CIO.com's Agentic Commerce Roadmap — How Enterprises Should Prepare for the End of Browse-and-Click

CIO.com's Series Maps the Full Picture of Agentic Commerce

Agentic payments are coming. Is your company ready?

Major tech companies and payment platforms are rolling out capabilities that allow AI agents to autonomously execute transactions on behalf of users.

www.cio.com

From late 2025 through 2026, CIO.com has published a comprehensive series exploring the structural shift toward agentic commerce through four distinct perspectives. Avery Dennison CIO Nick Colisto, Kendra Scott's digital leader Kamanasish Kundu, Trulioo CTO Hal Lonas, and tech journalist Maria Korolov each bring different angles, but read together, the four articles form a cohesive enterprise implementation roadmap.

"Your Next Customer Might Be an Algorithm"

Colisto opens his article with a fundamental observation: "Your next customer might not be a person at all. It could be an algorithm shopping on behalf of one." This is not speculation but data-backed reality.

Adobe's research found that traffic to US retail sites from generative AI browsers and chat services grew 769% year-over-year by November 2025. AI-referred traffic converts 31% higher and generates 45% longer session times compared to other sources. A BCG study from January 2026 further revealed that 43% of consumers already use generative AI for product research and purchase recommendations.

Colisto's core insight is that competition is shifting from "human clicks" to "algorithmic trust." Beautiful UI and marketing copy matter less than structured data and API quality that AI agents can interpret. Brand visibility increasingly depends on machine-readability.

The Three-Stage Maturity Model

Kundu outlines a three-stage maturity model for agentic commerce evolution in his article.

This model gives e-commerce businesses an objective benchmark for their current position. Most companies remain at Level 1, offering recommendations based on purchase history. But Kundu argues that Level 3's "Delegated Action" is the essence of agentic commerce.

At Level 3, a consumer simply says "replenish my running shoes," and the AI agent autonomously handles size selection, price comparison, payment, and delivery. Kundu calls this "zero-click commerce," stating that "the moment a customer signals interest through an agent, the transaction is already 90% complete."

Structural Transformation of Payment Infrastructure

Korolov's March 2026 article tackles the most concrete implementation challenges. It reports how technology companies and payment networks are rapidly deploying capabilities for AI agents to autonomously execute transactions.

Visa completed pilot programs in late 2025 and designated 2026 as the "year of mainstream adoption" for agentic commerce. Over 100 partners are collaborating, with 30+ developing in sandboxes. Meanwhile, Google announced UCP (Universal Commerce Protocol) at the January 2026 NRF conference, securing support from 20+ global partners including Shopify, Etsy, and Wayfair.

Particularly notable in Korolov's reporting is that existing fraud detection frameworks face fundamental disruption. Blackhawk Network's Nik Sathe warns that "traditional fraud signals like IP addresses and device fingerprints become ineffective in agent-mediated transactions." Payment infrastructure standardization also compels acquirers to build new guardrails.

The Unsolved Challenge of Agent Authentication

Lonas identifies three structural challenges blocking agentic commerce in his article.

The first challenge is agent authentication at scale. In a world where millions of agents attempt transactions daily, individual human verification is impossible. AI-to-AI authentication mechanisms are needed, but balancing rigor with flexibility remains difficult.

The second challenge is the merchant adoption gap. While large retailers move quickly on agent readiness, small businesses lack the capital for infrastructure investment. Korolov's article also cites Bobo Design Studio's Angie Chua reporting unauthorized data scraping by AI agents. The agent economy is expanding without reliable means to distinguish legitimate agents from malicious bots.

The third challenge is agent discoverability. Lonas predicts agent directories will evolve into something resembling app stores. A federated directory system searchable by name, capability, certification, and identifier would function similarly to the domain name system.

Data Quality as the "New Storefront"

A message common to all four articles is that data quality is the competitive advantage of the agent era. Colisto declares "data quality is the new storefront," while Kundu warns that "brands that are not indexed, accessible, and API-ready will become invisible."

These observations further reinforce the importance of data readiness for agentic commerce. McKinsey's October 2025 report projects that agentic commerce will influence $3-5 trillion in global retail sales by 2030. Given this market scale, data infrastructure is not a "someday" initiative but an immediate business priority.

PayPal's CEO predicts that 25% of e-commerce spending will be agent-driven by 2030. Bernstein analysts estimate agent-driven experiences could lift global e-commerce conversion rates by 1.5-2.5% annually, generating over $240 billion in incremental revenue.

The Enterprise Implementation Roadmap

Integrating the four articles reveals three phases for e-commerce implementation.

Phase 1: Machine-Readable Data

Structure product catalogs as machine-interpretable data. Explicitly describe attributes like size, dimensions, compatibility, and materials, adding semantic summaries explaining "who it's for" and "what problem it solves." Platforms like Adobe Commerce already offer UCP-compatible features worth leveraging.

Phase 2: API-First Design

Expose product data, inventory, pricing, and fulfillment status through APIs that AI agents can access programmatically. As Colisto notes, "retailers are reconfiguring their tech stacks to expose product data through standardized APIs." To address gaps in payment infrastructure, API design must consistently cover payment flows as well.

Phase 3: KYA (Know Your Agent) Framework

Build systems to identify, authenticate, and monitor AI agents accessing your digital channels. Lonas's phased approach provides a useful model: start with low-risk transactions (lunch orders, household replenishment) with human approval gates, then gradually increase agent autonomy as trust is established.

Conclusion

CIO.com's four-part series makes clear that agentic commerce has moved from discussion to implementation.

Adobe's 769% traffic increase from AI sources, Visa's 2026 mainstream adoption declaration, and PayPal's CEO forecasting 25% of e-commerce spending as agent-driven by 2030 all signal that the pace of change exceeds many companies' expectations.

At the same time, significant challenges remain: fraud detection obsolescence, the small-business adoption gap, and immature agent authentication standards. As Lonas notes, agentic commerce adoption will be gradual, not overnight.

For e-commerce businesses, the critical shift is treating agentic commerce as infrastructure work, not experimentation. Machine-readable data, API-first design, and KYA frameworks are investments that strengthen digital operations regardless of how quickly agentic commerce matures. The time to start is now.