How AI is changing marketing and why measurement needs a new role
Article 17 of 18 in series: The Great Reach Reset
AI is changing marketing measurement from a reporting function into a decision-support system. As AI increasingly helps marketers analyse, optimise and execute campaigns, measurement must provide reliable and comparable signals that help both people and machines make better media decisions.
For decades, we have designed measurement for humans.
We built dashboards to be read, reports to be presented and KPIs to explain what happened, because the purpose of measurement was to help marketers understand the past before deciding what to do next. Every major innovation, from impressions and reach to attention and business outcomes, has made us progressively better at answering that question.
That made perfect sense because the decision-maker was always a person. Today, however, that assumption is beginning to change.
AI is moving beyond content creation and campaign execution to become part of the decision-making process itself. AI can already analyse audiences, recommend budget allocations, optimise media plans and increasingly execute campaigns on our behalf. Whether that future fully arrives in two years or ten is almost beside the point. What matters is that measurement is no longer serving only marketers. Increasingly, it will also serve the systems making decisions alongside them.
Recently, Paula Despins, VP of Measurement at Amazon Ads, captured this shift when she observed that senior marketers are asking how quickly they can move from static reporting to systems that recommend actions and optimise campaigns in real time: Agentic AI isn’t a future-state conversation anymore.
I think Paula’s points to something much bigger than the adoption of AI. It signals a fundamental shift in what we expect measurement to do.
From reporting to AI-powered decision support
For decades, the purpose of measurement has been to explain performance. We have developed increasingly sophisticated ways of understanding campaign effectiveness, helping marketers interpret results and identify opportunities for improvement. That remains essential, but it is no longer sufficient when decisions are increasingly supported by AI.
Unlike humans, AI does not consume reports. It cannot combine a dashboard with intuition, commercial experience and context before deciding what to do next. Instead, it depends on structured, comparable and trustworthy signals that can guide its recommendations. The value of measurement is therefore no longer defined solely by its ability to explain yesterday’s performance. Increasingly, it will also be judged by its ability to improve tomorrow’s decisions.
That is a subtle shift, but an important one, because it means measurement is evolving from a reporting discipline towards a decision-support discipline.
Which marketing measurement signals matter most for AI?
This evolution also changes which measurements become strategically important as reach, frequency, attention and business outcomes all remain essential because they answer different questions about marketing effectiveness. However, when the objective shifts from reporting performance to supporting better decisions, some signals become considerably more valuable than others.
Incremental reach is one example: A traditional campaign report might celebrate another million impressions delivered. An AI optimising for business growth is likely to ask a different question altogether: Did those impressions introduce the campaign to new potential buyers, or were they largely delivered to people who had already been reached?
That distinction has always mattered, but it becomes significantly more important when media investments can be adjusted continuously. Growth rarely comes from repeatedly reaching the same people. More often, it comes from expanding the number of future buyers who know, remember and consider the brand.
The same logic applies to outcomes. Optimising towards clicks when the objective is sales creates one type of decision. Optimising towards attention without understanding whether that attention contributes to future business growth creates another. The real opportunity lies in connecting audiences, cross-media exposure and business outcomes, enabling optimisation decisions to be based on a complete understanding of marketing effectiveness rather than isolated platform metrics.
A different role for marketing measurement
We often talk about AI as the force that will transform marketing. I increasingly believe it will also transform what we expect from measurement.
For decades, the primary role of measurement has been to help marketers understand the past. That will not change, because understanding what happened will always matter. But as AI becomes part of the decision-making process, measurement will also be expected to provide the signals that help determine what happens next.
That does not make traditional metrics like reach, attention or outcomes less relevant. If anything, it increases their value. The challenge is no longer simply to measure them accurately, but to connect them in ways that enable better decisions across channels, and that is where our industry is heading.
For years, we have judged measurement by the quality of the reports it produces. In the years ahead, we will increasingly judge it by something different:
Does it help us make better decisions?
This article is 17 of 18 in our series: The Great Reach Reset
About this article series
In a fragmented media landscape, reach is no longer a simple KPI - it is a strategic growth lever. In this article series, AudienceProject explores why advertisers are not failing at reach, but at measuring it properly, and why incremental, deduplicated cross-media reach has become essential to driving penetration, controlling frequency, and unlocking sustainable growth.
You might also like:
How brands grow in the era of AI
Marketing mix modelling and cross-media measurement: Why marketers need both
Why omnichannel marketing isn’t about being everywhere
Why is advertising harder to measure and what can marketers do about it?
Naughty, naughty marketers? Or…
189 years later, advertising still comes down to trust and incremental reach
Signup for the AudienceProject newsletter
Industry relevant articles, truth-revealing studies, exiting new product launches and random thoughts from brilliant people - directly in your inbox. Free of charge and no spam guaranteed!