The Future of Consulting and Agencies

The future of consulting and agencies is being decided by one question: what are clients still willing to pay for once AI can produce the thing that used to fill the invoice? Consulting firms have always sold analysis, and agencies have always sold production. AI now does a large share of both, quickly and cheaply, so both business models are being repriced at the same time and for the same reason.

That does not mean either kind of firm disappears. It means the part of the work that was billed by volume is losing its price, and the part that was always the real product, judgment, taste and accountability, has to be sold on its own terms. Firms that can make that shift will be smaller and more profitable. Firms that cannot will spend the next few years defending hours their clients no longer believe in.

This briefing is the hub for that question. The consulting side is covered in depth in management consulting in the age of AI; this piece sets consulting and agencies side by side, because the forces are the same and the answers rhyme.

The pressure, from both directions

On the agency side, the shift has been building for years and AI has accelerated it. The Association of National Advertisers found in 2023 that 82% of its members run an in-house agency, up from 58% in 2013 and 42% in 2008. A 2026 survey of more than 200 agency professionals by the advertising software company Basis found 65% of agencies had clients move work in-house in the previous twelve months, nearly 40% had made layoffs, and nine in ten believed AI threatens their primary revenue streams. The largest holding companies are consolidating: Omnicom completed its acquisition of Interpublic in late 2025 and retired several of its best-known agency networks in the process.

On the consulting side, the pressure shows up in the pyramid rather than the headlines. The Big Four have slowed partner promotions, citing uncertainty about how much work will stay billable, and clients ask why they are paying for weeks of analysis their own team could now draft in a day.

Clients are pushing from the other direction too. Thomson Reuters’ Future of Professionals report 2026 found 78% of corporate clients say AI-enabled quality improvements from their outside firms are very important or essential, while only 6% say most or all of their providers deliver them. Clients are not asking for less expertise. They are asking for the saving to show up on their side of the table.

What is actually being repriced

Two things lose their price fastest, and they are the two things both industries grew large on.

Volume of analysis. Research, benchmarking, first-draft models, market scans, the slide that summarises forty interviews. This was the work of junior consultants, sold at a margin because it took time. It now takes much less.

Volume of production. Variations of an ad, resized assets, first-draft copy, routine design, reporting decks. This was the engine of agency retainers. The ad platforms now offer tools that generate it themselves.

What stays expensive is everything that was never really about volume: deciding what the client should do, knowing which idea is right for this brand, carrying the risk of a recommendation, and holding a relationship through a difficult quarter. Those were always the product. The volume was how the product got paid for.

The four models that are emerging

Across both industries, firms that are adapting tend to land on one of four models.

• Productised. A defined service at a fixed price, delivered with AI inside the process. The margin comes from the firm getting faster, rather than being given away with every hour saved.
• Outcome-priced. Fees tied to results the client can measure. Hard to structure well, and powerful when the outcome is genuinely within the firm’s control.
• Embedded. Senior people working inside the client part-time, owning decisions rather than delivering documents. This is the fractional executive model, and it is growing in consulting and in marketing leadership alike.
• Small and system-led. A handful of senior people running their own AI systems, doing work that once needed a floor of juniors. These firms compete on judgment and speed, not headcount.

The model that is not adapting is the one that sells hours of undifferentiated work at a markup, which is exactly the one most firms were built on.

If you buy from consultants or agencies

The buyer’s position has improved, and it is worth using. Ask how AI is used on your work and how it is checked. Ask what you will own at the end: the files, the systems, the reasoning, not just the deliverable. And ask how the saving from AI is shared, because if the firm is billing the same hours for work that now takes half the time, the saving is staying with them.

The practical checks for choosing a firm are set out in telling substance from hype in AI consulting firms. The wider picture of AI across professional firms, including law and accounting, is in AI for professional services firms.

If you run a consultancy or an agency

Three moves separate the firms that are getting stronger from the ones that are waiting.

First, decide what you are actually selling and price that, rather than the time it takes. If the honest answer is judgment, stop wrapping it in hours of production that clients now see through.

Second, own your own AI system: the instructions, the context about each client, the templates and the review steps, as assets the firm keeps. That is what makes a small team capable of work that used to need a large one, and it is what a competitor cannot copy by subscribing to the same tools.

Third, redesign how juniors learn. The work that used to train them is shrinking. Give them the checking, the client context and the explanation of why an answer is right, or the firm will have no seniors in ten years.

How this silo fits together

This is the hub for the future of consulting and agencies. Further briefings in the silo take the question to specific ground, starting with whether AI will replace consultants outright, and the industry trends worth watching. The broader shift in knowledge work that sits underneath all of it has its own silo in this pillar.

If you are deciding what your firm should actually sell from here, tell Daniel what you sell today, what you have built with AI so far, and what you need the business to do for you. Book 30 minutes with Daniel.

Frequently Asked Questions

What is the future of consulting and agencies?

A shift from selling volume to selling judgment. AI now handles much of the analysis consultants billed for and the production agencies were paid for, so both are being repriced. Firms that sell decisions, taste and accountability on their own terms, and run their own AI systems, are likely to be smaller and more profitable. Firms selling undifferentiated hours face continued pressure.

Will AI replace advertising agencies?

Not entirely, but it is taking much of their production work. The ANA found 82% of its members run in-house agencies, and a 2026 Basis survey of agency professionals found 65% of agencies had clients move work in-house in the previous year. Agencies that sell strategy, brand judgment and creative leadership remain valuable; agencies built on volume production are most exposed.

How is AI changing consulting firms?

It is shrinking the junior analysis layer that filled billable hours and pressuring the pyramid model built on it. Clients increasingly expect AI-enabled quality: Thomson Reuters found 78% of corporate clients consider it very important or essential, but only 6% say most or all of their providers deliver it.

What business models work for consultancies and agencies in the AI era?

Four are emerging: productised services at a fixed price, outcome-based fees, embedded or fractional senior roles, and small teams running their own AI systems. What they share is that the firm keeps the benefit of working faster rather than giving it away with every hour saved.

What should clients ask consultants and agencies about AI?

How AI is used on their work and how it is checked, what they will own at the end of the engagement, and how the time AI saves is reflected in the fees. If a firm bills the same hours for work that now takes half the time, the saving is staying with the firm.

How should a consultancy or agency prepare for AI?

Decide what it is really selling and price that rather than time; build and own its AI system, including instructions, client context and review steps; and redesign how junior staff learn, since the routine work that trained them is shrinking.