Agentic commerce: How to win when algorithms do the shopping

This is where traditional marketing thinking breaks down. Brand-building has long relied on emotional persuasion—compelling stories, beautiful imagery, carefully cultivated affinity. It still works for humans, but AI agents don't respond to any of that.

As Tomorrow's Commerce puts it: "You're not dealing with human customers who can be won over with a good brand story or the right appeal to sentiment. Agentic commerce is dispassionate and data-driven."

"Think of it this way," says Paul Appelhof, Agentic Commerce consultant at VML. "An AI agent evaluating your product doesn't care about your heritage, your campaign, or your brand ambassador. It cares about whether your data is complete, your reviews are verifiable, and your checkout is frictionless. In three seconds, it's made its decision—and moved on."

The implications are significant:

• Your product data becomes your primary marketing asset. Specs, pricing, availability, and policies must be machine-readable, accurate, and comprehensive.
• Trust must be verifiable, not claimed. Certifications, transparent sourcing, and third-party validations matter more than brand campaigns.
• Your digital presence must extend beyond your website. AI assistants learn from Reddit, review sites, and forums. Being present where AI learns is now essential.

But agent-readiness is just the starting point. The logical endpoint of agentic commerce is predictive fulfilment—where your customer's AI knows they're low on coffee, that their running shoes have hit their mileage limit, that rain is forecast for their commute. It doesn't wait for them to remember: it orders, optimizes delivery, and puts replenishment on autopilot.

Convenience becomes invisible. Shopping fades into the background of life.

For brands, this means designing towards autonomous transactions: APIs that allow agents to negotiate, subscribe, and auto-replenish. The winners will capture loyalty at the algorithm level.

Yet Tomorrow's Commerce reveals a counterintuitive insight: as AI proliferates, human curation becomes more valuable, not less.

When every feed is saturated with infinite options and AI-generated content, the differentiator becomes a human lens—taste, empathy, lived experience layered over machine scale. "Human-reviewed" labels, expert-curated bundles, and human arbitration for sensitive moments are emerging as powerful trust signals.

"The brands that will thrive aren't choosing between AI efficiency and human connection," notes Appelhof. "They're finding the moments where each adds the most value."

Four moves to make now

1. Audit your agent-readiness: Test your products on ChatGPT, Perplexity, and Google's AI Overview. Are you found? Correctly represented? Recommended? Identify gaps before your competitors do.
2. Make your catalog agent-native: Structure product data in machine-readable formats. Treat AI agents as customers: deliver agent-friendly APIs, deep links, and seamless handoff to checkout.
3. Build presence where AI learns: Engage authentically on Reddit, review sites, and forums. Publish authoritative answers that assistants reference and test how your brand is described today. You might be surprised.
4. Define your human-in-the-loop moments: Identify where human judgment changes outcomes: assortment decisions, sensitive categories, dispute resolution—and embed human oversight there first.