Change, Renew, Discover: New Ways of Working for Marketing Procurement inthe AI Era

By Helen Thompson

Analytics information printed out on paper on a table

2026 is the year trust becomes marketing procurement's most critical currency —

and seeing is no longer believing.

CHANGE: What’s Actually Shifting

The AI-driven sea plummet in agency pricing that many were predicting in 2025

hasn’t materialised. For Iain Seers, CEO of RightSpend, 2026 feels like it could be

different. “2026 is going to be a big year,” Seers asserts. “There’s been a lot of talk,

and not a lot that’s tangible, and I think this year things are going to change.”

Ask Seers where AI is making the most immediate impact, and the answer is

unambiguous. “Production has got to change,” he says. “The areas that are being

impacted are becoming clearer.” Retouching, CGI, asset creation at scale; these are

the categories where AI tools are already delivering results, and where the gap

between what agencies are charging and what the work actually costs is beginning

to show.

The kind of content that once required a full production crew can now be created in

seconds. The recent viral Brad Pitt/ Tom Cruise fight scene generated by two lines of

code in Seedance 2.0 provoked a media storm, a flurry of legal action, and Deadpool

Director Rhett Reese to proclaim that “Hollywood is about to be decimated.”

The pace of change is not uniform, however, and Seers is careful not to overstate

the disruption. Creative strategy, brand thinking, and original thought remain

stubbornly human endeavours. “It still requires humans to select the right tools for

the right job,” he says. “There are no new ideas — everything is based on something

that’s come before.” AI excels at recombination and execution. It does not yet

replace the spark of human creativity.

CHANGE: Pricing for New Value Chains

This distinction matters enormously for procurement, because it means a blanket

approach to AI pricing will not work. The production category needs to be re-

evaluated now. Strategic and planning services need a different conversation

entirely.

What makes this moment particularly sharp is the mismatch between cost reality and

commercial structure. “[Brand’s] budgets are not being halved. Costing isn’t

changing, but it should be,” Seers says plainly, and most procurement teams don’t

yet have the visibility to challenge this. The dominant pricing model, time and

materials, built around FTEs, persists not because it’s right, but because it’s familiar.

“Why are clients still using time and FTEs?” Seers asks. “Because the work being

 

done still requires people. It’s easy to calculate and let’s be honest it’s a global

model that works.”

New models are beginning to emerge. Output-based or deliverables pricing works

well for high-volume production. In this structure, images generated by a prompt

should cost less than images produced by a studio team, for example. Platform or

licence-based pricing is appearing too, with agencies charging clients to access

proprietary AI systems. And then there is what Seers calls the HTE model (Human +

Technology + Employee), a hybrid approach that averages out the blended cost of

people and tools, with lower day rates offset by a technology fee component. None

of these is a silver bullet, but they represent the direction of travel. The question for

procurement is whether they have the knowledge and confidence to lead those

conversations.

RENEW: Rebuilding the Agency Relationship Around Trust

If the commercial challenge is significant, the relationship challenge may be even

more complex. Seers identifies a shift in the nature of the problem that procurement

teams face. “Transparency is being overtaken by trust,” he says.

This is not simply about audit rights or contract compliance. It reflects something

broader happening in the world. “You can’t believe what you see,” notes Seers.

“From a brand perspective, you have to make sure people trust what you are

saying.” In a media environment where AI-generated content is proliferating and

audiences are growing more sceptical; brand integrity is at stake. Seers notes that

some brands are already choosing to hold back on AI-generated social content, not

because the quality isn’t there, but because their audiences don’t trust it. “They don’t

want their audience to not believe what they’re putting out.”

For procurement, this trust dynamic plays out in a specific and practical way.

According to Seers, “The first question you should be asking is: do you trust the

agency?” Agencies have developed their own AI tools and workflows, often

proprietary, and the visibility into how work is actually produced, (and at what cost),

is diminishing rather than improving. “Agencies have got their own flavours of AI.

That’s the conundrum… what worries me is a move to a world where it’s actually

less transparent.” The term Seers uses for the outcome of this dynamic is pointed:

ghost AI. Agencies deploying AI to complete work faster and capturing the time

saving as additional margin, while the client’s SoW remains unchanged.

The antidote is not more contractual complexity, rather it’s a more honest

commercial relationship. Seers advocates for agreed markups, performance-based

bonuses tied to turnaround and quality, and a move away from opaque blended

rates. It also means procurement teams getting to grips with the new roles appearing

on agency SoWs. Prompt engineers, AI content specialists, and automation leads

are all becoming standard line items. Understanding what these roles actually do,

and what a fair rate looks like, is rapidly becoming a core procurement competency.

DISCOVER: Practical Steps for Procurement

“Marketing procurement can do more,” says Seers. “There’s a real opportunity here,

not just to cut costs but to lead.” In a moment of genuine structural change,

procurement teams that understand the technology, challenge the pricing, and help

their organisations build trusted agency partnerships will be operating at a different

level to those still running RFPs against traditional rate cards.

But where to start? Seers recommends a structured discovery phase as the

foundation. That means mapping the areas of marketing spend most likely to be

affected by AI, particularly production categories, and then examining what’s actually

happening inside those budgets. Are AI-related roles appearing on SoWs? Are

agencies billing for tools or platforms on top of their fees? Is the time recorded on

jobs consistent with what AI-assisted production should realistically take? These are

answerable questions, and the answers are revealing.

From there, procurement can move to evaluation and negotiation. That means

building AI capability assessment into agency selection criteria. This can no longer

be a box-ticking exercise, but rather needs to become a genuine signal of how ready

a partner is to operate in this environment. It means asking direct questions about

how AI is being used in the production of work, what the cost implications are, and

how those savings are being shared. And it means being willing to put commercial

models on the table and propose alternatives.

The Watershed Year

Seers’s view of what lies ahead is optimistic but grounded. The technology is real.

The commercial pressure is real. And the opportunity for procurement to shape

fairer, more transparent models across the marketing ecosystem is real. “The brands

and procurement teams that engage with this now will have the competitive

advantage,” he says.

The question is simply one of timing. “Will 2026 be the watershed year? I think it

might be. The first movement at scale could come this year.” The conditions are

right: the tools are maturing, the conversations are happening, and the industry is

finally moving from talk to action.

Whatever changes, one principle will hold. “Human first and human last,” says

Seers. “That principle isn’t going away.” The brief still needs a human mind. The final

decision still needs a human judgement. Everything in between is up for grabs, and

procurement is better placed than most to shape what comes next.