March 9, 2026

Retailers are knowingly trading direct customer access for reach. Retail Dive reports that Etsy, Target, and Walmart all expanded their AI platform partnerships in early 2026, placing catalogs into Google's Gemini, Microsoft's Copilot, and OpenAI's ChatGPT. They are doing this, Retail Dive notes, "even at the risk of losing direct access to customers and control over their data." Amazon runs Rufus inside its own ecosystem while simultaneously feeding external platforms. The parallel bet is transparent. Maximize discovery through AI channels you do not own while trying to retain direct relationships for everything else. Deloitte research cited by Retail Dive makes the stakes explicit. About half of retail executives anticipate the multistep customer shopping journey will collapse by 2027, replaced by "a singular AI-driven interaction." The funnel is not being optimized. It is being eliminated.

Stefano Mormone gives the clearest name to what is happening. He calls it "off-site commerce," where the transaction migrates to a "blank page," a neutral, AI-curated space where every product sits directly alongside every competitor. Brand sites remain the home of storytelling, but the actual purchase happens in a layer controlled by the AI platform. Emotional appeal is no longer enough. Mormone argues success now requires "data integrity," meaning real-time, machine-readable product specs that survive the AI filter. Simona Fricano, commenting on NRF 2026 in the same thread, frames this as an operating model shift, not a technology shift. Fricano argues that executives must now own questions they previously delegated. Data ownership, decision logic, and pricing authority all change shape when AI agents become the primary customer interface, and margin models built around click-based funnels stop working in a journey with no clicks at all. The harder question is where automation should accelerate and where human judgment remains a genuine differentiator. Her warning is direct. "The real risk is not moving too fast, but moving in silos."

Sharper answers are emerging about where agents win and where humans still matter. In a thread aggregated by Peterson Technology Partners, Anthony Sapountzis of Aerion Technologies identifies long-tail discovery as the current sweet spot. "When an unknown random product like obscure rubber tap connectors are needed and the shopper has no idea where to begin the search, AI nails it in 30 seconds." But he adds a caveat that cuts against the hype. "Bots can research well, but vibe, taste, context often still needs humans. This is the messy middle stage before agentic commerce feels frictionless." Tasman Analytics, in the same thread, sharpens the point further. "Every demo uses travel, which is ironically the last thing anyone would delegate to AI. Picking your seat, frequent flyer perks, and hotels is highly individual experience." Routine, specification-driven purchases are ready for full agent autonomy. High-value, high-taste decisions are not, and the industry needs to stop pretending otherwise.

HelloRep argues the counter-move is building brand-owned agentic experiences, positioning its product as the solution. The pitch deserves scrutiny, but the underlying logic holds. Rather than ceding the entire customer journey to platform agents from Google or OpenAI, brands can deploy their own agents that carry context about loyalty tiers, promotional offers, and personalized pricing. Platform agents see catalogs as structured data in comparison tables. Brand-owned agents understand margin rules, inventory constraints, and individual purchase history. The proposed architecture is agent-to-agent, where a brand's agent becomes its representative in every AI-mediated interaction, surfacing information that generic platform agents cannot access or prioritize. Whether HelloRep or anyone else delivers this capability, the strategic need is real. Merchants who only exist as rows in a platform agent's comparison matrix have no leverage over how they are presented.

On the B2B side, agentic supplier discovery is going mainstream. Find My Factory reports that autonomous agents scanning 125 million company profiles continuously are reducing the research-to-quotation phase by up to 80%. Static supplier vetting is giving way to 24/7 monitoring of financial health, geopolitical risks, and compliance status, eliminating the three-to-six-month blind spots of annual audits. The same analysis finds that 86% of organizations plan to scale their AI implementations by the end of 2026, with 40% embedding agents directly into operational workflows. That report also describes standardized agent protocols enabling "buy in chat" scenarios where procurement agents create purchase orders, trigger workflows, and manage supplier interactions from within conversational interfaces.

The control question is now the defining question in agentic commerce. Retailers are placing catalogs into AI ecosystems they do not control, in exchange for visibility they cannot afford to lose. Many will find the trade worthwhile. But the retailers who build brand-owned agent experiences alongside platform distribution are the only ones whose strategy does not end in dependency. Deloitte's prediction that the shopping journey collapses into a single AI interaction by the end of 2027 is now less than two years away. The question is whether retailers will be hosting that interaction or serving as the fulfillment endpoint behind it.