Consumer agentic commerce has had a bumpy 2026 — ChatGPT's Instant Checkout launched with fanfare in September 2025 and was quietly shut down by OpenAI in March 2026. It would be a mistake to read that as agentic commerce stalling. The B2B side of this story is moving in the opposite direction, and the timeline is shorter than most vendors expect.
In consumer retail, agentic commerce usually means an AI assistant researching, comparing, and sometimes completing a purchase for an individual shopper. In B2B, the more consequential version is procurement and sales-research agents — software embedded inside CRMs, procurement platforms, and enterprise search tools — that evaluate vendors, compare pricing tiers, check integration compatibility, and summarize findings for a human buyer or negotiate terms directly.
Forrester's own forecast puts a sharper point on the timeline: 20% of B2B sellers are expected to face agent-led quote negotiations by the end of 2026 — meaning the counterparty on the other side of a pricing conversation may increasingly be software, not a person, evaluating your terms against a defined set of criteria.
It's worth separating the headline from the substance here. ChatGPT reached roughly 900 million weekly active users by February 2026, with over 800 million of those having shopping enabled and an estimated 50 million shopping-related queries occurring daily inside the app — about 2% of total ChatGPT query volume. Instant Checkout, launched in September 2025 with a 4% merchant transaction fee, was the consumer-facing execution layer for some of that volume.
OpenAI's own explanation for the March 2026 shutdown pointed at flexibility, not lack of demand: "We've found that the initial version of Instant Checkout did not offer the level of flexibility that we aspire to provide," the company said in its announcement, "so we're allowing merchants to use their own checkout experiences while we focus our efforts on product discovery." Read plainly, that's a retreat from owning the checkout step specifically — not from agentic shopping as a category.
Meanwhile, retail-side data shows the underlying behavior accelerating regardless of any single company's checkout product. AI-driven traffic to U.S. retail sites grew 393% year-over-year in Q1 2026, and Shopify reported that AI-driven traffic to its merchant stores grew 8x year-over-year in the same quarter, with orders originating from AI-powered searches up nearly 13x.
| Signal | Data point | What it means for B2B |
|---|---|---|
| Consumer AI shopping scale | 900M weekly ChatGPT users, 50M shopping queries/day | Establishes the behavioral pattern now migrating to B2B research tools |
| Retail AI traffic growth | +393% YoY (Q1 2026) | Agentic discovery is scaling fast even without a single dominant checkout product |
| B2B agentic deployment plans | 74% within 2 years | Most B2B vendors' buyers will soon be evaluated-by-agent, not just researched-by-human |
| Agent-led negotiation forecast | 20% of B2B sellers by end of 2026 | Pricing pages and terms need to be agent-legible, not just human-persuasive |
McKinsey forecasts $900 billion to $1 trillion in US retail revenue moving through agentic commerce by 2030, and $3 to $5 trillion globally. Morgan Stanley's AlphaWise survey work shows LLM adoption already approaching 50% in the US, with AI agents capturing an estimated 10-20% of e-commerce activity today.
One reason B2B vendors should act on fundamentals now rather than waiting for a "winning" platform: the technical standards for agentic commerce haven't consolidated. At least five protocol families are actively competing — ACP, UCP, AP2, Visa's Trusted Agent, and Mastercard's Agent Pay — and Google entered the field in January 2026 with its own protocol, backed by Walmart, Target, Shopify, and more than 20 other partners.
That fragmentation is actually useful information: it means no single integration decision today will be the "right" one for every future agent. The more durable investment is making your product data clean, structured, and consistently accurate everywhere — which pays off regardless of which protocol layer eventually wins.
With five-plus competing protocol families and no clear winner, building deep custom integrations against one specific agentic commerce protocol today carries real re-work risk. Prioritize protocol-agnostic fundamentals — clean structured data, current pricing, honest comparisons — over betting on ACP, UCP, AP2, or any single vendor's approach before the market consolidates.
Agentic commerce and GEO are converging on the same underlying requirement: content and data structured clearly enough that a machine, not just a human, can evaluate it accurately. The B2B companies that treat their pricing pages, integration docs, and comparison content as agent-readable infrastructure — not just marketing copy — will be positioned for both today's AI search citations and tomorrow's agent-led procurement conversations.
Agentic commerce refers to AI agents completing purchasing tasks on behalf of a human — researching options, comparing vendors, negotiating terms, and in some cases executing checkout — rather than a person browsing and buying directly. In B2B, this increasingly means procurement software and sales-research agents evaluating and selecting vendors with less direct human review of each option.
Industry forecasts project that 90% of B2B buying will be AI-agent-intermediated by 2028, driving more than $15 trillion of B2B spend through AI agent exchanges. Separately, 74% of B2B companies report plans to deploy agentic AI within the next two years, and Forrester predicts 20% of B2B sellers will face agent-led quote negotiations by the end of 2026.
ChatGPT's Instant Checkout launched in September 2025 and served ChatGPT's roughly 900 million weekly active users, with OpenAI charging merchants a 4% transaction fee per completed purchase. OpenAI shut the initiative down in March 2026 due to underwhelming performance, shifting focus toward improving product discovery and ChatGPT's broader apps ecosystem rather than in-chat checkout specifically.
Structure pricing, specifications, integration compatibility, and comparison data so an agent can parse it without human interpretation: use explicit schema markup, keep pricing pages current and machine-readable, publish clear integration and compatibility matrices, and ensure comparison-relevant details are present in actual page content rather than buried in PDFs or gated behind forms.
We'll audit how your pricing, integration docs, and comparison content read to an AI agent — and fix the gaps before your competitors do.
Book a Free AI Visibility CheckFree — No Commitment
We test 15 of your buyers' top queries across ChatGPT, Perplexity, and Google AI Overviews — and show you exactly where you appear, where competitors are being cited, and what's driving the gap.
Check My AI VisibilityTakes 2 minutes to request. We'll follow up within one business day.