A customer types a sentence into a chatbot: "I'm looking for an 18-karat gold wedding band, size 7, delivered before the 15th."
A few seconds later, they have an answer, with a specific product, a price and availability. They haven't visited a single website, compared any product pages or opened any tabs. The AI did all of that for them.
This scenario has a name: agentic commerce.
It is quietly reshaping the way customers find and buy products, including in jewelry and watch retail.
And it is not a fringe phenomenon. According to a Fevad/Odoxa study, 31% of French online shoppers already use generative AI when shopping online, a habit that barely existed three years ago. An iAdvize x Ifop study goes further: among generative AI users, 62% consult it before a purchase to research, compare offers or get a recommendation. The movement is underway, and it is accelerating faster than most of the technology shifts that came before it.
What exactly is agentic commerce?
Agentic commerce refers to a buying model in which an AI agent acts on the customer's behalf. It isn't a chatbot that answers questions, or a recommendation engine that suggests similar products.
From a simple request, an AI agent can:
- Search for products that match the request.
- Compare several offers.
- Check inventory and the product's actual availability.
- Suggest the option that best fits the customer.
What a customer does today on a search engine or comparison site, an AI can do for them, faster and with better targeting.
This relies on new technical protocols: OpenAI's Agentic Commerce Protocol, Google's Universal Commerce Protocol and the Agent Payments Protocol, which governs payments triggered by an AI.
In other words, the infrastructure of online retail is already adapting.
There is also a major difference with a traditional search engine: the AI remembers previous conversations. It already knows the customer prefers white gold to yellow gold, has a rough idea of their budget, or knows they usually buy gifts rather than for themselves.
The recommendation is no longer based only on the current request; it draws on a whole conversation history. That is what makes the suggestion truly personal, and much closer to what a good salesperson who knows their customers would do.
How does it work?
Let's go back to the 18-karat gold wedding band. The customer doesn't just describe what they want; they express a complete need: material, size, delivery deadline... The AI then analyzes several catalogs, compares available products against these precise criteria, and can guide the customer all the way to purchase.
The journey becomes much more direct. It's no longer about guiding a visitor through a website, but about answering an AI that queries, filters and partly decides on the customer's behalf.
A changing buying journey
Until now, the buying journey was a series of steps customers handled themselves: they search, filter, compare, check and choose. With agentic commerce, some of these steps can be handed over to AI.
The customer no longer searches; they express a need, and the agent finds the answer for them. This changes the very nature of the relationship between retailers and their customers: much of the decision now happens upstream, in a conversation the retailer never sees.
A new intermediary between customer and retailer
With agentic commerce, AI becomes a new gateway to a retailer's products. It can explore several offers, identify relevant products, compare them against the customer's request, and then present the best answers.
In practice, this means one thing: being visible is no longer enough.
For an AI agent to recommend a product, it has to be able to understand it, compare it and verify it. Products must be accessible, understandable and comparable by AI agents, not just presentable to the human eye.
And that starts with something familiar: solid SEO.
A poorly structured website, thin product pages or unclear content remain hard to use, whether for a search engine or an AI agent. SEO doesn't disappear with agentic commerce; it becomes the foundation AI relies on to understand an offer before recommending it.
The iAdvize x Ifop study also shows that 21% of consumers would already be willing to let an AI make a complete purchase for them, a figure that rises to 55% among regular AI assistant users. AI's role as an intermediary is only just beginning.
Data becomes essential
To recommend the right product, AI needs reliable, precise and up-to-date information.
For each reference, it needs to know:
- The product's characteristics.
- The current price.
- The inventory status.
- The sizes and variants available.
- Delivery times.
The more structured the data, the more relevant the agent's answer to the customer. Conversely, an incomplete product page or outdated inventory simply means a missed sale, without the retailer even noticing.
This data also has to be presented in a format AI can read. Text designed only for the human eye, buried in visuals or a complex layout, often remains invisible to an AI agent. Product information needs to be exposed in a structured way, through data feeds, standardized markup or AI-compatible formats, so it can be correctly interpreted and compared.
A new challenge for omnichannel retail
In jewelry as elsewhere, the same product can be available on several channels: physical stores, e-commerce websites, marketplaces.
To be usable by an AI agent, this information must stay up to date, reliable and synchronized.
The challenge is no longer just having good data somewhere. It's having consistent data, whatever channel is queried. Inventory that is correct in store but wrong online, and the AI drops the product, or worse, gives the customer an inaccurate answer.
Getting ready for tomorrow's uses
To take full advantage of these new uses, AI agents need data that is reliable, structured and continuously updated. This is where a solution like Commerce 360 really makes sense, by centralizing products, inventory, sales, orders, customers and e-commerce in one place.
By bringing business data together in a single environment, this kind of tool paves the way for new forms of commerce, driven by humans, but also, more and more, by AI.
Why not get a head start?
Agentic commerce isn't the norm yet, but it is moving fast. Centralizing your data, connecting your channels and upgrading your tools are already the first steps to anticipate these new uses rather than fall behind.
Frequently asked questions about agentic commerce
What is agentic commerce?
Agentic commerce is a model in which artificial intelligence acts as an intermediary between the customer and the retailer. It searches for products, compares offers, checks inventory and recommends the best option, based on a simple request in natural language.
Will agentic commerce replace traditional e-commerce?
No, it adds to the existing buying journey. The e-commerce website remains where the transaction takes place, but part of the research and comparison work, so far done by customers themselves, can now be handed over to an AI agent.
How can you make your products visible to AI agents?
It starts with solid SEO: a well-structured website, clear product pages and easy-to-understand content remain the foundation. Product data (price, inventory, sizes, delivery times) then needs to be exposed in a format AI can read, kept up to date and consistent across all sales channels.
Does agentic commerce also matter for small independent stores?
Yes. Store size matters less than the quality and structure of the data. A small store with reliable, up-to-date product information can be recommended by an AI agent as easily as a large chain, as long as its data is accessible and well organized.
Do you need specific technology to be ready for agentic commerce?
Not necessarily separate technology, but well-organized data. Centralizing product, inventory and sales information in a single environment, as an omnichannel management tool does, makes it much easier for AI agents to use.
Want to talk about your project? We can support you through this transition.
Back