set a purchasing-scope skill that determines which categories and budgets Gemini proceeds on without additional confirmation. The deprecation of Gems removes the previous per-conversation customization model; Skills are persistent across sessions.

  1. Google Blog
Issue 40
launch

Introducing Gemini 3.8 Live with Live Avatar

Google made Gemini 3.8 Live with Live Avatar generally available, adding animated AI presenters to enterprise voice agents in 97 languages at $1 per million video tokens.

Gemini 3.8 Live introduced multi-step voice reasoning for commerce in 2026-w38, scoring 68.6% on τ-Voice (Sierra's benchmark for AI agent completion of multi-step tasks). Live Avatar adds a video output layer: animated personas that speak and lip-sync in the same rendering pass as audio generation. Enterprise customers access it through the Gemini Enterprise console or Gemini Live API, with US and EU endpoints. Pricing adds a third cost dimension — audio in, audio out, video out — at $1 per million video output tokens. Commerce use cases include product walkthroughs and customer service flows where a visual persona reduces the perception of automation. Gemini 3.8 Live shipped in Search Live in 2026-w38; this release moves the same model into the enterprise API tier with avatar output.

  1. Google Blog
Issue 39
launch

Boost your holiday sales with these agentic commerce updates

Google makes AI performance insights GA in 5 markets, opens Business Agent for YouTube ads beta, and expands Universal Commerce Protocol integration with cart transfer and enhanced checkout testing.

AI Performance Insights in Merchant Center is now generally available in Australia, Canada, India, New Zealand, and the U.S., giving retailers their first quantitative measure of brand share of voice across AI Mode and AI Overviews. Business Agent enters YouTube ads as a beta, extending the in-query conversational capability first launched in Search earlier this year. The Universal Commerce Protocol (UCP) integration hub — which already enables direct checkout for hundreds of thousands of merchants — gains cart transfer to merchant sites and enhanced checkout flow testing. During testing, lululemon found that conversational attributes submitted via feed were incorporated 50% of the time in AI Mode product recommendations, the first reported conversion rate for structured feed data to AI shopping surface mentions.

  1. Google Blog
Issue 38
launch

Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking

Gemini 3.8 Live Extended Thinking ranks #1 on Artificial Analysis' Speech to Speech Quality Index at 82.6 and leads agentic task completion on τ-Voice at 68.6%; models roll out across Gmail, Gemini Live, Search Live, and Google Workspace.

The τ-Voice (a benchmark measuring AI voice agent completion of multi-step tasks defined by Sierra, the customer service AI company) score of 68.6% and the τ-Voice-banking score of 35.1% are the highest published results for any voice model on live agentic task completion benchmarks at time of announcement. Gemini 3.8 Live rolls out to Search Live — the voice interface layer over AI Mode — without a waitlist, adding real-time voice to the commerce surface where UCP-powered checkout and Business Agent are already deployed (2026-w38-aeo-google-holiday-agentic-updates). Extended Thinking targets high-complexity, multi-step reasoning by voice, positioning the model for delegated shopping tasks where an agent parses constraints — budget, availability, loyalty status — before selecting and confirming a purchase.

  1. Google Blog
Issue 38
launch

Introducing agentic video understanding with Gemini

Google launched agentic video understanding for Gemini models on September 1, cutting token consumption by up to 88%.

Agentic video understanding replaces static frame-per-second ingestion with goal-directed scanning—the model fetches only segments relevant to the query, reducing analysis costs by up to 66% and improving accuracy by up to 7%. For agentic commerce workflows, this reduces the token cost of processing product demonstration videos, unboxing content, and shopping guides that AI shopping agents use to build product context before making purchase recommendations. The feature launches across Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite via the Gemini API in Google AI Studio and the Gemini Enterprise Agent Platform at standard API pricing with no additional feature fee. Google confirmed the feature will extend to the Gemini app for all users and will power YouTube's Ask YouTube feature on the video watch page, grounding answers in video visuals rather than metadata alone.

  1. Google Blog
Issue 37
launch

Expert Intelligence: a new way for you to engage with trusted content

Google launched Expert Intelligence, grounding Gemini Notebook responses in licensed Google Play Books content.

Expert Intelligence grounds Gemini Notebook responses in content from books a user owns on Google Play, rather than general training data. Google describes the feature as preserving the value of expert published work while adding guardrails on how licensed material is used. Responses remain within the specific text the user has purchased. The feature launches in Gemini Notebook and will extend to additional Google surfaces. For AEO practitioners, Expert Intelligence establishes a precedent for AI product recommendations grounded in verified reference sources — a different data provenance model than the catalog-feed approach used by Google's Universal Commerce Protocol (UCP) for product discovery. Both approaches address the same structural challenge: giving agents reliable, verifiable product or reference data rather than generalized model knowledge.

  1. Google Blog
Issue 36
research

Gisting: Compressing LLM Agent context to increase throughput and cut cost

Shopify Engineering published Gisting on August 19, demonstrating LLM context compression that cuts agent token costs.

Token costs constrain the frequency and depth of LLM agent interactions in production commerce systems. Gisting applies compression to the context window an agent carries across a multi-turn session — the running history of user requests, tool outputs, and prior responses that the model reads at each step. Shopify Engineering's paper demonstrates that compressing this context reduces per-call token consumption and raises throughput without accuracy loss on tested benchmarks. For agentic commerce applications executing multi-step workflows — product lookup, inventory check, cart management, and payment confirmation — context window efficiency directly affects per-transaction compute cost. Shopify operates over 150 million merchant storefronts; a compression technique from its engineering team carries implementation weight for the commerce platform at the centre of the agentic retail stack. The technique applies to any agentic checkout system maintaining session context across multiple tool calls.

  1. Shopify Engineering
Issue 35
launch

Now you can connect even more of your favorite apps and services to Gemini.

Google adds 14 new third-party app connections to Gemini, including Ticketmaster, Zocdoc, Angi, GetYourGuide, and OpenTable for UK users.

The 14 new connections extend Gemini's commerce and service reach into four categories: productivity (Granola, Otter.ai, Wix), entertainment and local services (Fever, GetYourGuide, Localiza, OpenTable UK, Ticketmaster), music (iHeartRadio, Pandora), and home and health services (Angi, Thumbtack, Zocdoc). Gemini Spark users can instruct the agent to take autonomous actions through these connections. The additions follow Google's Ask Maps food ordering expansion (2026-w32) and the AI Mode OTA booking partner confirmations (2026-w26), extending agentic reach from food and travel into ticketing, healthcare appointments, and home services. Custom Model Context Protocol (MCP) servers are supported for Spark users. The combined effect of the 1 billion user milestone (2026-w33-aeo-gemini-app-one-billion-users) and this connection expansion places agentic bookings, ticket purchases, and service appointments inside a single agent interface available to the largest AI app user base currently disclosed.

  1. Google Blog
Issue 34
research

More than 1 billion people are using the Gemini app every month.

Google confirms 1 billion monthly active Gemini app users; the app executes automated tasks across 40-plus connected apps.

Google CEO Sundar Pichai announced the milestone on August 11, describing the Gemini app as the company's fastest-growing product in its 28-year history. The 1 billion figure puts Gemini's monthly active user count alongside ChatGPT, which crossed the same threshold in June 2026. Sixty-three percent of Gemini users interact via voice; one in five Gemini Live sessions involves camera or screen sharing. The app generates more than 150 million images per day. The scale matters for agentic commerce: the app now executes automated tasks across more than 40 connected apps, turning the user base into a live pool of potential agentic commerce participants. Gemini Spark, the 24/7 autonomous agent tier now running on Gemini 3.7 Flash (2026-w33), is available to Pro and Ultra subscribers in 160 countries, placing agentic shopping and service booking at mass consumer scale for the first time.

  1. Google Blog
Issue 34
launch

Google AI Mode expands store product availability checker to let agents call physical stores on users' behalf

Google extends its AI-powered store-calling feature — using Gemini models and Duplex technology — to AI Mode, enabling agents to verify local in-store product availability and deals.

Google's in-store availability checker in AI Mode marks the extension of agentic commerce from e-commerce to omnichannel retail: the agent closes the loop on a product search by autonomously calling physical stores to confirm stock before the user commits to visiting. The underlying technology — Gemini models layered on Duplex's conversational telephony stack — represents a decade of investment in voice-mediated commerce converging at AI Mode alongside Google's agentic booking rollout (2026-w26-travel-google-ai-mode-ota-booking-partners). For brick-and-mortar retailers, the feature surfaces inventory to AI-mediated demand without requiring API integration with Google Shopping or the Universal Commerce Protocol (2026-w02-standards-google-ucp-launch): any store with a phone line becomes accessible to the agent. This creates agentic commerce reach for the long tail of retailers with no structured catalogue connectivity, fundamentally changing the value of local store presence in an agentic search landscape.

  1. Google Blog
Issue 30
launch

Clustering billions of products for agentic commerce with Catalog API (2026) - Shopify

Shopify Catalog API unifies billions of products across millions of merchants under Universal Product Identifiers for AI agent search.

The Catalog API is part of Shopify’s Spring ’26 Edition (June 17), which shipped 150-plus updates. Shopify hosts billions of product listings across millions of merchants with no shared schema between them. Without a unified layer, AI shopping agents encounter incompatible data structures when comparing products across merchants. The Universal Product Identifier (UPI) groups the same real-world item under one ID regardless of how individual merchants structure their listings. Shopify’s current release covers intra-store clustering; cross-store clustering — allowing comparison across merchants — is described as the next phase. UCP integration, which Google opened to Shopify merchants via the Stripe partnership (2026-w21), provides the protocol layer through which agents access this catalog infrastructure. Shopify reports AI-powered search orders grew 15x year-over-year in 2025, making agent-readable product data a direct commercial priority.

  1. Shopify Engineering
Issue 28
research

IDC: agentic supply chain AI market to grow from under $2B to $53B by 2030 as autonomous procurement scales

IDC projects the agentic supply chain management software market will reach $53 billion by 2030, with autonomous agents accounting for approximately $10.6 billion — 20% of total software spend.

The IDC $53 billion projection contextualises why SAP, Coupa, and a16z-backed startups like Lio (2026-w10-procurement-lio-series-a-a16z) are all entering autonomous procurement simultaneously. The 2025 baseline of less than $2 billion places the market in its first doubling phase, with projected value concentrated in 2028–2030 when enterprise migration from assisted to fully autonomous workflows is expected to peak. The 20% autonomous-agent share — roughly $10.6 billion — represents software revenue attributable to agents making procurement decisions without human approval, a threshold that changes the risk calculus for enterprise compliance, audit, and financial controls. Regulatory frameworks for autonomous spending are largely absent in 2026; the scale of the IDC projection suggests legislative attention in 2027–2028 is likely. The market size also justifies the infrastructure investments: SAP's free Joule runtime (2026-w23-procurement-sap-joule-ariba-fieldglass-ga) is a land-grab strategy calibrated to this projected market.

  1. SupplyChainBrain
Issue 30
launch

New ways to find your favorite sources and original content in AI Search

Google brought Preferred Sources into AI Overviews and AI Mode on May 27, surfacing labeled links from preferred publishers in AI responses with a new Highly Cited badge.

Preferred Sources now injects publisher-level signals directly into AI Mode responses — the surface where agentic shopping queries increasingly land. Users who select preferred sources are twice as likely to click through to those outlets; 345,000 unique sources have been selected since launch. The Highly Cited badge identifies original reporting that other articles reference, giving AI Search a structural authority signal distinct from PageRank-style link count. For publishers and merchants optimizing for AEO (answer engine optimization), Google has added two new discovery controls: explicit user source preferences and editorial citation strength. The Universal Cart launch at I/O 2026 (2026-w21) established the transaction layer across Search, Maps, and Shopping; Preferred Sources now adds a discovery filter that governs which merchant and publisher content reaches the AI responses feeding into that transaction layer.

  1. Google Blog (AI + Search)
Issue 25
launch

Introducing the Universal Cart and more ways to help you shop

Google launched Universal Cart at I/O 2026, a cross-service shopping cart with AI price alerts and UCP-powered checkout.

Universal Cart is the first consumer surface built directly on the Universal Commerce Protocol (2026-w02-standards-google-ucp-launch), released at I/O 2026 alongside UCP's multi-item cart, Identity Linking, and BNPL feature additions (2026-w21-standards-google-ucp-gml). The cross-service cart unifies items collected across Search, Gemini, and YouTube into a single checkout, extending the AI Mode shopping expansion (2026-w07-aeo-google-ai-mode-shopping-expansion). The launch operates on the merchant-and-payments stack established by Ulta Beauty's pilot (2026-w17-pilots-ulta-beauty-google-gemini-checkout), Stripe's Gemini integration (2026-w18-aeo-stripe-google-gemini-checkout), and the UCP Tech Council expansion (2026-w17-standards-ucp-tech-council-expansion). The AI price alerts feature is the first agent-monitored pricing primitive shipped by a major search engine and parallels Amazon Rufus's auto-buy at target price (2025-w47-pilots-amazon-rufus-auto-buy).

  1. Google Blog
Issue 24
research

Shopify Delivers Again as Merchants Clear $100 Billion in Q1 GMV

Shopify Q1 2026: AI-driven traffic rose 8x YoY; orders from AI searches up 13x; GMV hit $100.7B on 35% growth.

The 8x AI-traffic and 13x AI-order figures are the first publicly disclosed merchant-side demand-curve numbers from a top-five global e-commerce platform, providing the empirical counterweight to McKinsey's $3-5 trillion sizing (2025-w42-aeo-mckinsey-agentic-commerce-study) and Walmart's 1.18% ChatGPT conversion disclosure (2026-w13-pilots-walmart-chatgpt-conversion-data). The data covers the same January-March 2026 window as Mastercard's Q1 Agent Pay milestone (2026-w18-payments-mastercard-agent-pay-q1-milestone). Shopify is also a launch merchant on the Stripe-OpenAI ACP (2025-w40-payments-stripe-openai-acp-instant-checkout), Microsoft Copilot Checkout (2026-w02-payments-microsoft-copilot-checkout), and the Universal Commerce Protocol (2026-w02-standards-google-ucp-launch), placing the platform across every major AI commerce surface. GMV growth of 35% year-over-year frames the underlying commerce demand against which the AI-traffic multiples sit. The disclosure is the first earnings quarter where AI-driven order volume on a major platform broke into double-digit-multiple territory.

  1. Shopify Newsroom
Issue 22
launch

Stripe Brings Merchant Checkout to Google's AI Apps

Stripe's Agentic Commerce Suite enables merchant checkout inside Google Gemini and AI Mode.

The integration places Stripe's Agentic Commerce Suite (2025-w50-payments-stripe-agentic-commerce-suite) inside Google's Gemini and AI Mode surfaces, extending the AI Mode shopping expansion (2026-w07-aeo-google-ai-mode-shopping-expansion) with a processor-grade checkout. Stripe now operates checkout across ChatGPT via the Agentic Commerce Protocol (2025-w40-payments-stripe-openai-acp-instant-checkout), Microsoft Copilot (2026-w02-payments-microsoft-copilot-checkout), and Google Gemini, covering the three largest US AI consumer surfaces in a single processor. The launch coincides with Stripe Link's agent APIs (2026-w18-payments-stripe-link-agents) and Stripe's seat on the Universal Commerce Protocol Tech Council (2026-w17-standards-ucp-tech-council-expansion). Ulta Beauty's same-month Gemini checkout (2026-w17-pilots-ulta-beauty-google-gemini-checkout) is the first live retailer flow using this integration. The integration uses Shared Payment Tokens introduced with ACP and operates inside the AI Mode Direct Offers and Business Agent surfaces that shipped in February (2026-w07-aeo-google-ai-mode-shopping-expansion).

  1. PYMNTS
Issue 21
research

Skift: 2% of US consumers vs 80% of executives plan agentic travel agents, exposing adoption gap

Skift analysis finds only 2% of US travellers willing to delegate full trip booking to autonomous AI, despite 80% of travel executives planning at-scale deployment.

The Skift 80/2 mismatch — executives deploying, consumers not adopting — parallels early e-commerce adoption curves where supply built ahead of consumer habit. The primary friction Skift identifies is accountability: no clear liability framework exists when an AI agent books a non-refundable fare or misses a preference. Regulatory bodies including the UK FCA (2026-w13-regulation-uk-fca-agentic-payments-priorities) have begun addressing agentic payment accountability but not travel-specific consumer redress frameworks. The data sets a measurable baseline: agentic travel success should be tracked against the 2% figure, not against executive enthusiasm. The Mindtrip Flights launch (2026-w19-travel-mindtrip-flights-launch) will be an early real-world test of whether a frictionless zero-redirect booking experience can shift consumer adoption beyond that threshold. The gap also signals an opportunity for regulatory clarity to unlock latent demand.

  1. Skift
Issue 30
research

eMarketer: 32.6% of US food shoppers willing to let AI auto-reorder staple groceries

Joint eMarketer/Amazon Ads survey finds a third of US food shoppers accept AI-managed replenishment for staples, with 45.8% open to AI recipe and cart recommendations.

The 32.6% auto-reorder acceptance rate — across all US food shoppers, not just early-adopter segments — establishes a credible consumer mandate for agentic grocery at scale. The figure is directionally consistent with the deployment timelines of Kroger (2026-w02-grocery-kroger-google-gemini-shopping-agent), Amazon Alexa+ (2026-w06-grocery-amazon-alexa-plus-prime-launch), and Carrefour (2026-w13-grocery-carrefour-chatgpt-integration): all three launched autonomous reorder features in Q1 2026. The larger 45.8% receptivity to AI recipe-and-cart tools suggests that the inspiration-to-purchase flow — not just replenishment — is the primary agentic grocery opportunity. Retailers relying purely on auto-replenishment agents may underserve a market that expects discovery and planning assistance, not just a smarter subscription service. The data also provides a benchmark for measuring consumer adoption: any deployment claiming mainstream traction should be evaluated against this one-third baseline.

  1. FMI
Issue 30
launch

OpenAI shuts down Instant Checkout merchant-of-record model, pivots to brand-owned checkout

OpenAI retires its centralised merchant-of-record approach after poor merchant onboarding; pivots to brand-native apps inside ChatGPT with Shopify, Etsy, and Walmart.

The retirement closes the centralised merchant-of-record experiment launched with Stripe in 2025 (2025-w40-payments-stripe-openai-acp-instant-checkout), citing low merchant onboarding. The brand-native pivot — apps inside ChatGPT — aligns with Walmart's parallel disclosure of one-third conversion versus its own website (2026-w13-pilots-walmart-chatgpt-conversion-data) and Walmart's pivot to embedding the Sparky agent inside ChatGPT. The Shopify and Etsy continuity preserves the original merchant relationships in a different structural form. The shift parallels the broader industry move toward merchant-owned agentic checkout flows visible in Stripe's Suite (2025-w50-payments-stripe-agentic-commerce-suite), PayPal's Agent Ready (2025-w44-aeo-paypal-agentic-commerce-services), and Google's UCP (2026-w02-standards-google-ucp-launch), which all keep the merchant as the transacting party of record. The pivot closes the merchant-of-record experimental period for OpenAI.

  1. CNBC
Issue 18
research

IDC predicts agentic AI will redefine travel and hospitality across the full booking pipeline in 2026

IDC research frames 2026 as the pivotal year for agentic AI in travel and hospitality, with full pipeline deployment from discovery through post-trip loyalty management.

IDC's hospitality forecast establishes a demand-side analyst baseline before the major supply-side launches of Q2 2026. IDC identifies the end-to-end agentic pipeline as the 2026 differentiator: in 2025, AI assisted individual steps (search, recommendation, customer service); in 2026, agents are expected to orchestrate discovery, comparison, booking, payment, itinerary management, and post-trip loyalty redemption in an integrated flow without human handoffs. The hospitality sector's data richness — loyalty profiles, past stays, preference histories, seasonal pricing — provides agentic systems with unusually deep personalisation inputs relative to first-time retail purchases. IDC identifies this as the primary competitive moat for branded hotel chains over OTA intermediaries: a Marriott agent with five years of guest stay data will outperform a generic booking agent, creating an incentive for hotel brands to own agentic relationships rather than cede them to Google or Booking.com.

  1. IDC
Issue 30
launch

Google activates AI Mode shopping with Direct Offers, Business Agent, and Gemini checkout

Google enables Direct Offers, Business Agent chat, and sponsored formats inside AI Mode Search; Etsy and Wayfair items become purchasable inside Gemini without leaving Google.

Direct Offers, Business Agent, and sponsored formats add the discovery, conversation, and monetisation layers to Google's AI Mode, extending the protocol-level Universal Commerce Protocol launch (2026-w02-standards-google-ucp-launch) four weeks earlier into a consumer surface. The Etsy and Wayfair purchase paths mirror Etsy's earlier launch in ChatGPT Instant Checkout (2025-w40-payments-stripe-openai-acp-instant-checkout), placing the same merchant inside two competing AI shopping ecosystems. Stripe's subsequent Gemini and AI Mode integration (2026-w18-aeo-stripe-google-gemini-checkout) and Ulta Beauty's UCP-powered Gemini checkout (2026-w17-pilots-ulta-beauty-google-gemini-checkout) build on this consumer foundation. The Universal Cart launch at I/O 2026 (2026-w21-aeo-google-universal-cart) extends AI Mode into a cross-service cart layer. The launch is Google's first cross-format consumer surface for answer engine optimization (AEO) and AI Mode shopping.

  1. Bloomberg
Issue 13
launch

OpenAI launches ChatGPT Shopping Research

Product discovery feature powered by a reinforcement-trained GPT-5 mini variant achieves 52% accuracy on multi-constraint queries; available free to all logged-in users.

ChatGPT Shopping Research is OpenAI's first dedicated product-discovery feature outside Instant Checkout (2025-w40-payments-stripe-openai-acp-instant-checkout). The reinforcement-trained GPT-5 mini variant and the published 52% multi-constraint accuracy figure place a concrete benchmark on agent product reasoning, contrasting with the conversion-rate data Walmart later discloses (2026-w13-pilots-walmart-chatgpt-conversion-data). The free-tier rollout matches Perplexity's free shopping agent positioning (2025-w47-aeo-perplexity-paypal-instant-buy) and contrasts with Amazon's account-gated Rufus (2025-w47-pilots-amazon-rufus-auto-buy). Shopping Research's discovery layer later becomes part of OpenAI's pivot to brand-native apps inside ChatGPT (2026-w13-aeo-openai-instant-checkout-shutdown), with discovery decoupled from the centralised checkout model retired the same week. Discovery survived; the merchant-of-record checkout did not. Shopping Research and Instant Checkout were the two consumer touchpoints of the Stripe-OpenAI partnership; the discovery layer persisted while the payment layer pivoted to brand-owned flows.

  1. OpenAI
Issue 8
launch

Perplexity launches free shopping agent with PayPal Instant Buy

Perplexity releases US shopping agent with real-time merchant catalog browsing and PayPal Instant Buy checkout inside search results.

Perplexity's shopping agent extends real-time merchant-catalog browsing into an AI search interface, joining ChatGPT's Instant Checkout (2025-w40-payments-stripe-openai-acp-instant-checkout) and Amazon's Rufus (2025-w47-pilots-amazon-rufus-auto-buy) on the same week. PayPal Instant Buy provides the agentic-payments checkout primitive, drawing on the Agent Ready surface coverage PayPal announced five days earlier (2025-w44-aeo-paypal-agentic-commerce-services). The free-tier positioning contrasts with subscription-gated agent surfaces and broadens the AEO lane's addressable user base. Perplexity's later Comet browser becomes the subject of Amazon's preliminary-injunction case (2026-w11-regulation-amazon-perplexity-court-order), establishing the company as both a payments partner and a regulatory test case for third-party agent access to merchant accounts. The third week of November 2025 forms the densest cluster of consumer-surface launches in the AEO lane to date.

  1. CNBC
Issue 7
launch

PayPal launches Agentic Commerce Services

Agent Ready and Store Sync tools let merchants accept payments on AI surfaces and make product data discoverable across channels including Perplexity; one integration covers multiple AI shopping surfaces.

Agent Ready and Store Sync sit on the merchant side of the agentic stack, complementing the consumer-side checkout work in Stripe-OpenAI's ACP (2025-w40-payments-stripe-openai-acp-instant-checkout) and Mastercard Agent Pay (2025-w18-payments-mastercard-agent-pay). The Perplexity surface coverage links to PayPal Instant Buy's later debut in Perplexity's free shopping agent (2025-w47-aeo-perplexity-paypal-instant-buy). The one-integration-many-surfaces model is later echoed by Visa's Intelligent Commerce Connect (2026-w15-payments-visa-intelligent-commerce-connect) and Stellagent's Commerce Studio (2026-w20-pilots-stellagent-commerce-studio). The release pairs with the same-week Mastercard wallet integration (2025-w44-payments-mastercard-paypal-integration), giving PayPal both the credential and the discoverability layer of agentic commerce within a single 48-hour window. PayPal Instant Buy later anchors Microsoft Copilot Checkout (2026-w02-payments-microsoft-copilot-checkout).

  1. PayPal Newsroom
Issue 6
research

McKinsey estimates agentic commerce at $3–5 trillion globally by 2030

Study forecasts up to $1 trillion in US B2C retail and $3–5 trillion globally redirected through agentic commerce by 2030, comparing the shift to the mobile era.

The McKinsey estimate is the first multi-trillion-dollar sizing of agentic commerce from a top-tier consultancy, anchoring later GMV-disclosure events: Shopify's Q1 2026 report of 8x AI-driven traffic and 13x AI-order growth (2026-w19-aeo-shopify-q1-2026-ai-traffic) and Mastercard's Q1 2026 disclosure that Agent Pay reaches nearly all cards globally (2026-w18-payments-mastercard-agent-pay-q1-milestone). The mobile-era comparison is structural rather than predictive — McKinsey maps the funnel-redirection mechanism, not the timeline. The report frames the addressable market that the answer engine optimization (AEO) lane's PayPal, Perplexity, and OpenAI launches (2025-w44-aeo-paypal-agentic-commerce-services, 2025-w47-aeo-perplexity-paypal-instant-buy, 2025-w48-aeo-openai-chatgpt-shopping-research) compete to capture. Publication falls in the same October week as Visa's Trusted Agent Protocol launch (2025-w42-payments-visa-trusted-agent-protocol), bracketing the week with both market sizing and network-level infrastructure.

  1. McKinsey & Company
Issue 5

What is AEO (Answer Engine Optimization)?

AEO marketing — short for answer-engine optimisation, sometimes called agentic engine optimisation — is the practice of structuring a merchant's product catalogue, content, and metadata so an AI shopping agent can find, evaluate, and recommend the merchant's offering. SEO targets human search-result clicks. AEO targets retrieval and ranking inside an agent's reasoning loop. The active surface includes structured product feeds, machine-readable price and availability, schema markup, and direct integrations with the major AI surfaces (ChatGPT Shopping, Google AI Mode, Perplexity, Gemini). A merchant that ranks high in AEO is one an agent will surface to its user even when no human typed a search query.

AI-surface product discovery has crystallised across a handful of platforms in late 2025 and early 2026. OpenAI launched ChatGPT Shopping Research in November 2025 using a reinforcement-trained GPT-5 mini variant that combs the web for product-by-product comparisons. Google activated direct shopping inside AI Mode in February 2026 with Direct Offers, Business Agent chat, and sponsored formats. Stripe disclosed a Google partnership in April 2026 enabling merchant checkout inside Gemini and AI Mode for Wix, BigCommerce, and WooCommerce stores via the Agentic Commerce Suite. Perplexity launched a free shopping agent backed by PayPal Instant Buy in November 2025. OpenAI retired its merchant-of-record model in March 2026, shifting to brand-owned checkout flows that hand the consumer back to the merchant after agent-led discovery. McKinsey estimated agentic commerce at $3–5 trillion globally by 2030. A unified feed standard for agent-readable product data remains unresolved at Schema.org. This hub tracks the platform launches, ranking signals, and feed-spec changes that shape which merchants agents surface. AEO marketing as a discipline now spans feed-spec engineering, schema markup, and direct platform integrations with the major AI surfaces.