The French online commerce sector surpassed 196 billion euros in revenue in 2025 according to Fevad, with an increase of about 7% compared to the previous year. Behind this growth lies a structural change: the average basket size is declining while the frequency of purchases is increasing. This dynamic alters the profitability levers for merchants and necessitates rethinking several technical components of the purchasing journey.
Decline in average basket size and increase in transactions: what this means for profitability
The growth of e-commerce no longer relies on high-value orders. Buyers are placing orders more frequently, but for lower unit amounts. According to Fevad data cited by several analysts, the average basket size has decreased by about 3% during the 2024-2025 period.
For a merchant, each order generates fixed costs: preparation, packaging, delivery, and returns management. When the basket size decreases, these costs weigh proportionally heavier. Two strategies can help offset this phenomenon.
- Group orders by free shipping thresholds calibrated to the actual average basket size of the store, not based on an arbitrary amount copied from a competitor
- Reduce the unit preparation cost by automating picking or by switching to a logistics provider whose pricing model is volume-based and decreasing
- Work on automatic replenishment (subscriptions, recurring baskets) to transform purchase frequency into predictable flows rather than random spikes
Following the news from Guide Info Commerce allows you to identify logistical solutions and management tools suited to this structural evolution of the market.
Agentic commerce: when AI buys on behalf of the customer
The term agentic commerce refers to a model in which an artificial intelligence agent executes all or part of the purchasing journey on behalf of a consumer. The agent searches, compares offers, selects a product, and can even finalize the transaction.
According to a KPMG report published in 2026, nearly one in three online shoppers is already using generative AI in their purchasing process (search, comparison, decision support). This figure marks a shift: the merchant site is no longer the only entry point in the journey.
Adapting your catalog to AI agent queries
An AI agent does not navigate like a human. It relies on structured data from product sheets: price, technical characteristics, availability, delivery time. If this information is missing or poorly formatted, the product becomes invisible to the agent.
In practical terms, this means that schema.org Product and Offer tags become as strategic as traditional SEO. A merchant who correctly fills in the “sku,” “availability,” “deliveryLeadTime,” and “returnPolicy” fields in their structured data increases their chances of appearing in the responses generated by these agents.
Agentic commerce does not replace natural referencing. It adds a layer of machine reading that rewards the quality of metadata.
Installment payments and alternative payment methods on online stores
Installment payments (Buy Now Pay Later) have established themselves as a measurable conversion lever. In a context of declining average basket size, offering three or four interest-free installments can be enough to turn a sale, especially for baskets between 80 and 300 euros.
Several recent payment solutions also target checkout fluidity. Wero, the instant payment system backed by European banks, is starting to appear on some online stores. Its main advantage: an almost instant transfer without sharing credit card data.
What this implies for technical integration
Adding a payment method is not limited to activating a module. Each additional option at checkout increases page load time and can create confusion if the interface is not clear. Limiting choices to three or four visible payment options remains a good practice, even if it means displaying alternatives in a secondary dropdown menu.
The cart abandonment rate increases with each additional second of loading time. Before adding a new provider, an A/B test on the payment page allows you to measure the real impact on conversion.
SEO and web traffic: the impact of AI-generated results on merchant sites
Search engines are increasingly integrating AI-generated answers directly into the results page. For an e-commerce site, this means that organic traffic can decline even if positions remain stable.
When Google or a competitor displays an AI summary that answers the query “best wireless headphones 2026” with a comparison directly in the SERP, the click to the merchant site becomes optional. The merchant loses the visitor before they even see the product sheet.
Two strategies to maintain traffic
The first is to target long-tail transactional queries rather than informational queries. A page optimized for “buy Sony WH-1000XM6 delivery 24h” captures a visitor further along in their journey, less likely to settle for an AI summary.
The second involves enriched content directly on product sheets: detailed customer reviews, demonstration videos, size or compatibility guides. This type of content offers value that AI summaries do not yet replicate accurately, and it generates engagement signals that strengthen positioning.
The proliferation of AI agents and generated responses in search results redistributes the cards of merchant web traffic. Stores that accurately structure their product data, optimize their checkout without overloading it, and adapt their SEO strategy to transactional queries have the technical foundations to absorb these changes without suffering a silent erosion of their sales.
