IA et retail Les centres commerciaux Nouvelle mobilité
According to Salesforce, AI is set to drive $263 billion in global online holiday sales this year, representing 21% of all holiday orders.
This is not an innovation trend. This is a behavioral shift. Consumers are no longer searching — they’re delegating.They start their journey inside conversational platforms like OpenAI’s ChatGPT, Gemini or Perplexity, formulating needs the way they would speak to a trusted advisor:
“Find me a meaningful gift under $30 for someone who loves cooking, lives in a small apartment, and already owns everything.” The outcome isn’t a ranking. It’s a decision. And increasingly, a purchase.
This shift has a name: agentic commerce. The traditional journey — search, compare, decide, buy — is collapsing into a single interaction, where the AI agent acts as a true personal shopping proxy. It understands intent, arbitrates between options, selects what truly matters and, in some cases, completes the transaction on the user’s behalf.
The result? Fewer steps, less noise and maximum intent. This isn’t just convenience — it’s delegated decision-making.
And the performance impact is already measurable. According to Adobe, traffic from generative AI platforms to U.S. retail sites surged by 760% in just one month. More importantly, these visitors are 30% more likely to convert, spend more time on retail sites and generate 8% more revenue per session.
In short, AI doesn’t attract browsers. It attracts buyers.
Retailers understand what’s at stake — and they’re moving fast. Walmart, Target and Etsy are integrating their offers directly into ChatGPT, allowing consumers to search — and sometimes buy — without ever touching a traditional website. Walmart deploys Sparky, Amazon relies on Rufus, and Target doubles down on its Gift Finder. At the same time, Amazon is locking down its data, blocking external AI bots to preserve control. Different tactics, same realization: if you’re not readable by AI, you’re not reachable at all.
This acceleration is rewriting the rules of visibility. Where SEO was built on keywords, bidding wars and placement, AEO — Answer Engine Optimization — puts relevance first. AI now favors brands that explain the problems they solve rather than simply what they sell, structure content to be understood rather than ranked, and answer real-life questions instead of search queries. Brands that have made this shift are already seeing AI-driven traffic increase by up to 90%. The rule has changed: not louder, not richer — clearer.
Agentic commerce marks a powerful shift — though not a flawless one. It is still evolving, and many consumers continue to value browsing, serendipity and discovery, while others find AI recommendations too safe, too predictable, or frustratingly repetitive. Emotion, impulse and surprise remain deeply human territories — for now.
But agentic commerce is not about replacing stores, brands or human relationships. It adds a new decision layer between intent and action, much like e-commerce did 20 years ago — only faster, more intimate and more decisive. With hundreds of millions of AI users worldwide and a growing share of web traffic flowing through LLMs, the implication is clear: if AI can’t understand your value, articulate it or recommend it, you risk disappearing from the journey. And in a world where shopping starts with a question, the brands that shape the answer will shape the market.
