2025 has brought us to the epicenter of a seismic shift in eCommerce, powered by artificial intelligence.
According to Adyen’s Retail Report, 36% of consumers have already used ChatGPT or other AI assistants for shopping - with adoption rates reaching nearly half among Gen Z and Millennials.

A recent McKinsey report forecasts that by leveraging agentic commerce, the U.S. retail market could generate $1 trillion in orchestrated revenue by 2030, and $3–5 trillion globally.
The stakes are high, and the direction is clear: AI is driving the next major evolution in commerce, following the rise of mobile shopping.

In this article, we’ll explore the difference between a traditional customer and an AI-customer, and how to make sure your products are discoverable for both.

Google defines AI agents as software systems that use artificial intelligence to pursue goals and complete tasks on behalf of users.

In eCommerce, this means that a traditional shopper looks for products for themselves, while an AI agent searches, evaluates, and decides which products to purchase or recommend to its user.

1. Intent and Behavior

A human shopper visits an online store to find a specific item or to browse for inspiration - influenced by ads, brand loyalty, or curiosity.
An AI agent, on the other hand, operates on a task-oriented basis. Let’s look at a few examples:

• "Find me a white evening dress under $100."
  A simple task - the AI agent filters and ranks products that match the criteria.
• "I need new outdoor running shoes that can be delivered in 2 days."
  A more complex scenario - it cross-checks delivery terms and user preferences, filtering out stores that don't specify this information explicitly.
• "I want to build muscle. What do I need to buy and do?"
  A full research project - the AI agent may create a multi-step plan covering nutrition, exercise, and equipment, pulling product data from multiple sources.

Depending on the task, the AI agent will show to user the best matching resources - including content, products, delivery options, and service terms - to fulfill the their request.

2. Perception

A traditional shopper experiences your website's design, visuals, and UX, influenced by your brand's aesthetics and storytelling.
An AI agent, in contrast, focuses on data quality and structure - it "sees" your store through structured text, APIs, and data feeds rather than images.
Some agents might use computer vision to analyze web-pages, but most rely on clean, machine-readable data.

Traditional eCommerce isn’t going away.
Just as online stores didn’t replace brick-and-mortar shops, and desktop users remain relevant in the mobile era - AI agents represent a new opportunity to attract customers to your webstore.
Therefore, store owners need to optimize for both humans and machines.

1. SEO Still Matters

High search rankings continue to drive traffic.
While some old tactics (like keyword stuffing) won't help AI agents, the core principles of good SEO apply to both worlds: solid structure, fast performance, relevant content, rich metadata and strong website authority.

2. Structure Your Content

The use of HTML metadata, Schema.org markup, and sitemaps makes your website’s data machine-understandable.

✅ Structured product information enables AI agents to interpret your offerings more accurately and present them in the right context.

3. Focus on Relevance and Use Cases

Provide detailed, context-rich product descriptions with real-world examples.
✅ This helps AI agents understand which scenarios your products fit - increasing the chances of being recommended.

4. Maintain Technical Health

Keep your site fast and fault-tolerant, use correct canonical URLs, and avoid broken links.
✅ These fundamentals affect both user satisfaction and AI visibility.

5. Build Trust and Transparency

AI systems increasingly prioritize trustworthy data sources.
Include clear authorship, company information, privacy policies, and verified reviews. This helps both users and AI systems assess your brand’s credibility.
✅ Bonus: Mark up reviews, ratings, and business info using Organization, LocalBusiness, or Review schemas to enhance trust signals.

6. Optimize for Conversational Queries

AI-driven search often works through natural language.
Adapt your content to match how people speak - include FAQs, question-based headers, and scenario-based examples.
✅ Example: Instead of just "Running Shoes", include content that answers "What are the best running shoes for beginners?" or "Which running shoes are good for wet weather?"

7. Enable Data Accessibility

Make it easier for AI agents to retrieve structured information.
Offer API endpoints, product feeds, or well-structured JSON-LD data that expose real-time inventory, pricing, and delivery options.
✅ This supports integrations with AI agents, shopping assistants, or comparison tools that fetch live data.

8. Stay Aligned with AI Ecosystem Integrations

Monitor developments like OpenAI's Merchant Program, Perplexity integrations, or Google's AI Overview feeds.
Early adopters of these ecosystems often gain visibility advantages.

The Agentic Commerce Protocol (ACP)

What can be better than to be recommended by AI? Being able to sell your products directly to AI.
Protocols like the Agentic Commerce Protocol aim to let AI agents communicate directly with merchants to make purchases on behalf of customers right from a webstore.

OpenAI now enables U.S. merchants using Shopify or Etsy to integrate with ChatGPT for embedded shopping and instant checkout. Merchants using other platforms can build their own integrations.

Similarly, Perplexity’s "Shop Like a Pro" feature allows major retailers to join AI-powered shopping experiences.

The Agent-to-Agent (A2A) Revolution

In April 2025, Google introduced the A2A Protocol - a standard that enables AI agents to communicate and collaborate with each other.
What can be benefits of such communication?

For example:

• If a product is out of stock, your AI agent could locate it from another retailer while still keeping your brand in the transaction loop.
• In supply chain management, it could find alternative suppliers, negotiate terms, and optimize logistics automatically.

While developing a brand-specific AI agent can be costly today, expect the rise of aggregated AI agents that represent multiple merchants collectively.
Soon, every business will face a key question:

"Should we build our own AI agent - or join an AI marketplace?"