Industry Perspectives | AI Should Simplify Advertising, Not Automate Its Complexity

As AI adoption accelerates across the advertising ecosystem, much of the focus has been on automation—optimizing bids, streamlining execution, and improving efficiency at scale.

Yet for many brands, a more fundamental issue remains unresolved. Despite having more data and tools than ever, it is becoming harder to answer a basic question: are we actually reaching the right audiences?

In this edition of Industry Perspectives, Nicolas Bidon, CEO at Ogury, explains why the industry is focusing on the wrong layer of the problem and how AI can create real value.

Why is it still so difficult for brands to know if they are reaching the right audiences?

Because the ecosystem is fragmented by design. Digital advertising today generates a wide range of signals, such as behavioral, contextual, transactional, or survey-based data. Each captures only a partial view of the user and is often unstable or difficult to interpret on its own.

What marketers are left with is not a lack of data, but a lack of consistency. Different platforms, datasets, and methodologies produce different (sometimes conflicting) views of the same audience. As a result, defining who to reach becomes increasingly uncertain, even as the volume of available data continues to grow.

What are brands still getting wrong when defining their audiences?

Many still prioritize precision over relevance. The assumption is that more granular data leads to better outcomes. In practice, this often results in over-segmentation and audience definitions that are too narrow or overly dependent on weak signals.

At the same time, many strategies still rely on proxies such as device IDs or inferred behaviors that no longer provide a stable foundation. Media buying has become highly optimized from an execution standpoint, but performance still depends on audience definition quality. If that definition is flawed, optimization will only amplify the problem.

How should brands approach signals in this new environment?

Signals are still essential, but they are only a starting point. On their own, they are incomplete, volatile, and often misleading. Their value comes from how they are combined and interpreted. AI enables a different model: instead of reacting to isolated signals, it can aggregate multiple data sources, identify stable behavioral patterns, and structure them into coherent audience models.

Persona-based approaches don’t remove signals; they turn scattered inputs into a stable foundation for audience planning and activation. It allows brands to move from micro-signals to a more consistent and scalable understanding of who they are trying to reach.

Where does AI create the most value today?

Most current applications focus on execution: optimizing bids, refining targeting, and automating campaign management. These use cases deliver efficiency, but they operate on top of existing complexity rather than resolving the fragmentation behind it. 

The greater opportunity lies upstream. AI helps extract and structure fragmented data, model personas, and clarify which audiences truly matter for a brand. It can guide strategic decisions before activation even begins, ensuring that execution is based on stronger foundations.

This marks a shift from operational optimization to decision intelligence. When AI improves audience understanding, execution becomes naturally simpler and more effective.

How should brands think about AI in an increasingly complex ecosystem?

A common misconception is that AI will simplify advertising by replacing the current stack of platforms and tools. In reality, the ecosystem is unlikely to become simpler at the infrastructure level. What can change is how these systems are connected and interpreted. For AI to deliver real value, it requires a consistent layer of interpretation—one that translates raw data into shared definitions usable across platforms.

The objective is not to multiply AI tools, but to improve how existing systems operate together. Interoperable architectures make that possible by allowing AI to enhance the ecosystem rather than add further complexity.

This is where Persona Intelligence becomes critical. By structuring and interpreting signals within a common framework, this technology enables consistent audience definitions across planning and activation, while integrating smoothly into existing technology stacks. With that foundation in place, intelligence becomes more portable, decisions more consistent, and activation simpler.

What does this shift mean for brands and agencies in practical terms?

It changes the point of leverage. Over the past decade, the industry has focused on improving execution and delivered significant gains in efficiency. The next phase is about improving the quality of inputs: how audiences are defined, prioritized, and understood.

Brands and agencies that invest in this layer will be better positioned to:

• build more consistent strategies across channels;
• reduce reliance on unstable signals;
• and focus investment on audiences that actually drive outcomes.

In this context, automation still matters, but it is no longer the primary differentiator.

What is the key takeaway for advertisers navigating this shift?

AI will not redefine advertising by making bidding faster or targeting more granular. Its real impact will come from its ability to transform fragmented signals into clear, actionable audience personas that can be activated consistently at scale.

Because in the end, performance doesn’t come from how efficiently you buy media. It comes from how well you understand who actually matters for your brand.