GEO BAROMETER BY VALIUZ
- Hélène Allouard
- 5 days ago
- 2 min read
Valiuz has analyzed data from 13 retail leaders throughout 2025 with one objective: to provide a reliable portrait of the ongoing GEO revolution.
Lille & Paris, February 4, 2026 – Within the framework of a barometer that will henceforth be published semi-annually, Valiuz—a retail media network and data alliance—has examined the traffic driven by generative AI to the e-commerce sites of 13 giants: Alinea, Auchan, Boulanger, BZB, Decathlon, Electro Dépôt, Jules, Kiabi, Leroy Merlin, Norauto, Rouge Gorge, Saint-Maclou, and Tape à l’Oeil.
What is the profile of these "augmented visitors"? What is their actual weight in overall traffic? Is their conversion rate comparable to so-called traditional traffic? And ultimately, how should retailers approach these new realities in 2026? Here is the current state of play.
18x: This is the acceleration of AI-driven traffic to e-commerce sites.
Between December 2024 and December 2025, these 13 brands saw their traffic sourced via various LLMs (Large Language Models) multiplied by 18! This figure reflects a powerful reality: the GEO channel is a booming channel with rapid adoption and increasingly effective conversion capabilities.
Notably, AI-driven traffic accounted for 72% new visitors, compared to 61% for "non-AI" traffic—a sign that brands have a significant interest in optimizing their discoverability there to attract new customers.
Primarily originating from ChatGPT (over 90%), this continuously rising traffic saw a peak last August with the release of ChatGPT-5 on the French market.
3.06 minutes / visit: -17% compared to the average time spent on sites.
At this stage, it should be noted that visitors reaching e-commerce sites via LLMs do not spend as much time as others: -17% compared to the average, or 37 seconds less than those from "non-AI" traffic (3.43 minutes). The same observation applies to pages viewed per session: -25% compared to the average, with 5.2 pages viewed versus 6.09 for non-AI traffic.
However, while these figures might initially suggest more superficial browsing, that is not the case. In fact, 90% of sessions via LLM land directly on a "product description page" (compared to 60% for other sources). The navigation is, in reality, more efficient.
Furthermore, the conversion rate over the period—1.7%—is drawing closer to the "non-AI" conversion rate month after month.
Grandparents: The demographic seeing the fastest acceleration in usage.
Unsurprisingly, the largest users of LLMs include "young people living alone," "high-income earners (CSP+)," and "urban executives," who are over-represented in the panel.
Nevertheless, the data shows a very strong acceleration in adoption across all consumer categories. Leading the way: mature couples, and even more so, grandparents!
"Given the extremely rapid adoption mechanisms, we can clearly state today that generative AI will have a far greater impact on our consumption than the Internet did. Furthermore, the level of trust is high, suggesting—especially with the deployment of agentic AI—a shift in consumer loyalty from the brand to the agent. Naturally, this paradigm shift imposes new challenges on retailers across all sectors. The absolute priorities for 2026: increasing visibility, adapting product offerings, and transforming physical stores, all with customer data at the core." — Amandine Launois, Head of Data & Insights Retailer at Valiuz.
Link to receive the infographic: https://www.valiuz.com/barometre-geo
MEDIA CONTACTS | AGENCE LEON
Gaëlle LEGRIS – Fiona THOMAS
