Strategy does not win. Execution does

And yet outcomes remain volatile.

Author: Carni Wilkens, chief growth officer at dentsu

Campaigns that look right in planning fluctuate once they are live. Performance varies from one burst to the next. When results improve, it is difficult to explain why. When they fall, it is even harder to understand what changed.

This is not because strategy has failed us.

It is because we are executing strategy inside algorithmic ecosystems that are not designed to reward advantage.

The reality we are working in

Modern media execution happens inside complex, platform driven systems. Platforms optimise to their own objectives. Signals are partial, delayed, and opaque. The same strategies, audiences and tactics are available to everyone and quickly become commoditised.

In this environment, having the right strategy is no longer a reliable source of advantage. Execution quality has become central.

This shift means execution can no longer be treated as an operational layer.

It requires an execution system intentionally designed to preserve strategic intent, enable learning, and deliver repeatable outcomes inside algorithmic environments.

That belief underpins how we have built our approach at dentsu.

This dynamic is particularly visible across African markets, where fast adoption, fragmented environments and diverse consumer behaviour coexist. Complexity is not the exception. It is the norm.

The real performance challenge is unpredictability

Marketing performance today does not fail because teams are inefficient.

It fails because outcomes are unpredictable.

Brands increasingly struggle to answer some fundamental questions with confidence:

• If we invest more, will performance scale?
  
• If results drop, what caused it?
  
• What is genuinely working versus what the platform is optimising?
  
• What outcomes can we expect next time, not just what happened last time?

Reporting tells us what happened. Predictability tells us what we can expect to happen again.

Measurement without learning does not create control.

At dentsu, the execution system designed to close this gap is Curator. Not by limiting activity, but by creating execution conditions that enable learning, consistent measurement, and more repeatable outcomes inside algorithmic systems.

Where strategies lose their power

Human behaviour has not suddenly become more rational. People still make fast, intuitive decisions shaped by emotion, experience, and context.

Where things break down is after strategy.

Audiences are translated inconsistently across platforms.

Activation decisions fragment.

Algorithms reinterpret intent.

Measurement frameworks differ by channel. By the time activity is live, the original strategic intent is often diluted.

The result is a familiar frustration across our industry:
“The strategy was right, but the results don’t reflect it.”

Curator exists to ensure that what is planned is what is executed as closely as the ecosystem allows.

One data spine, used consistently

Curator is not a buying model.

It is not a media shortcut.

And it is not primarily built to cut waste.

Curator is an end-to-end decisioning and activation ecosystem built around a single shared data spine.

That spine is the same foundation used by our strategists to develop audience and communication strategies. It is connected to over 140 million verified IDs across sub-Saharan Africa. And it enables consistent audience definition, planning, activation, and learning.

Strategy is activated against real people, across fragmented media environments, without reinterpreting the audience each time.

That consistency is the core differentiator.

Why this matters in an algorithmic era

We are operating in an environment defined by platform complexity, AI driven optimisation and diminishing transparency.

Advantage does not come from having more data. It comes from how coherently data is used and whether learning can be retained across time, platforms, and campaigns.

Curator is designed to operate with algorithmic systems rather than ignore them, while preserving strategic intent and enabling measurement that informs future decisions.

Curating for value, not restriction

Yes, Curator often results in less waste, fewer low-quality impressions, and more controlled environments. But these are outcomes, not objectives.

The objective is to unlock incremental value inside platform ecosystems: audiences that are strategically valuable but underleveraged, environments where performance can be stabilised, signals that improve learning rather than optimisation alone, and execution patterns that can be repeated with confidence.

This is how measurability and predictability improve over time not just efficiency in the moment.

One logic, multiple ecosystems

Every media environment has its own constraints. Algorithms behave differently. Inventory dynamics vary. Measurement is not applied equally. Attention shifts by context.

Curator does not force a single execution approach.

It applies one decisioning logic, grounded in a single data spine, across multiple ecosystems.

Across social platforms, it enables consistent activation while navigating optimisation systems and signal loss. In premium broadcast and video environments, it brings audience led decisioning into fragmented viewing landscapes. In inventory led environments, it curates’ brand safe, performance appropriate supply while preserving learning and control.

Different environments. Different constraints. One strategic execution system.

Who Curator is designed for

Curator is for brands that care about outcomes, not just activity. For teams that need performance they can explain and repeat. For organisations that want strategy to survive contact with platforms. And for those that value learning over isolated optimisation.

It is not designed for pure lowest cost buying, short term tactical bursts without learning intent, or situations where execution consistency does not matter.

That clarity matters. Over promising is the fastest way to undermine predictability.

Why we built it this way

Curator works because it is embedded inside a single data spine, a shared strategic methodology and an operating model designed to connect planning, activation, and learning.

It reflects dentsu’s broader focus on inclusive intelligence integrating data, technology, and human understanding into coherent systems rather than disconnected tools.

If these challenges reflect what you are experiencing; volatility, opaque performance and diminishing explanation, I will value the opportunity to compare perspectives.