Agentic commerce is changing how people discover and buy products. Conversations are replacing keywords. AI assistants are reducing friction. And new shopping interfaces are emerging across retailer sites, apps, and large language model (LLM) platforms like ChatGPT, Copilot, and Gemini.
Despite the hype, retailers are not becoming interchangeable endpoints in an AI-driven world. They remain the destination because they own the fundamentals of commerce: trust, curation, fulfillment, loyalty, basket-building, payments, and more. That trust is reinforced by brands, whose equity and product integrity work in tandem with retailer curation to guide confident purchase decisions.
Agentic experiences will reshape how shoppers discover products, but they will not remove retailers from the center of the transaction. Retail media’s next phase is about enabling this collaboration at scale.
Retailers are still the destination
Shopping is not a generic problem to solve. It is deeply contextual and constrained by real-world factors like availability, pricing, and delivery. Generic AI models are powerful for language, but they are not optimized for commerce.
Retailers have what LLMs do not: complete and accurate catalogs, localized inventory, real-time pricing, and a deep understanding of shopper behavior and preferences. That retailer-specific intelligence enables meaningful recommendations and reliable transactions. It is also why in emerging LLM integrations, retailers continue to control which products are recommended and in what order while the LLM presents those recommendations through a conversational interface. This is the foundation of retailer-built apps and business agent models, and it underscores why generic language intelligence alone is insufficient for shopping.
We have seen this pattern before. Comparison shopping engines, marketplaces, and social commerce added new discovery touchpoints and expanded ecommerce, but they did not eliminate retailer destinations. Retailers adapted by differentiating their experiences, while brands continued to build trust and demand within those environments. Agentic commerce follows the same trajectory. The experience evolves, and the retailer remains in control.
From keywords to conversations, built by retailers
The shift from keyword-driven search to conversational discovery is already underway. As companies like Amazon and Walmart introduce advertising within their shopping agents, retailers are rethinking how shoppers find products across their sites and apps. While adoption is still emerging, even among U.S. shoppers who regularly use agentic shopping assistants, more than 96%* continue to rely on other channels like search, brand sites, and retailer websites along the way, underscoring the need for agentic experiences that complement existing discovery. At Criteo, we are actively helping retailers build those experiences.
When we work with retailers on onsite agentic commerce, our focus is clear and deliberate:
1. Conversational shopping experiences that support advertising
Retailers are introducing shopper assistants that translate natural language questions into relevant product recommendations. We help power these experiences so advertising works naturally within conversations, using retailer logic around availability, pricing, shopper context, and business goals.
2. New, native-to-agent ad formats
Agentic commerce enables ad experiences beyond traditional onsite search. Formats like sponsored prompts allow brands to influence the questions shoppers are encouraged to ask, moving discovery upstream and creating incremental opportunities rather than simply shifting existing spend.
3. AI optimization for limited LLM real estate
Conversational interfaces surface fewer products, increasing the value of every recommendation. Using AI, we help retailers dynamically balance organic and sponsored results, manage trade-offs between customer experience and monetization, and adapt decisions over time as shopper behavior evolves, enabling incremental ad revenue without compromising the shopper experience.
Across all three, the objective is the same: help retailers deliver differentiated, trusted experiences while building sustainable economics. Basket size and retail media performance are both critical to overall profitability, particularly as retail media has become a high-margin part of a retailer’s business. Because we already power onsite discovery and sponsored products at scale, this is an exciting evolution of what retailers use today, not a reinvention.
Retailers are rightly focused on protecting consumer trust as these experiences evolve. The most effective way to monetize without compromising that trust is not fixed ad placements, but by giving retailers flexible, transparent controls that power personalized recommendations in real time. That is exactly what we are building.
Below is an early look at what a Criteo-powered conversational shopping experience could look like as retailers move toward agentic commerce.
Expanding offsite, without losing control
As agentic commerce moves beyond retailer-owned properties, one thing remains constant: the interface may be new, but the retailers remain in control. They determine which products are recommended and why, using their own catalog intelligence, availability, pricing, and shopper signals. LLMs provide the conversational interface, but retailers remain the system of record for customer relationships, product ranking, and transactions.
The opportunity ahead is to extend retail media into these offsite environments in a way that preserves retailer decisioning and supports retail economics.
We are early-stage and working with retailers that have built or are building ChatGPT apps to design ways to surface a merged set of organic and sponsored recommendations within these experiences. As agentic commerce expands offsite, our focus with retailers is on three priorities:
- Extending retailer recommendation logic into offsite agentic experiences, ensuring retailers manage which products are shown and why, even when discovery happens outside the retailer site.
- Introducing ad-supported recommendations that work economically, supporting incrementality and basket-building, rather than disrupting existing retail media performance.
- Addressing the requirements to scale responsibly, including transparency for consumers, measurement for advertisers, and the campaign management, reporting, and optimization needed to support long-term adoption.
Our role is to help retailers extend what already works onsite into offsite agentic environments. The interface may change, but performance, and accountability remain essential.
Incremental growth, not zero-sum shifts
A common misconception is that agentic commerce will cannibalize existing channels. As Criteo CEO Michael Komasinski has argued, agentic commerce is emerging, just not the way most people expect.
History supports his view. Aggregators like Google Shopping and Instagram Shops grew ecommerce by reducing friction and simplifying discovery without replacing retailer destinations. Agentic commerce follows the same pattern, unlocking online purchases that might otherwise be delayed or abandoned. By reducing friction and simplifying discovery, it has the potential to re-accelerate ecommerce penetration by creating new shopping occasions rather than replacing existing ones.
Engagement across LLM platforms should be viewed as additive rather than zero-sum. Discovery will remain fragmented because trust and choice remain critical. Consumers enjoy shopping and do not want to delegate every decision to an agent, especially for high-consideration purchases. At the same time, many low-friction purchases do not require agentic assistance and will continue to happen wherever the consumer experience is easiest. Both realities can coexist.
What this means for brands — and what comes next
As shopping becomes more conversational, brands face a familiar challenge in a new form: staying easily discoverable as channels multiply and fragment, while continuing to build trust and clearly communicate their brand promise.
Ads remain an important lever, but they are not the only one. In an agentic world, product content, attributes, and brand voice matter more than ever. AI agents need to communicate about products with the depth and nuance shoppers expect, similar to a knowledgeable in-store associate. This creates new opportunities for brands to influence how their products are represented as discovery evolves.
At Criteo, our focus is on helping retailers extend trusted, profitable commerce into agentic environments, while enabling brands to remain discoverable and drive measurable outcomes wherever shopping happens next.
*Source: Criteo Shopper Survey, September 2025. 10,170 respondents aged 18 and over across the United States, United Kingdom, Germany, France, Japan, South Korea, Australia, Singapore, and India.
Forward-Looking Statements Disclosure
This blog contains forward-looking statements within the meaning of the U.S. federal securities laws. Forward-looking statements are statements that are not historical facts and involve risks and uncertainties that could cause actual results to differ materially. Factors that might cause or contribute to such differences include, but are not limited to: failure related to our technology and our ability to innovate and respond to changes in technology, uncertainty regarding our ability to access a consistent supply of internet display advertising inventory and expand access to such inventory, disruption of current plans and operations by our planned transfer of the company’s legal domicile from France to Luxembourg via a cross-border conversion (the “Conversion”), failure to satisfy any of the conditions to and complete the Conversion, investments in new business opportunities and the timing of these investments, whether the projected benefits of acquisitions or strategic transactions, including the Conversion, materialize as expected, uncertainty regarding international operations and expansion, including related to changes in a specific country’s or region’s political or economic conditions (such as changes in or new tariffs), the impact of competition or client in-housing, uncertainty regarding legislative, regulatory or self-regulatory developments regarding data privacy matters and the impact of efforts by other participants in our industry to comply therewith, our ability to obtain and utilize certain data as a result of consumer concerns regarding data collection and sharing, as well as potential limitations in accessing data from third parties, failure to enhance our brand cost-effectively, recent growth rates not being indicative of future growth, client flexibility to increase or decrease spend, our ability to manage growth, potential fluctuations in operating results, our ability to grow our base of clients, as well as risks related to future opportunities and plans, including the uncertainty of expected future performance and results, changes in general political, economic and competitive conditions and specific market conditions; adverse changes to the marketing industry, changes in applicable laws or accounting practices, and those risks detailed from time-to-time under the caption “Risk Factors” and elsewhere in the Company’s SEC filings and reports, including the Company’s Annual Report on Form 10-K filed with the SEC on February 28, 2025, and in subsequent Quarterly Reports on Form 10-Q as well as future filings and reports by the Company. Accordingly, a forward-looking statement is neither a prediction nor a guarantee of future events or circumstances and those future events or circumstances may not occur. You should not place undue reliance on the forward-looking statements, which speak only as of the date of this release.
Except as required by law, the Company undertakes no duty or obligation to update any forward-looking statements contained in this blog as a result of new information, future events, changes in expectations or otherwise.
