AI strategy for agencies: decide what you sell before you decide what to automate.

I'm Yiannis Papastergiou, an independent AI consultant based in Thessaloniki, Greece. For marketing, creative and digital agencies of roughly €1M+ revenue I write one costed AI roadmap in two to four weeks: the offer, the pricing, the delivery work to automate and the rules for client material, in an order your partners can agree on. In person in Athens and Thessaloniki, remote everywhere else.

AI strategy · Marketing, creative and digital agencies

About 3–5 minutes, mostly one-tap answers, then your name and work email. Your result on screen, with a 7-day plan. No call required.

Take the free AI Growth GPS

See the AI P.R.O. Method

What is an AI strategy for agencies?

An AI strategy for agencies is a written plan that settles four things in order: what the agency sells once AI makes production faster, how that work is priced, which delivery and admin workflows are automated first, and what clients are told about AI. It ends with a costed roadmap, an owner for each move and a list of ideas to ignore.

Mine is the AI Strategy & Roadmap. Your investment is confirmed in a written proposal.

This page is about marketing, creative and digital agencies. It is not about AI strategy for government agencies.

See the AI P.R.O. Method

The short answer

Why does a marketing agency need an AI strategy, not more AI tools?

Because tools make the old product cheaper to produce, and the old product is hours. VoxComm, which represents more than 35 national agency trade associations, put the order plainly in its March 2026 guide, Redesigning the Agency Value Model: agencies must first redefine what they sell before changing how they price.

Clients have already moved. In the World Federation of Advertisers' September 2026 research among 46 of the world's largest brand owners, efficiency was the primary objective of AI for 22%, down from 70% two years earlier. A third now weigh efficiency, effectiveness and growth equally.

Your clients stopped buying AI for savings alone. An agency strategy that is only about savings is behind them.

And some would rather not buy at all. In Swydo's 2026 Agency Survival Guide, built on a survey of 1,306 PPC professionals, 20% of clients plan to replace agency work with AI instead of switching agencies. Without a strategy, an agency meets that with:

A tool subscription per team, and no shared view of what any of it saved.

Partners who each believe something different about AI and pricing, and a rate card that reflects none of it.

A new AI service announced on the website that nobody has yet delivered twice.

The workflows, one by one: AI consulting for agencies

The problem

What should an AI roadmap for agencies decide, and in what order?

Five decisions. An AI roadmap for agencies that starts at the third one automates a business model the first two would have changed.

Which services stay bespoke and senior, and which become repeatable packages a client can buy by outcome. VoxComm's guide describes the target as productised solutions designed around business outcomes rather than time and people.

1. The offer

Which work leaves hourly billing, account by account, and what you say at each renewal. This is the decision partners argue about longest, so it gets a date and an owner.

2. The price

Reporting, briefs, first drafts, variations, QA, timesheets. Each scored by hours a month and by how clean its inputs are. Three get built first. The rest wait.

3. The delivery work

What each contract allows, which tools are approved for which client, and one written AI policy you can hand over. In the WFA research, 69% of major brands require agencies to disclose when AI is used.

4. The client rules

Who reviews AI output before it leaves the building, and how juniors learn the craft when the first draft is no longer theirs to write. A plan that skips this tends to lose the team.

5. The people

And a sixth list beside them: the AI ideas I would tell you to ignore, each with its reason.

The order matters

What does an AI strategy for a marketing agency look like on one page?

An invented example, to show the form: a 35-person agency selling paid media, content and monthly reporting to mid-sized retailers.

Sell: reporting stops being a line of hours and becomes part of a fixed monthly package. Content variations become a priced unit.

Do first: report drafts pulled from the ad platforms and analytics, briefs built from call notes, and first-draft variations in each client's voice, all checked by a person.

Tell clients: one written AI policy, with the approved tools for each account, sent before renewal season.

Ignore for now: a chatbot product to resell, because nobody on the team has run one for a client.

Wait: campaign changes made by AI agents, until the client data rules are in place and the first three builds are measured.

The agency is invented and no result is implied. The point is the shape: a commercial decision first, three builds, one promise to clients, and a reasoned list of what not to do.

See a sample roadmap page

An illustration

What limits an agency's AI strategy: clients, contracts and the law?

In the WFA's 2026 research only 13% of major brands said they have a great deal of visibility into their agencies' AI use. A strategy that makes your use visible, on your terms, is an advantage before it is an obligation.

Clients can't see what you do with AI

In the same research, 59% of brands are introducing AI clauses on IP, confidentiality and training data, and 55% have a centrally approved list of AI tools, up from 18% in 2023. Your tool choices now sit inside your clients' rules.

Contracts are being rewritten

Forrester's 2026 study of US marketing agencies with the 4As names the barriers: accuracy and bias (63%), legal concerns (62%) and privacy and security risks (55%). For AI agents, lack of expertise (54%) and data infrastructure gaps (51%).

What stops agencies scaling AI

Article 50, which applies from 2 August 2026, requires anyone using AI to generate or manipulate image, audio or video deepfakes to disclose it, with a lighter duty for evidently artistic or creative work. For an agency producing for the EU, that belongs in the delivery plan.

The EU AI Act and synthetic content

This is information, not legal advice. Where a contract or the Act needs a lawyer, the roadmap says so.

Constraints

Who should write an agency's AI strategy?

Route

Best at

Check first

The table is my judgement.

Independent AI consultant (me)

A costed roadmap from someone who also builds, with nothing to resell.

I have not run an agency and have no published agency client. You bring the agency economics; I bring the AI and the numbers.

Agency growth consultant or trade body guidance

Positioning, productised offers and pricing. VoxComm's 2026 guide is an example.

Who says what is technically feasible, and who puts a range on each build.

A tool vendor's playbook

A fast start on one platform.

Whether every answer is that platform.

The partners, at an offsite

Full context and no fee.

Who scores the idea that belongs to a partner in the room, and who writes it up with owners and dates.

Your options

Who would be writing your agency's AI strategy?

Yiannis Papastergiou, independent AI consultant, based in Thessaloniki, Greece. I have not run an agency, and I have no published agency client. This is what I bring instead.

For Bronzed Studio I automated the Instagram booking flow, and the owner got back at least 10 hours a week. One workflow, one number.

A small business, measured

At ARAG UK, a 500-person legal insurance business, I started as the sole developer, became technical lead, and built and led a hand-picked team of contractors.

A team built from contractors

For UCL I led the commercialisation strategy and built the interactive demo dashboard for its Citizen Prosperity Index.

A client-side strategy

My father ran a butcher shop for forty years. At fifteen I was behind that counter. I think in margins, cash and time, not in technology for its own sake.

Raised behind a counter

None of that proves what AI would do in your agency. The free AI Growth GPS lets you test my thinking on your own agency first.

Who you'd be dealing with

AI strategy consulting for agencies: where the roadmap sits, and the investment

About 3–5 minutes, mostly one-tap answers, then your name and work email. Your result on screen, with a 7-day plan. No call required.

Take the free AI Growth GPS

Your investment is confirmed in a written proposal. Two more services stand alone: AI Governance & EU AI Act Compliance and AI Training & Executive Workshops.

See the AI P.R.O. Method

Each step uses what the last one found. None is a condition of the next.

Not sure the agency is ready at all? A readiness score across data, people, process and governance, and your first three moves. 2 weeks.

AI Readiness Assessment

Before

A prioritised, costed, sequenced roadmap, including new AI services and packages instead of hours. It contains the Assessment. 2–4 weeks.

AI Strategy & Roadmap

The plan

A delivery or admin workflow from the roadmap, built in the tools you already use and measured in hours saved. 2–4 weeks.

Automation

A larger build, piloted with real users against an agreed metric. 4–8 weeks.

AI Agents & Custom AI Solutions

An owner for the plan, one to three days a week. Monthly, no contract.

Fractional Chief AI Officer

After

Services and investment

What happens after you click

No step is a condition of the next. Stop after any step and keep what you hold.

your growth constraint, your AI readiness and a 7-day plan, on screen

Step 1

The free AI Growth GPS. A 3–5 minute conversation about your agency.

the investment, confirmed in writing

Step 2

Ask for a proposal.

one decision page: three moves, each with an owner, a date and a budget

Step 3

The roadmap, over two to four weeks: deep-dive, opportunity map, roadmap, executive readout.

No call required

Questions agency owners ask about AI strategy

Five decisions and one list. What you sell once production is faster. How it is priced, account by account. Which three delivery or admin workflows are automated first. What each client contract allows and what clients are told. Who reviews AI output and how juniors keep learning. And the list of ideas to ignore. Each move gets an owner, a date and a budget.

What should an AI strategy for a marketing agency include?

Decide it before, and change it account by account. VoxComm's 2026 guide puts the order as redefining what you sell first, then how you price. My view is the same: if delivery gets faster while the rate card still sells hours, the gain goes to the client at the next renewal. I can't promise how your clients will react.

Should an agency change its pricing before or after it adopts AI?

Plan starts from €1,500 (the Numbers Round Table, refunded if it isn't worth it within seven days). Everything after that is quoted as one fixed figure in a written proposal, sized to your business, before anything starts.

What is the investment for AI strategy consulting for agencies?

Two to four weeks: a deep-dive, an opportunity map, the roadmap and an executive readout. From you I need time with the partners and team leads for the interviews, someone to walk me through your delivery process, tools and a sample of client contracts, and the leadership team together for the readout.

How long does it take, and what do you need from us?

That is when it pays most. Nine in ten US marketing agencies already use generative AI, according to Forrester and the 4As, so using it is no longer a position. The strategy decides what the tools are for: which hours come out, where they go, what the client pays for and what you can prove.

We already use AI tools in every team. Do we still need a strategy?

Yes. It includes a review of what your contracts already say, the approved tools for each account and a written AI policy you can hand to a client. In the WFA's 2026 research, 59% of major brands were adding AI clauses to agency contracts. Where a clause needs a lawyer, I say so.

Does the strategy cover what we tell clients about AI?

Not the way I work. The aim is the same team with hours of repetitive work taken off them, measured in hours saved. What the agency does with those hours is a decision for the partners, and the roadmap puts it on the table instead of leaving it to drift.

Will an AI strategy mean fewer people in the agency?

Not as a published client, and I have not run an agency. My one published result is small: for Bronzed Studio I automated the Instagram booking flow, and the owner got back at least 10 hours a week. The services are new as a set, which is why the first page about your agency is free.

Have you written an AI strategy for an agency before?

Questions

Not ready for a roadmap? Start with a free 3–5 minute conversation about your agency.

About 3–5 minutes, mostly one-tap answers, then your name and work email. Your result on screen, with a 7-day plan. No call required.

Take the free AI Growth GPS

A guided conversation with my AI Growth GPS. Tap or type your answers. Why it's free: you're pitched by AI consultants every week, so you get something useful before we ever speak. Nothing is shared.

Discover · The free AI Growth GPS

Decide what you sell. Then decide what to automate.

Take the free AI Growth GPS, or see how the AI P.R.O. Method works first. Neither starts with a call.

About 3–5 minutes, mostly one-tap answers, then your name and work email. Your result on screen, with a 7-day plan. No call required.

Take the free AI Growth GPS

See the AI P.R.O. Method