I spent yesterday at the Prolific North event, courtesy of my clients at Twisted Loop. If you want a snapshot of where the UK agency world sits with AI, it’s a study in optimistic paralysis.
We are all excited. We are all experimenting. But the gap between our strategic intent and our operational execution is now the single biggest commercial risk we face.
The AI readiness paradox: we’re all using it, but few are ready
The data shared by Twisted Loop’s recent report – The State of AI Adoption in UK Agencies 2025 – brings together insights from across the sector, highlighting where momentum is strongest, where gaps remain, and what’s shaping the next stage of AI adoption.
It paints a fascinating, almost contradictory, picture of the market:
| The Opportunity | The Reality (The Gap) |
| 90% of professionals are already using AI daily. | Only 2% feel their organisations are truly prepared. |
| 89% feel positive about AI’s potential to raise value. | 66% are terrified of over-reliance on it. |
| 70% prioritise internal efficiency. | 83% haven’t built a single piece of bespoke tooling. |
We’re doing stuff. We’re deploying generic tools for internal efficiencies, then wondering why we aren’t seeing transformational change.
Meanwhile, the exponential curve looms large: AI capability is doubling every seven months. We are 18–24 months away from 10x productivity for knowledge workers who know how to harness it.
The gap between knowing and doing is not closing.
The billing model is dying
Data shared highlighted the core commercial tension:
- 75% of agencies use AI, but only 23% have a public-facing proposition.
- Why? Because the billing model is broken. Time + materials doesn’t scale when AI delivers a three-month project in three days.
The growth is going to Growth Partnership clients who treat their agency as an embedded team, focused on value, outcomes, and equity. You have to rewrite your pricing around time + machine, not just time.
The operational blueprint: build, don’t buy
A lot of discussion around getting started but also how do you know you’re getting it right? Some are in experimentation, some are building infrastructure. Some are outsourcing to other markets.
Investment ranged from carving out time away from client work to significant hundreds of thousands of investment in people and infrastructure. Clearly it is hard to complete with the millions being spent by the HoldCos.
While a lot of the discussion naturally veered to the tools, especially with the Gemini 3 announcement dropping the night before, the real conversation is about the why and the use cases. Identify and solve the problems. Start there, not which tool. Linking it back to the commercial discussion, better strategy, creativity, and insight that guides commercial decisions is billable. This is the high-value human output.
Summary
- Increase fluency: Every week you delay costs you compounding advantage.
- Operationalise value: Build an operational blueprint that translates AI capability into client value propositions today, not “when the tools are perfect.”
- Ship fast: The market will reward the agencies that productise insight and speed, not just headcount.
- Re-think pricing: Time + materials is dying. Value and outcomes are the future.
- Hire curious people: The role is now curator, translator, and tastemaker.
Treat AI as a toy, you will become one. Treat it as infrastructure and problem solving, it could be the greatest force-multiplier this industry has ever seen.
Thanks to Twisted Loop for the invitation, and to all the speakers for the provocations.
