Public Sapient and Salesforce have released a whitepaper presenting a comprehensive technology strategy to help business and technology leaders to prepare for agentic commerce, or what they describe as the next major evolution in retail and commerce where AI agents handle the cognitive tasks of shopping, such as product comparison and validation.

According to the paper, agentic commerce is moving at an unprecedented pace. While the widespread adoption of the Internet took two decades, and mobile phones eight years, agentic commerce is expected to compress into just 24-26 months. This is because agents leverage existing infrastructure, including APIs, payment systems, and fulfillment networks built in place, preventing the need for businesses to rebuild foundational systems from scratch or wait for new digital rails to be established.

The A.C.E. Framework

To support this shift, the paper introduces the A.C.E Framework, an architectural model for building agent-ready commerce ecosystems across three critical layers.

First, the Agentic Experience Interface focuses on enabling products and services to be discoverable and actionable by AI agents. Second, the Composable Micro-Apps expose core commerce capabilities through modular, interoperable services. Finally, underpinning both is the Enterprise Context Orchestration, which represents the central nervous system of the agentic system. Situated between the micro-apps and legacy enterprise systems, this layer governs data integrity, security, compliance, and contextual intelligence across the entire agent ecosystem.

Beyond the architectural foundation, the paper also identifies three emerging categories of AI agents that organizations must increasingly optimize for.

The first category is the platform agents, such as ChatGPT and Google’s Gemini. These platforms sit between consumers and merchants, effectively functioning as the new “headless browser” for the AI era. As they control an increasingly dominant share of product discovery, becoming agent-addressable on these surfaces has become critical for businesses. ChatGPT, for example, processes 1 billion web searches weekly, with several sources claiming approximately 50 million shopping-intent queries every day.

The second category is brand agents. These systems operate on merchant-owned assets, serving customers directly within the brand’s ecosystem. Unlike platform agents, they offer differentiated experiences, capture first-party data, and strengthen direct relationships.

Notable examples include Sephora Virtual Artist, an augmented reality feature in Sephora’s app that allows users to try on makeup products in real time and which formulates product recommendations; and Zalando Advisor, an AI-powered conversational tool providing personalized fashion advice and product recommendations based on the customer’s questions, context, and shopping history.

Finally, the third category is personal consumer agents. These agents act on behalf of individual shoppers, managing preferences, executing purchases, and coordinating across multiple merchants. They are being enabled by platforms like ChatGPT Projects, and Gemini Personal Agents, alongside automation tools and open-source ecosystems that allow for self-hosted, action-capable agents, and long-running effort.

Though this category is still in its infancy, it is evolving quickly and warrants close monitoring from business leaders.

Agentic commerce market outlook

Agentic commerce refers to a shopping experience powered by intelligent AI agents capable of anticipating, personalizing, and automating every step of the process to create frictionless, proactive experiences. It’s enabled by cutting-edge advancements in AI, and represents a transformative shift in the entire commercial ecosystem.

Consumer behavior is already evolving in response to these capabilities. According to a 2025 survey by McKinsey, 44% of users who had tried AI-powered search said that it had become their “primary and preferred” source for Internet searching, compared with 31% who prefered using traditional search.

In Europe, AI adoption shows up most strongly upstream of the transaction. In particular, European consumers most frequently use AI for comparison brands, models, prices and reviews (63%), learning about a category or product (55%), and discovering new products and getting inspired (46%), according to separate 2025 study by McKinsey.

However, trust drops sharply when it shifts from supporting judgment to taking decisions, particularly when using AI to pre-fill baskets, complete checkout, or automatically reorder items.

Despite current challenges, McKinsey forecasts robust growth for the sector. Given the growing availability and current adoption levels of AI-powered discovery tools, it estimates that by 2030, the global B2C retail market alone could see between US$3 trillion and US$5 trillion in orchestrated revenue from agentic commerce.

Featured image: Edited by Fintech News Switzerland, based on image by DC Studio via Magnific