State of Data 2026: IAB examines marketing measurement in the AI era
Artificial intelligence can speed up analysis without solving a question: what decision should the figure inform? The IAB report offers a starting point for resuming attribution, uncreativeness and marketing mix to their right place in the dialogue between advertiser and agency.
Study published on 2 February 2026. Documented recapture: the date of the facts is separate from the date of publication of this section.
Article prepared with artificial intelligence assistance. The sources and method are indicated in the text.
Clarify the decision-making and measurement methods before comparing figures produced by different tools.
The signal: what is documented
L的IAB published on 2 February 2026 a State of Data edition dedicated to the transformation of measurement through artificial intelligence. The public presentation of the report links attribution, Incrementality testing and marketing mix models. It places this work in a context of data fragmentation, signal loss and the evolution of advertising platforms. This review is based on the presentation and the document made public by the organization.
The interest of an agency is to start from the decision to make: arbitrating a budget, evaluating a creation or understanding a commercial result does not require exactly the same evidence. Artificial intelligence can help process data, without removing the limits of collection, assumptions of a model or the need for comparison.
Three questions, three tools
The allocation of credit is divided between contact points according to a model. A test of Incrementality seeks to estimate what would have happened without the action studied. Marketing mix modelling analyses relationships at a more aggregated level, with own assumptions and constraints. These approaches therefore do not produce three interchangeable versions of the same truth. The choice depends on the decision, the granularity available and the quality of the device.
The very convincing dashboard trap
A staff member can write a smooth comment from poorly reconciled data. If the expense is one week and sales to another, the quality of the text does not fix the error. Before any synthesis, it is necessary to harmonize dates, currencies, conversion definitions, refunds and market boundaries. The team must also be able to trace the comment back to the file or request that justifies it.
Practical case: compare two markets
Imagine a brand present in Tunisia and Morocco. The first market benefits from a promotion, the second from a better product availability. Simply comparing their ROAS can lead to paying the context rather than the media. A useful protocol documents these differences, defines a comparable period and formulates a testable hypothesis. No rate of progression is proposed here: This is an example of a method, not a campaign result.
What the marketing manager needs to ask
A budget recommendation must explain its data, assumptions, uncertainties and what might contradict it. The speed of production of a reporting is an indicator of operation; the improvement of the decision is considered otherwise. We recommend that a decision be kept as a test: What arbitration would the team have taken without the new analysis, and what information has actually changed its choice?
The point of vigilance
Ask each measurement provider to explain the data used, the uncertainties and what its indicator actually allows to decide.
Sources and method
Documented recapture of facts of 2026. The date of the event or source document is separate from the publication of this article. The pedagogical scenarios and recommendations are proposed by Le Fils de Pub; they do not describe a test or presence on the spot of our editorial staff. Article prepared with artificial intelligence assistance.
Continue with a practical method
Suggested routes from the story to briefing, production or measurement. These features are separate from the original publication.