Digital Shift Q3 2025: Brand building in the age of AI

1. The growing automation of advertising

This Shift asks what the emergence of fully-automated ad campaigns means for the future role of marketers and agencies.

Both Meta and Google have plans to launch, or have launched, one-click solutions that promise to automate almost all elements of advertising campaigns: from content creation through to optimisation.

While it is likely that this announcement will be primarily of interest to smaller advertisers with fewer resources at their disposal, it speaks to the broader trend of an increasing level of automation in the advertising process.

It’s a trend with significant implications for marketers and agencies, including:

• Agencies shifting upstream or downstream: Traditional media buying and creative execution will become commoditized, meaning that media agencies may need to reposition themselves either upstream (offering strategy, insight, or innovation) or downstream (handling content ecosystems or experience design).
• Ethical and cultural risks increase: Fully-automated ads present potential pitfalls with quality control, misinformation, and loss of cultural nuance if not vetted carefully. Marketers will need to take responsibility for auditing AI outputs and setting ethical boundaries.
• The role of the marketer evolves: Marketers could well spend less time managing campaigns and more time training AI systems, defining brand guardrails, and curating prompts to guide AI outputs.

In a mostly-automated landscape, what will that mean for creative differentiation? If AI-generated creative becomes prevalent, organisations will need to invest more in brand identity, storytelling, and upstream strategy to stand out in a sea of algorithmic content.

2. Context required

GenAI relies on context to shape its outputs. This has led to the emergence of a new capability that has been dubbed ‘context engineering’.

This is the deliberate shaping of the environment in which AI systems operate to guide outcomes in a business’s favour. This means designing the inputs, constraints, and situational cues that steer AI-generated outputs toward brand-aligned, effective results.In a world where marketers may no longer write the ad or choose the audience directly, their influence could shift to setting the right prompts, defining brand parameters, curating data, and shaping the ‘frame’ within which AI works.

It will be increasingly important for marketers to ensure that the AI can interpret the business objective, cultural tone, brand voice, and audience context well enough to generate accurate and effective outputs.

This could require marketers to develop new skills in narrative framing, strategic data use, and systems thinking. For brands, the onus will shift to clearer articulation of identity, values, and market positioning. Engineering context could become a powerful new differentiator as people increasingly rely on AI-generated outputs for work and information retrieval.

3. Avoiding ‘Bottom-up trap, top-down fantasy’ for AI integration

According to IBM’s recent 2025 CEO study 60% of CEOs find themselves still in the pilot phase with AI, despite two-thirds having expected to have been out of the proof-of-concept stage by this year. In addition to this, over the past three years, CEOs say only 25% of AI initiatives have delivered expected ROI.

As many organisations learned from digital transformation, large-scale change takes time, and so while this presents a cause for concern, a way forward may be possible.

One of the key challenges for leaders is how fast this space is moving. Almost two thirds of leaders (61%) say that competitive advantage depends on who has the most advanced generative AI capability, yet a similar proportion (64%) say that the risk of falling behind is driving them to ‘invest in some technologies before they have a clear understanding of the value they bring to the organization.’

Together, these two patterns result in a disconnect: innovation without strategy at the bottom, and strategy without execution at the top.

AI strategist Karl Yeh has named these the “bottom-up trap” and the “top-down fantasy.”

The ‘bottom-up trap’ happens when teams begin experimenting with AI tools in isolated, grassroots ways, often with enthusiasm but little strategic alignment. While these pilots might produce local wins, they rarely scale or integrate into core business processes.

On the other hand, the ‘top-down fantasy’ is when senior leaders set grand visions for AI transformation, possibly in response to hype or competitive pressure, but then fail to understand the operational realities, data readiness, or cultural change required to make them work.

Successful AI transformation happens when organizations bridge this gap, combining clear strategic intent with empowered experimentation, supported by cross-functional teams and strong governance

For marketers, the lesson is clear: AI transformation won’t succeed through isolated tools or vague ambitions. To unlock its real potential, marketing leaders must connect experimentation to business outcomes, ensure teams have the capabilities and data foundations they need, and build a culture of learning and iteration. Ultimately, transformation isn’t only about the technology: it’s about leadership, structure, and follow-through