Sponsored by
Happy Wednesday! Adyen launched Agentic, a payments and checkout layer for enterprise merchants that want AI assistants to sell without handing away the boring parts of commerce.
Microsoft is turning Copilot work into metered labor, X is trying to make ad signals credible again, and marketplace advantage is moving from searchable supply to executable demand.
In today's MarketMaze:
🛒 Adyen sells agentic rails
🤖 Microsoft meters work
🧩 X fixes the signals
🤖 Marketplaces become agents
🛒 Two retailers harden power
🧮 AI ROI meets finance
LET'S ENTER THE MAZE!
The Maze: Adyen launched Agentic, a U.S.-limited enterprise API suite that helps merchants sell through AI shopping agents without rebuilding checkout for every platform.
Adyen splits the journey into Feed, Cart, and Payments: product data, tax/shipping/order logic, and fraud-safe authorization.
The point is control: merchants need AI discovery to connect with live inventory, payment rules, and merchant-of-record logic, not just a prettier chatbot.
Early participants include American Express, Mastercard, Salesforce, Visa, ESW, Scheels, Sézane, and SharkNinja, while Adyen says the product supports modular adoption.
Why it matters: Agentic commerce will not scale on demos alone. The boring layer, data, tax, fraud, routing, and payment authorization, may decide who captures the margin.
10,000+ brands found their 3PL with Fulfill.
From Liquid Death to Longchamp, thousands of brands trusted Fulfill to find their fulfillment partner. Tell us your products, order volume, and shipping zones, and our team screens a network of 2,800+ vetted 3PLs to hand you a ranked shortlist of your best-fit partners. You talk to the matches you like, get pricing directly from them, and decide on your own terms. No fee, no pressure, and most brands get their shortlist within days. Find your best-fit 3PL.
The Maze: Microsoft made Copilot Cowork generally available worldwide, turning enterprise AI from a seat-license assistant into metered workflow labor.
Cowork runs long-running, multi-tool tasks inside Microsoft 365, returning completed work for review rather than just drafts or suggestions.
The pricing model uses Copilot Credits, with task cost based on model use, context retrieval, tool calls, and runtime; PayGo is priced at $0.01 per credit.
Admins can set access, spending caps, usage alerts, and reporting through Microsoft 365 cost controls, making AI work easier to govern before teams let it loose.
Why it matters: Microsoft is trying to own the place where AI work is requested, measured, approved, and paid for. That is enterprise software gravity.
No theory. No slides. Just pipeline.
Most founders know their product. Few know how to get it in front of the right people. In this hands-on session, Clay + HubSpot for Startups walk you through ICP definition, prospect list enrichment, and AI-personalized outreach. You launch your first sequence before the session ends. June 18. 11am ET / 4pm GMT.
The Maze: X is adding the dull but essential layer its ad comeback needs: cleaner setup, cleaner event signals, and faster diagnostics inside Ads Manager.
X added Google Tag Manager setup, unified Conversion API resources, and live Pixel/CAPI diagnostics, with rollout during June 2026.
The GTM workflow lowers the developer burden for server-side tracking, while the CAPI hub gives technical teams one place to manage documentation, configuration, and controls.
The move builds on X's April Ads Manager rebuild and narrows a measurement gap with platforms that already made CAPI setup easier.
Why it matters: Advertisers do not scale spend on attention alone. They scale when conversion signals are clean enough to defend budgets and optimize campaigns.
The Maze: Marketplace advantage keeps moving closer to the customer job: from local trust, to searchable supply, to vertical apps, to retention, and now to AI agents that may execute the transaction.
The five-era model moves from bazaars and classifieds to Amazon, eBay, Uber, Etsy, and AI assistants, showing how marketplaces evolved from access layers into task-specific operating systems.
The 2020-2022 reset matters because COVID pulled demand online while Apple’s IDFA rules weakened cheap tracking, forcing marketplaces to care more about retention and efficiency.
The AI era is not just better search; OpenAI’s Agentic Commerce Protocol and Google’s UCP point toward agents that discover, buy, and support.
Why it matters: If agents become the buyer interface, marketplaces will compete less on “how many options do we show?” and more on “can we finish the job?” Clean catalog data, checkout, inventory, logistics, and support become the new shelf space.
The Maze: Amazon and Walmart are forecast to top 51.0% of US retail ecommerce sales in 2026, up from 41.3% in 2019. That is not just share gain. It is distribution power hardening into infrastructure.
Amazon still does most of the heavy lifting at 39.5%, but Walmart's 11.5% is enough to push the pair over the half-market line in the forecast.
The metric includes direct and marketplace sales, so the pressure lands on brands twice: they need platform demand, then they need visibility inside the platform.
The same value-led concentration is showing up in stores, where Walmart, Target, Costco, and Dollar General lifted their visit share from 16.8% in 2019 to 17.5% in Q1 2026, per Placer.ai.
Why it matters: Smaller retailers are not dead. But vague channel strategy is. When two retailers control half of online sales, brands need to know where they win discovery, where they buy visibility, and where they still give shoppers a reason to leave the default path.
The Maze: AI has reached the boardroom. Now it has to survive the finance meeting.
In Oliver Wyman Forum and NYSE's CEO data, AI-driven revenue gains above 10% fell from 9% to 5%, while the zero-revenue-impact group rose from 15% to 23%.
Cost savings look less ugly: the 1%-10% savings band stayed near 40%, which suggests AI is easier to prove in efficiency work than in new revenue.
Why it matters: The AI budget is moving from demo logic to capex logic. Commerce leaders need baselines for conversion, AOV, support cost, media efficiency, returns, and margin before AI gets asked the obvious question: what changed?
Everything else you should know about
🛒 Kroger and Wegmans test agentic grocery carts. Cooklist is powering automated cart building for grocers, turning recipe intent and pantry data into checkout-ready baskets. Grocery AI is getting operational.
🇬🇧 UK publishers invoice AI scrapers. Publishers are billing AI crawlers GBP500 per article through county-court claims. The cheap-content era is meeting small-claims enforcement.
📈 AI-referred shoppers spend more. Retail sites saw 53% more revenue per visit from AI-referred shoppers. The traffic is small, but the basket signal is getting harder to ignore.
🇨🇳 China adds banks to digital yuan rails. China recruited 26 banks for its cross-border digital-yuan project. Payments infrastructure is becoming geopolitics with APIs.
📊 NIQ and Unlimitail join retail-media measurement. NIQ and Unlimitail are combining shopper data and omnichannel measurement across retailers and countries. Retail media keeps professionalizing because advertisers eventually ask for proof.
🧠 Databricks pushes agentic customer data. Databricks launched CustomerLake with IAS signals, linking first-party profiles to media buying. The CDP pitch is becoming less database, more decision engine.
🚢 Customs risk moves up the shipping stack. Shippers are rechecking importer-of-record exposure, recordkeeping, and visibility after U.S. customs-rule changes. Compliance is now a logistics product feature.
THAT’S IT FOR TODAY!
You’re the reason our team spends hundreds of hours every week researching and writing this email.
What do you think of this issue?
See you next time in the maze!
MarketMaze team
