Agentic Commerce Reality Check: Why 89% Prep, Only 3% Deliver

Agentic commerce is the practice of using autonomous AI agents to research, compare, and complete purchases on behalf of consumers, often without a human clicking "buy." The infrastructure for this future just arrived in force. In the same October week, Stripe, TikTok, Mastercard, Constructor, and Gap all shipped new agentic commerce capabilities, from one-click checkout to AI identity scoring for payment networks.

Yet the number that should stop every Fortune 500 CMO mid-scroll is this: only 3% of merchant transactions involve AI agents, even though 89% say they are actively preparing for agentic commerce . That gap, 89% investment versus 3% delivery, is the defining business story of the AI commerce era, and it is exactly the kind of hype-versus-behavior disconnect that Matt Britton has spent his career decoding for brand leaders.

Britton, founder of consumer intelligence platform Suzy and author of Generation AI, built his reputation studying the space between what technology makes possible and what consumers actually do. As an AI keynote speaker, he tells Fortune 500 audiences the same thing he is telling readers now: the agentic commerce stack is production-ready, but trust and habit formation remain the real bottleneck to ROI. Enterprises racing to deploy AI shopping agents without understanding this behavioral lag are building infrastructure for a customer who has not shown up yet.

This matters because the dollars at stake are enormous. The US agentic commerce market is forecast to be worth $300-500 billion by 2030 and account for 15-25% of total eCommerce sales, according to Bain. Gartner's own forecasts anticipate 20% of digital commerce transactions will be executed through AI platforms, either via on-platform checkout or via AI agents by 2030 . The destination is not in question. What is in question is how companies survive the long, awkward middle period between launch and adoption, and Britton's keynote work exists precisely to help leadership teams plan for that gap rather than get blindsided by it.

What Is Agentic Commerce and Why Did Everyone Launch at Once?

Agentic commerce describes AI systems that move beyond recommendation into execution: finding a product, comparing options, and completing a transaction with minimal human intervention. The infrastructure race accelerated dramatically this month. Agentic commerce had its busiest launch month on record, with Stripe, TikTok, Mastercard, Constructor and Gap all moving at once, yet only 3% of merchant transactions involve agents and liability remains unresolved.

Each player attacked a different layer of the stack. Constructor announced a checkout integration so shoppers on retail sites and apps can discover products, ask questions, get personalized recommendations, and refine their choices, then securely check out and pay, all without leaving the agent experience . The company claims this is a first: Constructor is the first to offer Stripe-powered checkout directly within an onsite shopping agent .

TikTok moved on a parallel track, debuting conversational shopping and one-tap purchasing. TikTok launched a Shopping Assistant that understands conversational context, remembers shopper preferences, answers questions on products, sizing, shipping, and availability, and helps users complete purchases, alongside a separate one-click checkout feature that allows purchases from brands directly from the For You feed, with Shopify, Salesforce, Shoplazza, Stripe, and other commerce providers supporting the initiative. Mastercard, meanwhile, tackled the question nobody else was answering: how do you even know a transaction came from an agent? Mastercard expanded Agent Pay with services combining identity, intent, behavioral, and fraud signals, with the first service being tested in the U.S. producing a probability score estimating whether a transaction was initiated by an AI agent.

Gap rounded out the week by deploying conversational AI across its portfolio brands, pushing agentic tools into mainstream apparel retail rather than leaving the category to pure-play tech. This is infrastructure convergence at a scale rarely seen in commerce technology, five major players, five different layers of the stack, one single week. It is the clearest signal yet that enterprises believe the agentic era is arriving, which makes the 3% transaction number even more striking.

The Enterprise AI Adoption Gap: Why Readiness Does Not Equal Revenue

The enterprise AI adoption gap is not a technology problem. It is a behavioral one, and that distinction changes how every CMO should be budgeting for 2027. Only 17% of organizations have deployed AI agents today, although more than 60% plan to do so within two years , according to Gartner research, confirming that ambition is sprinting far ahead of production reality.

Where agents are live, their behavior tells an important story. Human Security's traffic analysis found that from January to August 2025, approximately 87% of all pages browsed by agents were related to products , but the majority of these interactions are related to product research and recommendations, not full transactions, with just 2.2% of agents interacting with shopping carts, checkout, and payment funnels . Agents are prolific window-shoppers. They are not, yet, buyers. Britton frames this pattern for his keynote audiences as a classic technology adoption curve, except compressed and reversed. Normally infrastructure lags consumer demand. Here, infrastructure has lapped it. His talks on AI and consumer behavior walk Fortune 500 leaders through why spending on agent-ready checkout without parallel investment in trust-building is a recipe for stranded capital.

The data backs the caution. A recent Checkout.com report adds important context to the merchant side of this story. PYMNTS Intelligence, surveying more than 2,000 US consumers and 60 merchants, finds that around 132 million American adults have used AI to help with a retail purchase, but 59% of those AI-assisted purchases still end on Amazon. Discovery is agentic. Checkout, mostly, is not.

Consumer Trust in AI: The Real Bottleneck to Agentic Commerce ROI

Consumer trust in AI agents to complete financial transactions remains the single largest barrier to agentic commerce ROI, and it is not closing quickly. Forrester's benchmark research found only 24% of US online adults trust AI agents to act on their behalf to make routine purchases . A follow-up Forrester survey sharpened the picture further: three-quarters of consumers say they are uncomfortable allowing an AI agent to complete a purchase and pay autonomously, even if they could set spending limits and rules in advance . Trust is not just stalling, in some measures it is reversing. Riskified's quarterly tracker shows that consumers' comfort with agents making purchases fell from roughly 70% in late 2025 to just 45% by early 2026 . That is a significant regression for a technology category enterprises are betting billions on. Checkout.com's own research reinforces the ceiling: one in four (24%) consumers say they will never delegate purchases to AI, and 27% trust no organization to operate an AI shopping agent.

Four concerns surface repeatedly across the research, and Britton cites all four in his boardroom briefings:

• Loss of control over the final purchase decision and spending amount
• Errors and liability when an agent buys the wrong item or quantity
• Distrust in agentic decision-making, particularly around undisclosed sponsored recommendations
• Data security and privacy concerns about what information agents access and share

Liability sits at the center of the stalemate. Merchants and AI providers cannot agree who eats the loss when an agent gets it wrong. In the PYMNTS merchant survey, 93% said the AI or agent provider should bear the loss when an agent makes a wrong purchase, 80% expect providers to verify an agent's authority, and only 28% are willing to offer agents their full product range on the same terms as other channels. Until that question is resolved contractually, enterprises will keep building rails that few consumers are ready to ride.

ChatGPT Commerce and the Discovery-Over-Checkout Pivot

ChatGPT commerce offers the cleanest case study of why discovery is winning while checkout stalls. OpenAI's original Instant Checkout launch promised native, in-chat purchasing. Within months, the company pulled back. OpenAI said the initial version of Instant Checkout did not offer the level of flexibility it aspired to provide, so it allowed merchants to use their own checkout experiences while focusing its efforts on product discovery.

The reason behind the pivot is instructive for every enterprise watching OpenAI as a bellwether. OpenAI observed that product discovery within ChatGPT drove more significant user engagement than actual conversions, prompting a strategic redirection focused on enhancing discovery features. Even the company building the most advanced consumer AI interface on earth could not force checkout adoption faster than consumer comfort allowed. This mirrors what Britton has long argued on The Speed of Culture podcast: technology companies consistently overestimate how quickly consumers will hand over transactional control, even when the user experience itself is frictionless. The lesson for retail, finance, and real estate leaders exploring agent partnerships is to prioritize AI-assisted research experiences now, since that is the stage consumers actually trust, and treat autonomous checkout as a multi-year behavioral build rather than a feature flag to flip on. Industry-specific implications vary widely, whether a brand operates in the financial services space managing payment liability or in real estate where agentic tools are reshaping lead qualification rather than transactions.

Where Agentic Commerce Is Actually Working Today

The gap between hype and behavior is not uniform across every product category or demographic. Some segments show real, measurable momentum, and enterprises chasing agentic commerce ROI should concentrate resources there first rather than spreading thin across every category simultaneously.

Low-risk, repeatable purchases are leading the way. The findings suggest that agentic commerce will scale unevenly, starting with low-risk, repeatable categories before expanding further, with consumers most willing to delegate shopping for groceries (41%) and household supplies (31%). Convenience, not novelty, is driving the willingness that does exist. Agentic commerce for consumers is being driven by convenience as they want to automate routine tasks to get time back, with one in four (25%) saying saving time is the top motivator for them to use an AI shopping agent.

Younger consumers and higher-value AI-assisted shoppers also show distinct behavior worth watching. Research shows that millennials are notably more receptive when safeguards exist: forty-two percent of American millennials said they would allow AI agents to buy items for them within a $250 budget if they could return the purchase within seven days. This is a brand strategy for any CMO targeting Gen Z or millennial consumers, a space Britton covers extensively through his Gen Z keynote speaker platform: build return policies, spending caps, and transparency features into the agent experience itself rather than treating trust as a marketing message layered on top.

Conversion quality among AI-assisted shoppers who do convert is also meaningfully higher than skeptics assume. AI‑sourced traffic to US retail sites was up 89% YoY during peak events like Prime Day 2026, and those visitors converted 40% better than non‑AI channels. The opportunity is real. It is simply concentrated and conditional rather than universal, which is exactly the nuance that gets lost when headlines compress "agentic commerce" into a single, monolithic trend.

Key Takeaways for Business Leaders

• Audit your agentic commerce investment against actual transaction data before expanding checkout infrastructure further, since discovery and conversion require fundamentally different trust thresholds.
• Prioritize low-risk, repeatable product categories like groceries and household goods for early agentic checkout pilots, where consumer willingness is already highest.
• Build spending limits, return guarantees, and transparent disclosure into every agent experience, since liability clarity is what unlocks consumer willingness, not feature sophistication.
• Treat AI-assisted discovery as the primary near-term ROI opportunity, following OpenAI's own pivot away from forced in-chat checkout toward richer comparison and research tools.
• Prepare leadership teams now for a multi-year adoption curve by bringing in outside expertise, such as a dedicated AI keynote speaker, to align the C-suite on realistic timelines before budget cycles overcommit.

Frequently Asked Questions About Agentic Commerce

What percentage of transactions actually involve AI shopping agents today?

Current merchant data shows only about 3% of transactions involve AI agents, despite 89% of merchants actively preparing infrastructure for agentic commerce. The gap exists because consumers use AI heavily for product research and comparison but still complete most actual purchases through traditional checkout, often redirecting to familiar platforms like Amazon even after AI-assisted research.

Why don't consumers trust AI agents to make purchases?

Forrester research found only 24% of US online adults trust AI agents to make routine purchases on their behalf. The primary concerns are loss of control over spending, unclear liability when agents make mistakes, undisclosed sponsored recommendations, and data privacy around what information agents access during the shopping process.

What happened to ChatGPT's Instant Checkout feature?

OpenAI scaled back Instant Checkout after finding that product discovery drove far more consumer engagement than actual in-chat conversions. The company shifted its strategy toward richer product discovery tools, letting merchants handle checkout through their own existing systems rather than forcing native in-chat purchasing.

Which product categories are seeing the most agentic commerce adoption?

Low-risk, repeatable purchases lead adoption, with 41% of consumers willing to delegate grocery shopping to AI agents and 31% willing to delegate household supply purchases. High-consideration purchases like electronics and apparel still see far more AI-assisted research than AI-completed checkout.

The Bottom Line on Agentic Commerce

The infrastructure race is over for now. Stripe, TikTok, Mastercard, Constructor, and Gap proved in a single week that the technical capability to run agentic commerce at scale already exists. What remains unproven is whether consumers will trust that capability fast enough to justify the billions enterprises are pouring into it, and the data says that gap will not close on its own.

Matt Britton built his career translating exactly this kind of signal for Fortune 500 leadership teams, turning consumer behavior data into strategic clarity long before competitors catch on. Companies that win the next phase of commerce will not be the ones who deployed agents first. They will be the ones who understood the trust gap, built for it deliberately, and earned consumer confidence one transparent, low-risk interaction at a time.

For organizations ready to align leadership around what agentic commerce really requires, Britton's AI keynote speaker presentations turn this research into an actionable roadmap. Visit Matt Britton's speaker platform to book a keynote, or explore Generation AI for a deeper look at how enterprises can close the distance between AI ambition and consumer reality.