What Is Criteo? Products, How It Works, and Setup Requirements
Published:
Last Updated:
Category: Ad Operations
Authors: Shusaku Yosa
Criteo is a commerce media company whose services use shopping and product-interest data. Although associated with dynamic retargeting, its offering also covers customer acquisition, retail media and monetization for retailers and media owners.
Start by identifying your role: are you buying advertising or monetizing inventory? Then define the business outcome, available data and people responsible for maintaining the setup.
Which Criteo product serves which purpose?
Product | Main purpose |
|---|---|
Commerce Growth | Customer acquisition and retention for advertisers and agencies |
Commerce Max | Buying commerce media for brands and other advertisers |
Commerce Grid | Supply-side inventory infrastructure |
Commerce Yield | Advertising monetization for retailers and marketplaces |
These broad roles follow Criteo’s platform overview. Inventory, interfaces and commercial terms vary by product and contract. The rest of this guide focuses on an advertiser selling products through its own website.
Understand the feed-to-event relationship
A dynamic product campaign needs reliable product information and a way to associate relevant site activity with the correct products. A tag firing is not sufficient evidence that this relationship works. A mismatch between variant IDs and parent-product IDs, for example, may prevent the intended association.
Check | What to reconcile | Typical failure |
|---|---|---|
Product IDs | Feed IDs and event IDs | Variants mapped to the wrong parent |
Price and currency | Feed, site and update time | Expired sale prices |
Availability | Stock and discontinued items | Advertising an unavailable product |
URLs and images | Reachable, matching product pages | Redirecting to an unrelated category |
Purchase events | Order ID, value and cancellation rules | Duplicate purchase on page reload |
Consent and exclusions | Applicable purpose and audience restrictions | Suppressed records re-entering elsewhere |
This is an implementation review, not a universal feed specification. Use the requirements for your selected product when defining fields and event payloads.
Define the launch conditions
Choose one initial objective: acquiring new buyers, encouraging repeat purchases or helping recent visitors return. Agree on revenue, order and margin definitions, including returns. Assign an owner to the feed, site measurement and consent workflow.
Budget separately for media, creative, campaign management and data preparation. Do not treat an old published minimum spend or traffic threshold as a current eligibility rule. Confirm the relevant product, market and contract with Criteo. Likewise, there is no universal number of days after which learning is guaranteed to be complete.
Before launch, run a permitted test order and verify product identity, value, duplicate prevention and exclusions. Name the person who can pause delivery if data becomes unreliable.
Reported ROAS is not incremental revenue
For a fictional example, spending JPY 200,000 and receiving JPY 800,000 of platform-attributed revenue produces a reported ROAS of 400%. The yen amounts illustrate a Japan-based campaign; they are not a recommended budget.
That revenue is not necessarily the amount caused by advertising. Some customers may already have intended to buy or may also have interacted with other channels.
Evidence | What it supports | Remaining question |
|---|---|---|
Platform ROAS | Results under that platform’s attribution rules | Click/view windows and overlap |
First-party orders and margin | Overall commercial performance | Returns, costs and customer mix |
Controlled comparison | An estimate of incremental impact | Design, bias and sufficient observation |
Avoid adding conversions reported by different platforms. Reconcile what you can using first-party order records and document what cannot be reconciled. Differences from GA4 can also reflect attribution and coverage, not only implementation errors.
Review operations before optimizing outcomes
Early reviews should prioritize feed errors, ID consistency, destinations, duplicate events and stock updates. Once measurement is stable, examine product groups, customer acquisition versus repeat purchasing and margin after costs.
Keep a change log containing the date, products, budget, expected effect and review date. A case-study result is not a forecast for your account. The useful decision is whether your own evidence supports continuing, changing or stopping the campaign.
Related guides
- Native Advertising: Formats, Disclosure, and Measurement
- How to Read PageSpeed Insights: Scores, Field Data, and Fix Priorities
Review your results consistently
For ongoing data aggregation, comparison and reporting, review NeX-Ray’s integrations and features. Check the plan requirements for your analysis, then start for free.
Sources and verification
References checked on October 1, 2026. The examples and worksheets in this article are editorial illustrations, not reported customer results.
