AI is becoming the new front door to commerce. As discovery moves to LLMs and agents, the deterministic funnel breaks and the branded site loses its monopoly on discovery — the contest becomes intelligibility and trust. Build for humans and agents to avoid the “Agentic Ghettos.”
AI becomes a concierge that aggregates, curates, and validates purchases before handing high-intent shoppers to merchants. So the job of your site splits in two: a machine-readable source of truth for agents, and a destination worth choosing for humans. Serve both, or get sorted into an Agentic Ghetto.
When the buyer is “h,” your product data is the storefront. Structured feeds and clean markup decide whether an agent can recommend you at all.
Agents own speed and convenience. What they can’t replicate is brand, story, taste, and trust — the reasons a person picks you over the cheapest match.
The essays, research, and episodes that define Future Commerce’s position on AI and agentic commerce.
Phillip Jackson’s foundational essay: agents as a new species of shopper, the third great platform shift, and the walled-off zones every brand should design against.
Read the essay →
A 2,000-consumer study and the name FC gave it: AI aggregates, curates, and validates before handing shoppers to merchants.
Read →
FC’s Holiday AI Shopping Benchmark — the hard data that AI is now the habitual front door to commerce.
Read →
Sharon Gee on the three truths of agentic commerce: data is the storefront, agents are customers too, and sites must be built for eyeless robots.
Listen →
Mike Mallazzo on AI’s eras in commerce and how generative AI decentralizes product search.
Read →
What the branded site becomes when AI mediates discovery — plus the product-feed and schema moves that matter.
Read →Sharon Gee on the three truths of agentic commerce — data is the storefront, agents are customers too, and what trust looks like when an agent checks out for you.
Six questions we and their agents buy.
How shoppers actually use AI to research, compare, and buy — from our own consumer surveys.
ACP, UCP, Instant Checkout, and the new plumbing that lets agents pay.
Feeds, schema, and clean data — so an agent can find you and recommend you.
Open versus walled platforms — and who decides which agents get in.
Non-determinism and the death of the deterministic performance-marketing funnel.
Google, Shopify, OpenAI, Perplexity, Amazon and Walmart fighting to be the front door.
A deeper cut across essays, member briefs, and podcasts in this frontier.
Ghe ‘Commerce 4’ foundations of agentic commerce feeds, first-party data, AI ads, and UCP.
Read →The SensesOpenAI × Stripe Instant Checkout and the Agentic Commerce Protocol — conversations as the new storefront.
Read →EssayWhy probabilistic, LLM-mediated discovery breaks the deterministic performance funnel.
Read →EssayFrom Shoptalk Spring 2026 — where the agentic hype ran hottest, and the human, trust-building uses that actually work.
Read →PodcastThe live conversation behind the thesis — how to keep agents and humans shopping side by side.
Listen →The SensesWhat an answer-engine shopping flow reveals about discovery, citation, and merchant control.
Read →The vocabulary we coined for the agentic commerce shift.
Walled-off zones where platforms ration agent access and keep agent shoppers separate from humans — the failure mode to design against.
AI aggregates, curates, and validates purchases before handing high-intent shoppers off to merchants.
People buy across modes and moments instead of moving down a tidy funnel — the behavior agents speed up.
Shopper, brand, and agent — trust and data rights renegotiated now that the agent is a customer too.
Probabilistic, AI-mediated discovery that breaks the predictable performance-marketing funnel.
When the buyer is “eyeless,” your machine-readable product truth — not your homepage — is what sells.
Commerce mediated by AI agents that research, compare, and increasingly buy on a shopper’s behalf. We see it not as “a bot buys everything” but as a change in where the buying decision gets made — with AI as the new front door.
No — but the site’s job splits in two: a machine-readable source of truth an agent can use and recommend, and a destination that gives humans a reason to choose you over the cheapest match.
Phillip Jackson’s term for walled-off zones where platforms ration agent access and keep agent shoppers separate from humans. The goal is to keep agents and humans shopping side by side, in the open.
In our research, about 70% use AI as a regular shopping habit, 77% click through to a merchant before buying, and 23% will buy on an AI recommendation — rising to 44% when the AI finds the best price.
Start with structured product feeds, clean schema and markup, and crawlable content, so an agent has unambiguous product truth to work from. “The eCommerce Site’s New Existential Crisis” has the checklist.
AI doesn’t live in a silo — it reshapes the channels, marketplaces, and brands around it.
What happens to sponsored placements and RMN economics when the shopper is an agent, not a human.
Explore →FrontierAgentic surfaces — ChatGPT, Perplexity, Rufus — as a net-new channel with their own listing discipline.
Explore →FrontierWhy meaning, lore, and a differentiated promise are the moat when agents commoditize the cheapest match.
Explore →Future Commerce publishes the research, essays, and podcasts on how AI is changing the culture of commerce. Free, every week.
Subscribe free →