The AI Agent Basket: How Agentic Commerce Is Reshaping B2C Digital Economy Spending in 2026

The AI-Assisted Shopper Is Already Here

The standard story about agentic commerce is still “it’s coming in 2027 or 2028.” The data from Q1 2026 contradicts that. Adobe Analytics’ tracking shows AI-driven traffic to US e-commerce sites grew 393 percent year-over-year in the first quarter of 2026. Salesforce’s own measurement recorded a 119 percent increase in AI-assistant-driven retail traffic during H1 2025 — numbers that accelerated, not slowed, as the year progressed.

The best benchmark for what this means at the order level comes from Salesforce’s Cyber Week projections: intelligent agents were expected to drive 22 percent of global orders during the holiday period. In parallel, one 2026 agentic commerce study found that 45 percent of shoppers now use AI for at least part of their buying journey.

What Changes in the Basket

The economics of an AI-agent-initiated purchase differ meaningfully from a human-browsed one. Adobe’s 2026 data describes the shift:

• AI referrals produce 254 percent more revenue per visit than traditional traffic
• Shoppers referred by AI spend 45 percent more time on-site
• AI-referral traffic converts 31 percent higher than baseline

Agentic commerce specifically — where the AI agent is not just advising but executing the purchase — is projected to lift average order value by 15-25 percent for early adopters, with retailers targeting more than 15 percent AOV uplift as a baseline integration goal, according to industry analysis.

The compounding is substantial. If a retailer had a $X AOV with a 2 percent conversion rate on traditional traffic, agentic traffic can plausibly deliver $1.2X AOV at a 2.6 percent conversion rate — more than 50 percent more revenue per visit.

The Protocol Layer Matures in 2026

The infrastructure question was “how do AI agents actually purchase?” — and in 2026 it got answers. Adobe Commerce publicly committed to agentic commerce standards with protocol support for machine-readable product catalogs, verified agent identities, and authenticated transaction flows. Google’s Universal Commerce Protocol, Salesforce’s Agentforce Commerce, and Visa’s Intelligent Commerce SDK all rolled out parallel stacks.

The significance is that merchants no longer have to build one-off integrations per AI agent platform. A retailer that implements the emerging agentic commerce schema inherits compatibility with ChatGPT, Perplexity, Claude, Gemini, and emerging pure-play shopping agents simultaneously.

The Measurement and Attribution Problem

Agentic commerce breaks traditional marketing attribution. When a shopper asks an AI agent “what’s the best sunscreen for sensitive skin under $30,” the agent may consult 8 sources, reason about the user’s context from prior conversations, and execute the purchase on one retailer — without any classical “referral” attribution signal that marketers can measure.

This is why retailers are investing in “AI Engine Optimization” (AEO) — the successor to SEO focused on ensuring brand and product surfaces are discoverable, citable, and purchasable through AI agents. RingLy’s 2026 retail analysis notes the dual pressure: SEO budgets are being reallocated toward agentic optimization, and traditional media mix modeling needs an agent-exposure dimension that did not exist in 2024.

Why B2C Gets Disrupted First

B2C commerce has three traits that make it the natural first market for agentic disruption. First, the transactions are small and repeatable — low-risk enough for a shopper to delegate to an AI. Second, the product catalogs are well-structured — SKUs, prices, reviews — which AI agents can parse reliably. Third, consumer preferences are personal and difficult to articulate in search queries but easy to express to an AI that remembers prior context.

Forrester’s B2C analysis published in 2026 maps the current state: agentic commerce is past the experimentation phase, into production in specific verticals — consumer electronics, beauty, home goods, grocery replenishment — and is reshaping the competitive map faster than the retail incumbents expected.

Who Wins, Who Loses

Three retailer archetypes are in different positions:

• Large platforms with rich data (Amazon, Walmart, Alibaba) win because their catalogs are the default product universe AI agents consult.
• Brands with direct-to-consumer channels win if they invest in AEO and structured product data — agents can surface them for specific needs.
• Mid-market retailers with thin digital presence are structurally exposed — agents bypass them unless they adopt the protocols.

The practical implication for smaller retailers in markets like North Africa is stark: the 15-25 percent AOV uplift is accessible only to those who implement the emerging schemas. Retailers still running static HTML catalogs in 2027 will be invisible to the agents shopping on their customers’ behalf.

The Longer Arc

Morgan Stanley’s projection — nearly half of online shoppers using AI agents by 2030, accounting for 25 percent of their spending — is aggressive but plausible given Q1 2026 growth rates. The question for the industry in 2026 is less “will this happen” and more “how fast does the schema ecosystem consolidate, and who owns the identity layer for AI-executed purchases.”

Sources & Further Reading

• Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites — Adobe Business Blog
• Salesforce Unveils Agentforce Commerce Capabilities — Salesforce News
• Adobe Commerce Commits to Agentic Commerce Standards — Adobe
• Agentic Payments in B2C Commerce: Where We Are Now — Forrester
• How Agentic Commerce Will Transform Buying Decisions in 2026 — Codewave
• 42 AI in Retail Statistics You Need to Know in 2026 — RingLy
• What Is Agentic Commerce? 45% of Shoppers Use AI — Ekamoira
• Agentic Commerce: How AI Shopping Agents Can Change Retail — McKinsey