January 19, 2026

The question has changed. A week after NRF, retailers are no longer debating whether agentic commerce will reshape their industry. They are fighting over who controls the layer between customer and checkout. The acquisition of Manus by Meta for $2 billion signals how high the stakes have climbed. Manus agents can break down goals into tasks, interact with multiple data sources, and deliver completed work with minimal prompting. Integration into Facebook, Instagram, and WhatsApp will create end-to-end shopping flows where discovery, negotiation, and purchase all happen without leaving the feed. The social commerce wars just became agentic commerce wars.

Consumer behavior is validating the thesis. An Accenture survey cited by Observer found that 74% of consumers abandoned shopping baskets in the previous three months because they felt bombarded by content, overwhelmed by choice, and frustrated by the effort required to make decisions. The 2025 holiday season offered a preview of the alternative. One in three shoppers used AI tools to generate gift ideas, compare prices across stores, style outfits, and build personalized wishlists. Among Gen Z, a majority did so. What once required thirty open browser tabs now happens in a single adaptive conversation. Consumers will delegate friction to whatever system earns their trust.

Observer frames this as a fundamental shift in who the customer actually is. Retailers now serve two buyers simultaneously: the human who ultimately approves the purchase, and the AI system that filters, evaluates, and increasingly authorizes or executes the transaction on their behalf. Brands that optimize for both can shape the path. Those that ignore the agent become interchangeable SKUs competing mostly on price. Observer's language is blunt. When merchants "embrace the fact that the most important buyer in the market is no longer a person, but the A.I. that earns that person's trust, they move back in the driver's seat."

The protocol layer is already generating controversy. The Interline's coverage of Google's Universal Commerce Protocol highlights a feature called "Direct Offer" that tailors discounts at an individual-agent level. Critics argue this crosses the line from personalized pricing into legally questionable surveillance pricing. The mechanism is technically elegant but ethically murky. An agent negotiates a price that a human shopper cannot independently verify or compare. Regulatory attention seems inevitable. For retailers adopting UCP, the early-mover advantage comes bundled with potential scrutiny over opaque, individualized pricing that only machines can see.

Shopify is positioning itself carefully between full agentic embrace and brand preservation. President Harley Finkelstein told Retail Brew that agentic commerce is a "new, highly meaningful spoke" around the online store hub where videos, newsletters, and product drops still live. He raised the possibility that agent-mediated discovery could lead to "merit-based shopping" where AI weighs specs, customer satisfaction, and durability more heavily than brand marketing. For merchants with strong fundamentals, this is liberation from the advertising tax. Those dependent on marketing to compensate for mediocre products should be worried.

The B2B side may see even more profound disruption. Presta describes AI agents functioning as "Virtual Procurement Officers" that monitor stock levels, detect potential shortages, autonomously negotiate with approved vendors, and manage purchase orders end-to-end. They call this "Intelligent Auto-Scaling" for physical goods. Second Talent's statistics compilation projects that 20% of B2B transactions will be driven by autonomous agent-led negotiations between buyers and sellers by year end. Manufacturing is expected to see factory agents negotiating production schedules and reducing maintenance costs by roughly 25%. The "cost of coordination" that has defined enterprise procurement for decades could compress toward zero. If these projections hold, the human role in B2B purchasing shifts from transaction execution to policy governance and exception handling.

The market projections keep accelerating. The dedicated market for autonomous AI and agent software will reach $11.79 billion in 2026 according to Second Talent. Gartner forecasts that 40% of enterprise applications will use AI agents by year end, up from less than 5% in 2025. Companies plan to double their AI spending this year, with over 30% of that budget directed specifically to agentic AI. About 90% of executives expect agents to deliver measurable returns within the calendar year. This is no longer venture capital optimism. It is CFO-approved operational spending.

The ROI case is becoming concrete. Organizations deploying agents report 40-60% reductions in operational workflow costs compared to brittle legacy RPA. Contact centers using autonomous agents see cost-per-contact drop 20-40%. Retail's "invisible agents" running inventory and pricing yield an approximate four-percentage-point uplift in in-store conversion. These are not futuristic projections. They are early deployment numbers from companies that moved first. eMarketer captures the strategic implication cleanly. According to one executive, the competitive battlefield is moving "from human attention to agent-to-agent negotiation." Another put it even more simply. "Consumers won't browse; they'll delegate." The retailers who understand that distinction are building for it now. Those who do not will spend the next few years wondering where their customers went.