The Agency Operating Model Has a Problem. Some Are Starting to Admit It.
Why are agencies struggling to prove value? This article examines the challenges facing the modern agency operating model, from campaign validation and ROI measurement to stakeholder alignment and AI disruption.
There's a particular kind of meeting that agency leaders dread. The one where a client looks at twelve months of work and asks: "But what did we actually get for this?"
It's not a hostile question. It's an honest one. And the fact that it's hard to answer says something uncomfortable about how a lot of agencies still operate.
The economics of the traditional agency model were built for a different era — when campaigns ran on annual cycles, channels were few, and "did it work?" was a question you could confidently answer at any given time. That world is gone. And yet many agencies are still running the same operating system underneath a fresh coat of AI-generated content and retooled service decks.
Hans Piechatzek, Managing Director at move:elevator, describes the shift plainly:
"Marketing has become much more complex. We have far more channels, constantly new formats and new platforms. Marketing teams everywhere are dealing with an explosion of tasks — but no explosion in budget, and certainly no explosion in headcount."
Piechatzek has spent years working out a structural response to that problem — distilled into a framework he calls W.A.V.E. But the four pressure points it addresses aren't unique to move:elevator's clients. They show up across the industry, and they're worth thinking about on their own terms.
Campaigns still get built on assumptions nobody has tested
Ask any performance marketer what they've learned from running campaigns and you'll hear some version of the same story: what the client was certain would work, didn't. What nobody expected, did.
This isn't surprising. What's surprising is how rarely that lesson travels upstream, into how briefs are written, how concepts are developed, and how strategies get signed off. Entire campaigns get built on anecdotal experience: something that worked for a different client three years ago, a hunch from a senior stakeholder, a creative instinct that felt solid in the room.
The fix doesn't have to be expensive. Usability research has long established that five user tests will surface the majority of structural problems in a concept. That's not a large budget line, but it's the difference between validating a direction and discovering it was wrong after six weeks of production.
The real cost of skipping it isn't the research fee. It's the rework, the missed performance, and the gradual loss of client trust that follows a launch that didn't land.
The stakeholder who appears at the end isn't going away
Briefings travel through organisations and lose fidelity at every step. By the time one reaches an agency, it often reflects one team's interpretation of a business need — not the full picture. Sales sees it differently than marketing. Marketing sees it differently than the C-suite. Nobody is wrong. They just weren't in the same room.
The cost surfaces late: a near-finished campaign pulled back into review, a launch that lands awkwardly because the product team had context nobody shared, a strategy that makes sense to marketing but doesn't connect to how the business actually sells.
Getting all relevant stakeholders aligned before work begins — on what success looks like, why the project exists, and what the shared path forward is — sounds elementary. In practice it's one of the most consistently skipped steps in agency-client engagements. And one of the most expensive to skip.
"What's the ROI?" is no longer a question you can answer vaguely
Expectations for proving marketing’s impact on revenue are higher than ever. Boards expect numbers that connect to business outcomes. Marketing leaders are increasingly being asked to justify budgets in terms that go beyond reach, impressions, and brand sentiment.
Most agencies — and their clients — still don't have the infrastructure to make that connection clearly. Measurement gets added later as an afterthought, reported selectively, or left to the client to figure out. The result is that marketing decisions get made on incomplete information, and budget conversations become negotiating positions rather than data-driven discussions.
Building agreed KPIs and measurement frameworks into an engagement before campaigns run changes the nature of the relationship. Results become visible to both sides. Conversations shift from defending what happened to improving what's next. That's a more durable basis for a client relationship than goodwill alone.
AI is already separating agencies — just not in the way most are talking about
Every agency is talking about AI. The conversation tends to cluster around efficiency: doing the same work faster, at lower cost. That's real, but it's the smaller part of the story.
The more consequential question is what happens to agency value when execution gets cheaper. If drafting, formatting, coding and basic research compress toward commodity cost, the work that remains — strategy, judgment, context, quality assessment — becomes both more important and harder to fake.
Piechatzek quotes Lance Miller, CEO at Surex:
"AI won't replace companies, but companies that use AI will replace companies that don't."
The agencies finding their footing here aren't necessarily the ones with the most sophisticated tooling. They're the ones that have worked out what they're actually selling – and made sure it's something AI can't replicate on its own. The ones that haven't answered that question yet are going to find the next few years clarifying.
This article draws on insights shared in the TYPO3 Business Insights podcast. The full episode is available here, along with a complete English transcript available here.
