Imagine witnessing a century’s worth of progress in a single decade. It sounds implausible, yet it becomes increasingly realistic when viewed through the lens of AI acceleration – particularly if artificial general intelligence is achieved.
This is a thought experiment referenced by Will MacAskill, who points to the scale of change experienced over the last hundred years: nuclear weapons, birth control, the internet. Now consider these developments being compressed into a single decade and how it would fundamentally change the way decisions are made.
When the rate of progress increases, the margin for error narrows. The speed at which organisations need to make decisions, and make the correct decisions, becomes a crucial competitive advantage.
What if it all arrived at once?
It’s interesting to consider this compression across media planning and buying. Picture the next ten years of platform, product and behavioural change all arriving in 2026.
- What systems would need to be in place to cope with that reality?
- How would brands continue to grow in a landscape defined by constant disruption?
- Which strategies would allow teams to respond, rather than react?
Viewed through that lens, a number of structural shifts become unavoidable.
AI agents change how work gets done
Historically, media execution has relied on fixed processes and workflows. Campaign builds, QA, pacing changes and optimisation logic were defined upfront, embedded in tools or platforms, and adjusted only when performance forced a rethink.
This model assumes there is enough time between decisions to review performance, diagnose issues and manually intervene. AI agents can monitor performance continuously, detect emerging patterns earlier and adjust execution logic in response – without waiting for scheduled reviews or manual intervention.
AI agents and no-code solutions materially change the dynamic. Media practitioners can now prototype workflows, surface insights and automate optimisation directly – without waiting for development cycles or new tools. In practice, this means execution logic can be adapted much quicker as conditions change.
As the landscape fragments across more platforms, formats and signals, operational load increases even as platforms automate more of the execution layer. Without internal systems that reflect how teams actually work, complexity compounds. AI agents allow teams to shift from managing processes to managing decisions – reducing friction in day-to-day execution while preserving strategic control.
Creative diversity is now the growth lever
Creative has become the focal point of media performance discussions – and for good reason.
Platform algorithms have deprioritised granular audience targeting in favour of learning from creative signals. Meta’s latest Andromeda update has shifted the emphasis a step further. Incremental tweaks to headlines, descriptions or assets are no longer enough to drive meaningful performance gains. We analysed 1,000 ads in Q4 and found that 40% of the assets received less than £50 in spend, accounting for just 1% of total investment. In practice, these assets failed to meaningfully enter the auction. The level of creative diversity was not sufficient for algorithms to differentiate, prioritise or scale delivery – reinforcing the point that incremental variation alone is no longer enough to drive performance.
The introduction of an Entity ID makes this explicit. By grouping similar creatives under a shared identity, Meta has fundamentally changed how variation is interpreted. One hundred subtle variants deliver far less value than 10-15 clearly differentiated creative themes.
This places new demands on creative and media teams alike. Creative teams need clearer guidance on asset volume, formats, video lengths and variation. While media teams need to obtain faster insights on what works so new creative angles can be developed.
One step closer to keywordless search
If it weren’t for advertiser dependence on legacy models, Google would likely have moved away from keyword-based targeting years ago. Instead, the transition has been incremental: broad match expansion, Performance Max, and most recently AI Max. Taken together, the direction of travel is clear. Keywords are becoming less central, replaced by inferred intent, behavioural signals and content relevance.
Continuous experimentation with new formats, campaign types and creative inputs therefore becomes essential. In a keywordless environment, delayed learning or incorrect set-up increases the margin for error, making fast feedback loops critical to more effective optimisation. That same capability becomes even more important as search expands beyond traditional engines.
LLMs such as ChatGPT, Perplexity and Claude are evolving, advertising opportunities are set to follow in 2026. Early advantage won’t come from tactical execution alone, but from organisational readiness to test, learn and adapt quickly.
MMM that’s fast, frequent and actionable
Many existing measurement frameworks are still rooted in click-based attribution and rely heavily on modelled data provided by ad platforms. This inevitably skews investment toward lower-funnel acquisition channels, reinforcing short-term performance at the expense of long-term scale. As a result, brands struggle to justify investment in new audiences, emerging formats and upper-funnel activity – even when these are critical to future growth.
As the pace of change accelerates, measurement must shift from retrospective reporting to informed decision making. Faster, more actionable insight into incrementality, cross-channel effects and diminishing returns will be essential. This enables brands and agencies to move beyond reactive optimisation and toward scenario planning and predictive modelling – testing how changes in spend, creative or channel mix are likely to impact outcomes before decisions are locked in.
In an AI-driven environment, the role of measurement is not simply to validate performance after the fact. It is to guide investment, where to push more, where to hold back, and where to experiment next – at a speed that matches how quickly platforms and algorithms are already operating.
Building for speed, not certainty
If the next decade of progress really does compress into the next few years, the challenge for brands won’t be access to technology. It will be the ability to make good decisions under sustained pressure. As AI accelerates execution, multiplies creative output and reshapes discovery, the organisations that win will be those built to learn quickly, adjust course and act with intent.
Preparing for what comes next is about designing systems and teams that can keep pace when the margin for error disappears.
