by Dana Lee on 16th Sep 2026 in Video Interviews

In this episode of A Coffee With... Tim Conley, VP, CGrid, EMEA at Criteo, joins ExchangeWire's John Still over a cold brew to chat how AI is transforming commerce, programmatic, and measurement.

AI is reshaping the commerce landscape, driven by rapid consumer shifts towards LLM product discovery. Tim advises brands to identify effective prompts and creatives in these environments so they are ready to scale as the technology matures. Early adoption and a test-and-learn approach are crucial for brands to build skills and insights early, giving them a competitive edge in market share and expertise.

The conversation outlines Criteo's commerce intelligence layer and early partnership with ChatGPT Ads, as well as how its retail media presence helps brands to fulfil programmatic measurement. The future of media trading is envisioned as an agent-driven ecosystem where value lies in layers of measurement, verification, and intelligence.

AI Product Discovery and the Collapsed Purchase Funnel

Criteo research shows that 39% of people now use AI in their product discovery and shopping journeys. With AI Overviews, AI is now integrated directly into information retrieval without users explicitly seeking it. LLM advertising functions as a discovery channel designed to introduce unexposed consumers to a brand, bringing them into an ecosystem where they can be captured and retargeted elsewhere.

The nature of consumer queries is also changing from generic keywords to specific, conversational problems. This enables targeted, efficient discovery and collapses the traditional marketing funnel, creating a highly compressed customer journey where the brand name is secondary to solving the immediate problem.

Criteo's Commerce Intelligence Layer

Criteo became ChatGPT Ads' first partner due to its vast commerce dataset of 4 billion globally mapped product SKUs. This granularity allows Criteo to serve highly relevant ads within the LLM, matching specific product models to conversational contexts instead of generic ones. With over $1 trillion in annual online transaction data, Criteo's commerce intelligence layer lets them map out-of-vertical consumer behaviours such as what electronics a user purchases, their overall purchase power, and their travel habits. These non-endemic insights help build comprehensive profiles that enable advertisers to target audiences actively in the same purchase journey.

Criteo is also designing its tools to MCP-first, injecting its commerce insights, curated deals, and audience segment recommendations directly into agency workflows at the point of decision-making. This shift is aligned with agency buying trends, as major agencies invest heavily in building mature, proprietary agency operating systems for planning and activation.

The Future of Media Trading and Advertising

Tim predicts that media trading will consolidate within agency operating systems to activate budgets directly. In a mature agentic ecosystem, seller agents will represent publisher inventory and optimise yield within ad servers while buyer agents negotiate and source the right supply for campaigns. This automated, agent-based framework will streamline processes and reduce the tech tax.

The role of DSPs and SSPs will be creating value through measurement, verification, and intelligence. Establishing these layers of trust and mediation – and eventually, performance – will be essential to preventing the fraud and misrepresentation seen in the early days of programmatic, while enabling a new era of agentic performance trading.