Back

The Big AI Secret · ongoing research

Agencies: AI Adoption Guide 

How whole teams turn AI into value, not just how individuals adopt it.

Researched by Maybe* research team, Sector Analyst

Research underway

This Guide is being built from continuing interviews. It grows sharper as they come in, and moves from estimated to measured. Apply to contribute.

A candid look behind the headlines at how AI is really being used in teams today. The finding underneath all of it is simple: the gap is not adoption, it is integration across the team. The most AI-mature sector yet agencies run the most tools and pay the highest disconnection tax. The gap is integration, not adoption.

ShareXLinkedIn
AI does not replace people. It amplifies those with the confidence to experiment and the courage to lead change.Polly Barnfield OBE · CEO, Maybe*

Scores in this Guide are shown on the Integration Index, 0 to 100. Earlier editions used an 18-point scale; it is the same measurement expressed differently. This is ongoing research, updated as new interviews come in. All contributors are anonymised to protect the honesty of what was shared. That is what makes it unusual: it captures the unfiltered truth of AI in action.

Drawn from over 3,138 in-depth interviews and questionnaire responses. Findings are cross-referenced against published research from other organisations, including Stanford, McKinsey and BCG, where comparable figures exist.

Where agencies stand

Agencies lead, and that is why the gap bites hardest

Agencies are the most AI-mature sector in the research, and smaller agencies are often ahead of larger ones. But leading on adoption is not the same as integrating it, and the very speed of agency adoption has left many running a sprawl of disconnected tools. The agency question is no longer whether to use AI. It is how to make it hold together.

82%
use AI for content variation and adaptation, the highest-adoption use case.
Agency patterns
ShareXLinkedIn
74%
use AI for performance analysis and improvement.
Agency patterns
ShareXLinkedIn
67%
use AI for creative concepting and ideation.
Agency patterns
ShareXLinkedIn
63%
use AI for multi-channel distribution optimisation.
Agency patterns
ShareXLinkedIn
In fact the smaller agencies tend to be much more on it with AI than the bigger agencies.Agency leader · agency interviews

The hidden cost

The disconnection tax

Because agencies adopted fastest, they run the most tools, and pay the highest price when those tools do not talk to each other. One agency running twelve separate AI apps found it took longer to manage the process than before AI. Across the research, roughly a fifth of team hours is lost to this fragmentation, time that delivers no incremental value. For an agency, that is billable capacity.

4 to 7
AI tools a typical agency runs at once, with some running far more.
Tool overload
ShareXLinkedIn
~18%
of team hours lost to disconnected tools, the disconnection tax.
Hidden cost
ShareXLinkedIn
We're producing more content than ever, but it takes longer to manage the process than before AI.Operations lead · digital agency
We've got 18 AI logins, 7 dashboards, and no shared strategy.Agency operations lead

Perception

Quiet wins, loud worries

Agencies are getting real results, and often hiding them. Many will not mention AI in a pitch, because clients still hear the word and imagine shortcuts rather than strategy. The wins are quiet and the worries are loud: brand safety, authenticity, and a creative team more afraid of losing its voice than its job.

Our AI use is sophisticated. But if our clients heard 'AI,' they'd imagine lazy automation, not the tailored, strategic work we're actually doing.CMO · agency

The prize

The capacity prize

The agencies that move from experimenting to integrating do not cut their teams. They do more with them. The recurring pattern is the same team handling around a third more clients without adding headcount, and creative people spending far more of the week on the work that bills. The prize is capacity, the equivalent of more people, without hiring them.

30%
more clients handled without adding headcount, by agencies that integrated AI.
Agency ROI
ShareXLinkedIn
40%
faster first drafts once AI was integrated into content.
Agency ROI
ShareXLinkedIn
380%
ROI within six months in one integrated agency content operation.
Agency ROI
ShareXLinkedIn

Consolidation

From tool chaos to a connected stack

The winning move is consolidation. The agencies pulling ahead cut from a sprawl of overlapping tools to a connected stack with shared prompt libraries and one source of truth, then govern it. One firm cut tool spend by 60% and lifted adoption by 40% in the same move.

60%
cut in AI tool spend after consolidating a fragmented stack.
Tool overload
ShareXLinkedIn
85%
agency AI adoption, up from 20%, after a culture-led rollout.
Human element
ShareXLinkedIn
40%
more creative output per team member with a collaborative agent approach.
Agents
ShareXLinkedIn

Client trust

Client trust and brand voice

For agencies, the output is the product, so on-brand quality and client trust decide everything. Most creatives treat AI output as a base to shape, not a finished article, and the strongest agencies build brand-aligned prompt libraries and human review into every workflow. The line that keeps clients is simple: AI does the admin, the agency brings the brand story, the strategy and the audience.

It's very rare that I'm happy with what AI has created. I always just use it as a base or a foundation and then tweak it how I want it.Agency creative
You are the chef and AI is the sous chef. It will do all the admin, but you bring the brand story, the strategy, the audience.Agency founder

What to do next

The agency roadmap

The path from Explorer to Integrator is well worn: start where the pain and the volume are highest, usually content and reporting; consolidate onto a connected stack; put brand and governance rules in first; and integrate AI into the tools the team already runs, rather than adding more. Prove one Task, then scale.

Research established

Agencies were the majority of the sample behind The Big AI Secret, so this Guide is measured, not estimated. New voices still refine it, and take-part stays open at the foot of the Guide.

This research is underway

Apply to contribute

This Guide is being built from live interviews. Two ways to shape it, both credited.

For agency teams

Get your Stage

Ten minutes of questions about how your agency actually uses AI. You see where you sit on the maturity model, which dimension is holding you back, and how that compares with the agencies already ahead of you.

Reward · Your stage, your gap by dimension, and the Tasks that close it.

FIND YOUR STAGE

For researchers and graduates

Author a published Guide

Join the growing research team. You lead a Guide from interviews to publication, credited on it as the Sector Analyst. A published Guide is a real, permanent credential.

Reward · Author a published Guide as its named Sector Analyst.

APPLY TO BE AN AI ANALYST

Get the findings as they land, one email a month.

The Big AI Secret · living research, updated monthly · drawn from over 3,138 in-depth interviews and questionnaire responses. All figures are sourced to the research; quotes are anonymised and verbatim. Produced by the team at Maybe*.

Where you stand

Where does your business stand in marketing?

Two minutes against the same maturity model this research uses. See your stage, where your gaps are, and a short list of what to do next.

FIND YOUR STAGE →