Turn intent into outcomes with audiences built on real commerce data

Commerce Audiences are built from consented deterministic and/or probabilistic signals observed across Criteo’s network of 17,000+ ecommerce sites, 200+ retailers, and thousands of publisher sites. We shape those signals into two attribute types:

• Product attributes. Categories, brands, and price quantiles.
• User attributes. Country, gender, buying power, and content consumption.

Signals come from real activity like searches, product views, purchases, content views, and ad interactions.

It depends on the segment and how shoppers behave. Some audiences are defined by one strong action (like a specific product search or cart activity). Others blend multiple signals (search + views + purchases + content + ad engagement). In short: each audience uses as many data points as needed to reflect real intent, not a fixed number.

Commerce Audiences are built with machine learning that turns consented commerce signals into dynamic audience segments. Our models analyze patterns in shopper behavior, then score and rank users based on how likely they are to take specific actions, like engaging, converting, or buying.

These audiences aren’t static lists. They’re automatically generated and refreshed twice per day (every 12 hours) based on the latest user interactions, so your targeting stays aligned with real, current shopping intent.

Pricing depends on the audience type and the solution where you activate it. Here’s the general guidance from the table:

• Commerce Growth: Criteo, Advertiser, and Custom Audiences: No additional fees. Retailer Audiences: n/a.
• Commerce Grid: Criteo and Custom Audiences: Flat CPM. Retailer Audiences: Flat CPM or share of media cost (varies by retailer). Advertiser Audiences: No additional fees.
• Commerce Max: Retailer Audiences: Flat CPM. Other audience types: n/a.

For exact rates and retailer-specific terms, contact the relevant product marketing team for each Commerce Media Platform solution.