AI agents are the new front door of commerce: How to prepare for Agentic Commerce?
- alissahilbertz
- 21 hours ago
- 4 min read
We are on the brink of a new era in digital trade. Increasingly, autonomous AI agents, not humans, are deciding which products are visible, which alternatives are considered, and which purchases are ultimately executed. The traditional e-commerce funnel, which was built on traffic, clicks and conversion rate optimization (CRO), is being replaced by something radically different: an agent-driven, hyper-personalized journey.
From funnel-thinking to agent-thinking
AI is the new driver of discovery. With the rise of ChatGPT, Google’s Gemini, Perplexity and specialized generative AI platforms, consumers are increasingly outsourcing the "legwork" of searching and comparing to digital agents. These agents don't just know a user's preferences and history; they have real-time access to a near-limitless global marketplace.
For years, retailers focused on Search Engine Optimization (SEO), Search Engine Advertising (SEA) and on-site conversion. With LLMs, the buying journey has become more layered. Now, discoverability is shifting to a new domain: AI Agent Relevance.
What is Agentic Commerce, really?
Agentic Commerce goes beyond basic chatbots or recommendation algorithms. It centers on autonomous AI agents capable of independent reasoning, decision-making and executing actions on behalf of the user.
The scenario: Instead of visiting ten webshops and manually clicking through filters for a new backpack, you give your agent a prompt:
"I need a waterproof backpack under €80, that fits a 15-inch laptop, with a minimalist professional look."
The agent then:
Scours the entire online market (not just the top Google results)
Compares technical specs, durability reviews, and real-time availability
Weighs your personal style history
Executes the purchase autonomously (if authorized)
The interaction shifts from "scrolling and clicking" to a single high-intent conversation. Instead of the webshop, the agent orchestrates the entire journey.
Product data: the new marketing channel
This shift exposes a fundamental weakness in many current retail and Product Information Management (PIM) landscapes: product data is often available, but rarely agent-ready.
Traditional marketing focuses on brand image and emotional positioning for humans. Agentic Commerce focuses on machine-readability. How well does an AI agent understand your product?
Basic attributes like dimensions, color and material are now the bare minimum. To win, agents need contextual data:
For whom is this product ideal?
In which specific scenarios does it outperform the competition?
What is the "reason to buy" that a reasoning engine can verify?
Without rich, distinctive and consistent product information, your brand simply won't make the agent's shortlist.
The impact of Agentic Commerce on retailers
1. From SEO to AX (Agent Experience)
Gartner predicts that by 2030, a significant portion of e-commerce transactions will be handled by GenAI platforms. If your data isn't structured for LLMs, your product effectively doesn't exist.
2. High precision, low returns
AI agents match products with intent using surgical precision. This leads to higher customer satisfaction and a radical reduction in return rates due to product mismatches, provided your data is accurate. The opposite also holds true: if the agent is "lied to" by poor data, the brand penalty will be severe. Imagine poor data causes an AI agent to confidently recommend the wrong running shoes. Your frustration, product return, loss of trust and the AI agent’s algorithmic feedback loop ultimately create a brand penalty far greater than a traditional mis-sale that went unrecorded.
3. The decline of the traditional funnel and rise of the API
Expect fewer pageviews and fewer direct clicks. AI agents compare specifications, price, delivery times and policies directly via Application Programming Interfaces (APIs). Therefore, APIs, rather than pretty landing pages, will become your primary commercial channel.
4. New growth opportunities
Agentic Commerce levels the playing field for:
Niche assortments that were previously too hard to find, example: “Find me a vegan, chemical-free leather repair kit that ships within two days.”
Dynamic bundling is based on the user's specific context, for example: “Prepare everything I need for a three-day ski trip under €400”. The agent then bundles thermal base layers, gloves, a helmet renting option and travel-sized sunscreen.
Specialized advisory agents that act as automated personal shoppers. Mirroring the role of a high-end retail advisor, the agent can source clothing that fits your size profile, filter on ethical sourcing, evaluate material quality and color matching while also recommending outfits for upcoming events.
How to prepare for Agentic Commerce: Your three-step roadmap
Transform data into "agent-centric content"
Move beyond technical specs. Enrich your PIM with "lifestyle attributes" and "use-case scenarios." Ensure your data is not just "clean," but "descriptive" enough for an AI to reason with. While size, cushioning and whether a shoe is waterproof is relevant, so is “best suited for runners doing 10-30 km per week on asphalt” or “ideal for warm climate due to breathable mesh”.
Adopt an API-first architecture
Agents make decisions in milliseconds and do not tolerate uncertainty. If your API cannot confirm stock levels instantly, you will be skipped in favor of merchants whose data is verifiable and up to date. Although the universal protocol for agent-to-business communication is still evolving, your catalog, pricing and inventory must be accessible in real-time via machine-readable endpoints.
Launch agentic pilots
Don't wait for a "Big Bang." Start building small-scale agents for specific use cases, like a specialized "Gift Finder" or a "Technical Support Agent." This will highlight the gaps in your current data landscape before the market shifts entirely.
The direction is clear: AI agents are the new "front door" of commerce. And it is important to note that this is not a distant outlook. Various sources state that between 25-30% of consumers are already comfortable with agents shopping for them. Retailers who invest in rich product data and optimize their tech stack for autonomous agents today will be the ones who remain visible tomorrow.
Authors: Bert Kwanten and Wouter Verstege
Curious about Agentic Commerce? Get in touch
Bert Kwanten
Senior Manager Retail & Products
T: +31 638602329