The Path to Agentic Commerce: How Brands Can Win Discovery, Earn Trust, and Prepare for AI Buyers

Agentic commerce is emerging fast, even if purchases are still small today. While artificial intelligence agents only make a small minority of purchases on commerce sites, traffic generated by AI agents is growing rapidly. The shift in product discovery is already underway, and it will reshape how customers buy and brands sell.

What 'Agentic Commerce' Really is

In this context, agentic commerce refers to a buying journey where a consumer uses an AI platform or AI agent to discover products, negotiate, decide, or transact. AI platforms include providers such as OpenAI, Google, Microsoft, and Perplexity, which offer search/answers, checkout programs, and agent capabilities that can influence and enable agentic commerce.

The 4 Formats: From AI-Influenced to Fully Agent-Led

There are four formats of agentic commerce:

1. AI checkouts: customers buy directly on AI platforms via product feeds and integrated ordering.
2. AI answers (AI referrals): AI-generated answers drive click-through to merchant sites for on-site buying.
3. On-site buying with “agentic experience”: Direct-to-consumer sites blend agentic capabilities with navigational experiences for rich customer engagement.
4. Agentic buying: AI agents act as customers with minimal human involvement — expected to take five-plus years to reach meaningful adoption.

Step 1: Win product discovery on AI platforms.

To compete in AI checkouts, brands should focus on factors they can influence — e.g., product attributes, price, availability and shipping details, while other factors are harder or impossible to control, like aggregated signals and model evolution.

For AI answers, traditional search engine optimization evolves into answer engine optimization (AEO). Focus on creating “4C content”: content that is clear, comprehensive, credible, and current, with a stronger push toward natural language-optimized content, organized in short snippets and high-density information (every word must carry value).

Step 2: Enable an agentic experience to retain control of product discovery.

Customers want AI capabilities, better self-service, and access to a real human, but they’re also skeptical about AI trustworthiness and value. An effective agentic experience has three characteristics:

• Personalized dialogue that understands intent and asks probing questions.
• Problem-solving via self-service tools (e.g., product finders, configurators, FAQs/how-tos) plus human help when needed.
• Confidence through transparency such as explaining recommendations, tool limitations, and links to source information to reduce hallucination risk and build trust.

Step 3: Set up for the future — control AI traffic and implement protocols.

By 2030, 20 percent of digital commerce transactions are projected to be executed via AI platforms or AI agents. Yet AI agent behavior today skews heavily toward product search, with fast growth in account/registration activity, raising real security and customer-account integrity concerns.

A practical starting point is bot management that distinguishes good bots from bad and grants granular permissions (browse, login, transact) so you don’t accidentally block AI platforms while protecting customers. Also implement protocols (e.g., MCP) that have wide adoption and good utility.

Become the Trusted 'Agent' for Your Customer

The brands that win won’t just chase AI visibility, they’ll earn discoverability, deliver trusted brand experiences, and set up technologies and governance for autonomous agentic buying when it matures.

Sandy Shen is a vice president analyst in the Gartner for Marketing Practice who presented live on this subject at the Gartner Marketing Symposium/Xpo, June 8 - 10 in Denver.