What 2025 taught us, and what 2026 demands
2025 was the year the old marketing playbook stopped working. 2026 won't be about incremental adaptation. It will be about reconstruction.
2025 was the year the old marketing playbook stopped working. Not dramatically. No single event broke it. But the accumulation of shifts reached a tipping point that most marketing leaders felt, even if they couldn’t name it precisely.
Signal loss accelerated. AI intermediation went mainstream. Measurement got harder. And the agency model that’s served the industry for decades started showing cracks that couldn’t be papered over.
2026 won’t be about incremental adaptation. It will be about reconstruction. Rebuilding how brands discover and win customers in an environment where the old inputs no longer exist.
What 2025 made clear
The signals aren’t coming back
Privacy regulation tightened. Platform compliance increased. Third-party cookies survived in Chrome, but the signal loss they represented is happening anyway. Privacy regulation, consent frameworks, and platform changes are eroding the raw behavioural data marketers relied on.
2025 was the year CMOs stopped waiting for a workaround and started accepting this as permanent. The brands that moved early on first-party data strategies gained ground. Those still hoping for a reprieve fell further behind.
AI entered the customer journey and hid it
AI Mode, ChatGPT search, Perplexity, Copilot. 2025 saw AI move from novelty to genuine channel. Customers started getting answers without clicking through. Discovery happened inside the model, invisible to attribution systems.
This is what I call the Invisible Middle. The part of the journey where intent forms, options are evaluated, and decisions crystallise. It now happens in environments marketers can’t observe directly.
Measurement broke and started being rebuilt
Deterministic attribution now captures a shrinking share of customer journeys. Multi-touch models couldn’t account for journeys they couldn’t see. In response, sophisticated marketers returned to approaches that had fallen out of fashion: marketing mix modelling, incrementality testing, econometric analysis.
2025 wasn’t the year measurement died. It was the year inference-led measurement started replacing observation-based attribution.
The agency model hit the wall
Retainers that couldn’t demonstrate clear value got cut. Hourly billing made less sense as AI compressed delivery timelines. CMOs under budget pressure started asking harder questions: what exactly am I paying for, and what’s it worth?
The agencies that thrived were those offering genuine expertise, clear value articulation, and flexibility across engagement modes. Those still selling hours struggled.
In-housing continued, but capability gaps remained
Operational marketing moved further in-house. But strategy, data architecture, measurement science, and AI adoption exposed gaps that internal teams couldn’t fill alone. The need for external partners didn’t disappear. It changed shape.
What 2026 demands
Discovery must be reconstructed
The mechanics of how brands find and win customers are fundamentally changing. 2026 is the year to rebuild. New data foundations. New measurement approaches. New channel strategies. New operating models.
This isn’t optimisation. It’s reconstruction.
First-party data becomes non-negotiable
If you haven’t built a durable first-party data asset by now, 2026 is your last window before the gap becomes insurmountable. This means identity strategy, data architecture, and the organisational capability to activate what you collect.
Measurement shifts from attribution to inference
Stop trying to observe every touchpoint. Start building the capability to infer impact through modelling, experimentation, and econometric approaches. The brands that master inference-led measurement will make better decisions than those still chasing deterministic attribution.
AI is both threat and solution
AI intermediation will continue to hide customer journeys. But AI tools will also power the inference capabilities marketers need: predictive modelling, automated analysis, content at scale. The winners will be those who use AI to solve the problems AI creates.
Content and creative must scale, with governance
Campaign automation (Performance Max, Advantage+, Andromeda) demands volume. AI can produce it. But without quality governance and structured testing, volume becomes noise. 2026 requires a mature approach to AI-assisted content: scale with control.
Paid media diversifies
Retail media continues to absorb budget. AI placements emerge as a new channel. Traffic volatility increases as platforms shift. The paid media mix in 2026 will look different from 2025, and the measurement challenges will be harder.
Agency partnerships get restructured
The retainer is dead. Hourly billing is dying. CMOs should demand partners who can teach, guide, and enable, not just execute. Value must be clearly articulated, delivered, and measured. If your agency can’t explain the value they create, find one that can.
Operating models must evolve
Leaner teams. New skills: data science, AI fluency, experimentation design. Faster governance. The marketing org chart that worked in 2022 won’t work in 2026. Structure follows strategy, and the strategy has changed.
The bottom line
2025 was the year the old model broke. 2026 is the year to build what replaces it.
The CMOs who treat this as a moment of reconstruction, not just adaptation, will create advantage that compounds. Those who keep optimising a broken playbook will fall further behind.
One thing hasn’t changed: marketing exists to find customers, earn their attention, and create value. The mechanics are being reconstructed. The mission isn’t.