Marketing Procurement

All change please: There’s a new model in agency land

“You are now buying something different. You’re buying new services. The ones you were buying historically don’t go away but spend is now going into experiential AI, analytics, data, marketing technology and more. That requires new skill sets, even new ways of paying for these services. It’s changing the dynamic.”

AI isn’t why the agency model is evolving. It’s just a catalyst for change that has been needed for years. Agency Mania’s Co-Founder and Principal, Bruno Gralpois argues that, while the new agency model finally has potential to show its true value to marketing procurement, for those who don’t evolve there’s danger ahead.

The agency world would like to think it has moved on from the Mad Men era. The lunchtime martinis may have been replaced by a kale smoothie and the corner offices swapped for digital nomads but in many respects, little has changed from the fifties.

Billings are fee-based, parsed out in hours from juniors right up to agency head. The brief is still ultimately, ‘make us an ad’. But as the number of channels and speed of communication has ballooned, the agency model and ecosystem has become increasingly complex – and stressed.

Unlike the fifties when the treatment was revealed in the boardroom one week, and in a magazine on the newsstands the next, the workings of today’s model have become opaque. Client marketers and procurement specialists are sceptical – What exactly am I paying for?

AI is reshaping all of this.

The first, very obvious, change is that AI is a time-saving tool.

While some of its more complex potential applications are yet to meet expectations, it has been most effective in crushing time spent on repetitive tasks, powering through volumes of data in a fraction of the time it takes a human and condensing workflows from months, to weeks, days, even hours and minutes.

Now clients pose that question with even more urgency – What exactly am I paying for?

Agency Mania’s Co-Founder and Principal, Bruno Gralpois reveals that time isn’t actually what clients have been paying for. “It’s natural to believe that if an ad takes 20% less time to produce, should it not be 20% cheaper?

The old model of billing by time would suggest so. But this misses the point,” he warns. “The client has always been paying for the strategic and creative expertise of the agency, the data it has access to which enhances campaign performance.

The agency has simply been sucking up the fact that it had several weeks of tedious grunt work to generate the work. Several weeks where it could have been adding more value to the client project or boosting its portfolio. With AI, that is now an option.”

The evolution is best explained through the six pillars of agency model change:

1. Transactional to Transformative: Moving the mindset from paying for items – staff billable hours, creative and media services – to investing in data-driven, multidisciplinary services which include consulting, data management and analytics, marketing technology and omnichannel marketing experience.

2. Complicated to Simplified: In almost every aspect of the agency-client relationship, there is complexity. Multiple contracts, multiple suppliers – even within the same holding company – multiple points of contact.

“Simplifying how you work with agencies, how they interface with you, how you set up a contract and how you pay them all has to be simplified,” Gralpois insists.“Establish a single point of contact, use AI to parse contracts to a single document, define what value really looks like, measure and report on that basis only.”

3. Commercial to Collaborative: Fees and labour-based billings put clients and agencies on a combative footing, each looking to the bottom line before the value. A collaborative approach means defining the value to be derived from the relationship, the skills and tools needed to achieve it and remunerating on that basis.

4. Longevity to Adaptable: A benchmark of agency success was in length of tenure. “That can be a good thing. There is a lot of value in building institutional knowledge of a client’s business, particularly if client teams change frequently. But we also need fresh ideas, access to specialist skills,” Gralpois says.

Agency reviews are now weeks rather than months long, using AI for the RFPs, getting into the chemistry check and starting to do the work right away. He adds: “Agencies must be more adaptable, regularly bring in new talent to answer brands’ evolving needs.”

5. Transparency to Trust & Verify: Letting go of granular measurement doesn’t mean letting go of understanding.

Gralpois believes agencies are trying to understand the wrong aspects of the process: “While opacity in a process isn’t to be welcomed, you must make sure you’re striving to understand the right thing. One doesn’t always need to know how the sausage is made, simply that it’s the best tasting sausage you’ve ever had, measured against all previous sausages.”

6. Retained to Outcome: The retainer model, where agencies are compensated according to how many hours and staff are involved in creative production, is responsible for a misunderstanding of what clients are paying for.

You are not paying for the 20 hours it takes a junior member of staff to accomplish a task. You are paying for the task. Introducing AI to reduce that 20 hours to minutes does not diminish the value of the task that has been accomplished.

Gralpois notes that this move to outcome-based understanding is long-overdue: “Not only does outcome-based compensation reframe the agency’s work more logically (paid for what they do, not how they do it), it opens the door to more effective collaboration that delivers even higher value work.”

Invest to secure higher value work

“You are now buying something different. You’re buying new services. The ones you were buying historically don’t go away but spend is now going into experiential AI, analytics, data, marketing technology and more. That requires new skill sets, even new ways of paying for these services. It’s changing the dynamic,” Gralpois claims.

Just look at the landscape at the moment. It goes without saying that the tech giants are investing in AI infrastructure – more than $300bn in 2025. But so are the large agencies. WPP increased its AI investment to $381m, acquiring InfoSum in 2025 to boost its data capabilities. Publicis has dropped $327m for a three-year plan to enhance its AI offering.

Also telling is that, in the list of the top 15 agencies by revenue, 2024, Accenture Song was number two, and Deloitte Digital number five. Two heavy investors in AI, data, tech and strategy.

As a client, your investment is also moving away from rote work and into upskilling. Understanding and using new tech, delving deeper into data than was possible before, reaching new audiences, and engaging deep personalisation at scale.

He adds a word of warning, “leading agencies are all investing in this and more. Instead of asking why your marketing spend is being invested here, you should be asking questions if it isn’t.”

Break down complexity

A new model is also needed because, frankly, the ecosystem is now too complex to manage sensibly. Large brands are, themselves, complex with multi brand, multi region arms.

Even the big holding companies are finding their own, internal brands hard to navigate – M&A activity has created monsters. Both parties are then working together across an almost incomprehensibly large and multifaceted media landscape.

“The irony is that tools introduced to help marketing procurement navigate these relationships more transparently are just adding to the complexity,” Gralpois admits. “In part, this has happened because clients are asking the wrong questions.

In a bid to gain control over spend, they want to see inside the black box. You tell me programmatic advertising works, but I need to know how. You can understand it better, perhaps, but that doesn’t mean that’s where your focus should be.

It’s unsurprising that after rising for years, performance-based compensation is declining in popularity.” That’s because performance is often conflated with value and they’re not the same. The latter is also harder to define, one reason perhaps why agencies and procurement shy away from the discussion.

What is my definition of value and a brand and how do you manifest that as an agency?

Because there’s very little definition of value that has been agreed upon, it’s very hard for the buyer to know what they’re getting, and for the agency to price. It may seem like a brave new world, but Gralpois is convinced that we’re moving towards better clarity as well as better work.

“By moving away from the minutiae that are little more than a snapshot, and into building stronger KPIs and more robust measurement that means we can articulate and quantify that value, clients and agencies can cut through that complexity and build a better understanding of their needs.”

Marketing procurement should of course welcome the changes that AI is bringing to the agency landscape. But not because of the prospect of bargain bin discounts. Instead, AI is bringing a better understanding of how to define, deliver and improve the value of agency work.

It is creating more streamlined, simplified paths to collaboration that will be to everyone’s benefit.