Across retail, payments, and big tech, AI-enabled checkout, automated product comparison, and autonomous purchasing tools are becoming critical for e-commerce. The industry has moved beyond talking about AI to managing the transition to agentic commerce.
The audit and assurance, consulting, and tax services firm PwC sees agentic commerce influencing what products get surfaced, compared, and chosen, often without humans in the loop. This change is rewiring retail power, loyalty, and visibility in ways many brands are not yet prepared for.
The National Retail Federation’s recent research insights and its latest Digital Trends reports show that AI agents are rapidly moving from experimentation to execution in commerce. The findings suggest that agentic commerce is no longer a future concept. It is the new infrastructure, anchored by protocols like Google’s Universal Commerce Protocol.
The emergence of agentic commerce is creating a new retail environment where the gatekeeper isn’t the store selling the product but the digital entity that decides whether the product is even seen.
Eric Shea, principal for customer and commercial excellence at PwC US, observed that what changed this year is that the infrastructure stack finally started to mature at scale. AI agents are no longer operating as isolated chat interfaces.
“They are increasingly connected to real-time inventory, pricing, fulfillment, payments, identity, and product data ecosystems through standardized protocols and API frameworks,” he told the E-Commerce Times.
Drivers of the Gatekeeper Trend
According to Shea, the emergence of interoperable commerce standards is helping AI agents move from answering questions to executing transactions. At the same time, retailers have spent years modernizing cloud, data, and digital commerce infrastructure, which is now enabling non-human agents to act on behalf of consumers in real-world environments.
This trend also reflects changes in consumer behavior. Shoppers increasingly expect commerce to be frictionless, personalized, and conversational. Agents are becoming the orchestration layer between intent and transaction, he added.
“We’re also beginning to see measurable behavioral shifts already emerge, including increased shopping intent within AI prompts and growing referral traffic from AI platforms to retail and brand websites,” Shea said.
As AI agents begin to handle end-to-end processes from discovery to checkout, the definition of customer loyalty changes when the decision-making shopper is an algorithm rather than a human. Customer loyalty is evolving from emotional affinity alone to trust earned through algorithmic recommendations.
“Historically, brands competed for attention and emotional connection with human shoppers. In agentic commerce, brands also need to compete for machine confidence,” he noted.
The New Economics of Discovery
AI agents typically provide a highly curated shortlist of one to three items rather than a page of search results. Shea estimated that 25% to 40% of shoppers already use these tools.
As a result, the economics of digital discovery are beginning to change. AI agents are collapsing traditional search lists into a single decision loop, undermining the billions of dollars brands have invested in top-of-funnel web traffic.
He noted that traditional funnels were built around generating clicks, impressions, and site visits. Agentic commerce compresses that journey into a much smaller decision loop, in which the agent curates options before the consumer ever reaches a website.
“We are moving toward a world where discoverability depends less on who wins the click and more on who earns the recommendation,” Shea predicted.
He added that brands may need to prioritize structured product data, trust signals, interoperability, and post-purchase performance. They will need to place these marketing factors alongside brand and product storytelling embedded in that data and across owned digital channels.
“In many ways, organizations optimized heavily for transactions and conversion over the past decade, sometimes at the expense of richer brand narrative and contextual content that AI agents increasingly rely on to interpret value,” he said.
Cost of Missing the Agentic Cut
According to Shea, the competitive pressure intensifies significantly in a shortlist economy. Historically, search engines and marketplaces gave consumers dozens of visible options. AI agents compress that visibility into a very small set of recommendations.
“Brands that fail to become machine-readable, trusted, or operationally competitive risk becoming effectively invisible in certain shopping journeys. That visibility gap can directly affect traffic, conversion, and long-term brand relevance,” he said.
He warned that this dynamic may intensify as AI systems become more personalized and consumers increasingly default to the top recommendation. AI-enabled wearables, voice interfaces, and ambient commerce experiences will grow, leaving less space and time for long recommendation lists.
“This creates a more pronounced winner-takes-most environment over the next one to three years,” Shea predicted.
How to Make Brands Visible to Agents
Shea noted that poor data quality is the biggest barrier to ROI. For a brand to be visible to an agent, it must provide structured, machine-readable signals. Agent SEO may become the most important discipline in retail marketing.
“Generative engine optimization (GEO) is quickly becoming foundational, but it is broader than traditional search optimization. This is really about machine-readable commerce readiness,” he explained.
AI agents need clean, structured, trustworthy, and increasingly near-real-time data to evaluate products, compare alternatives, and make recommendations. Brands risk being excluded from the recommendation layer if pricing, specifications, inventory, delivery windows, reviews, or fulfillment data are inconsistent or incomplete.
“In many ways, the future of visibility may depend less on keyword optimization and more on operational data quality across the enterprise,” he said.
Reconfiguring Trust, Ads, and Influence
Brands must adapt marketing strategies to favor AI agents as discovery veers from paid ads to algorithmic trust. This involves the same ways they build trust with consumers: consistency, transparency, and reliability, Shea offered.
Agents evaluate signals such as product accuracy, fulfillment reliability, customer satisfaction, return experiences, review authenticity, sustainability claims, and pricing consistency. The stronger and more verifiable those signals are, the more likely a brand is to be recommended repeatedly, he detailed.
“This shifts the conversation from pure advertising spend toward enterprise trust architecture. Every operational touchpoint becomes part of the brand’s discoverability strategy,” said Shea.
He added that the most important trust signals are typically those tied directly to consumer outcomes and transaction confidence. That includes inventory accuracy, verified reviews, fulfillment reliability, pricing transparency, return simplicity, and product consistency.
“Agents are designed to reduce friction and minimize the likelihood of a poor customer experience. If a brand consistently delivers accurate information and reliable fulfillment, the agent has more confidence recommending it,” he said.
Shea added that trust signals are not limited to a brand’s owned properties. AI systems continuously evaluate signals across social platforms, creator ecosystems, reviews, forums, and third-party sources. So brands must syndicate, monitor, and moderate content consistently across the broader digital ecosystem.
Where Human Influence Still Matters
According to Shea, as agents become the primary researchers, the era of the human influencer is morphing. That change is not replacing human influence. It is redefining it.
“Human creators, communities, and social content still shape consumer preference and cultural relevance. But AI agents increasingly act as interpreters of that information,” he said.
That means technical documentation, structured product content, verified reviews, and machine-readable attributes become significantly more important because agents rely on them to make recommendations, he clarified. In the future, brands may need to optimize simultaneously for emotional influence with humans and informational clarity for machines.
Brand Trust in an AI Shopping World
Shea sees trust scoring becoming more dynamic and predictive over time. It will likely combine operational performance, customer sentiment, and contextual relevance.
Trust will ultimately determine whether agentic advertising succeeds or fails. PwC advises brands to balance sponsored placements in an agent’s response without breaking users’ trust in their AI assistants.
“Consumers are willing to accept recommendations and sponsored placements if they believe the system remains transparent and aligned to their interests,” Shea said.
He suggested that brands and platforms should clearly distinguish sponsored recommendations from organic ones and ensure that advertising does not compromise the quality of recommendations. If users begin to feel that agents are prioritizing monetization over relevance, trust erodes quickly.
“That sensitivity is one reason why many AI platforms today have been cautious about placing traditional ad units directly alongside prompts and responses. Preserving user trust and recommendation integrity remains foundational to long-term adoption,” he offered.
Shea added that the long-term winners will likely be organizations that balance commercial opportunity with transparency, governance, and consumer confidence.
