It raises the question of whether the whole marketing operating model is still fit for purpose. In an AI-enabled environment, the real work isn't choosing between in-house and outsourced, or between creative, media, content, technology and production partners. It's designing a system where all those capabilities create value together, without confusion or uncontrolled risk.

That's where the conversation has shifted. Clients don't only need agencies that respond well to briefs. They need clarity: what sits inside the marketing department, what stays external, what gets automated and how decisions, data, governance and accountability flow across the whole ecosystem.

A "big brief" used to be a communications challenge. Now it's often an operating model ask. Marketing teams want faster content, sharper performance media, more integrated customer journeys, lower production costs and smarter use of AI and they're asking their agency partners to help model all of it. That's a major shift in the agency-client relationship and in the commercial model underneath it.

Scratch the surface and what's revealed is rarely about AI. It's unclear decision rights. Capability gaps on both sides. Slow approvals. Fragmented data. Duplicated partner roles. Under-used technology. A procurement model that still rewards activity over outcomes.

The agency relationship becomes the visible symptom of a deeper design problem. AI is merely exposing these issues. What looks like an agency challenge is usually an alignment challenge between strategy, people, process, systems and governance.

The Traps AI Sets

AI won't remove the need for external partners, but it will remove the tolerance for vague roles, inflated teams and work priced as though nothing has changed.

Clients and agencies need to agree, together, what gets automated, what gets augmented, what moves in-house, where external expertise still earns its keep and how efficiencies should show up in fees, as well as who owns data protection, brand governance, IP, approvals and the quality of AI-assisted output.

The danger is treating AI as a cost-cutting exercise. That's the fastest route to thinner thinking, more content and less impact. The opportunity is to use AI to redesign the operating model around higher-value human contribution: sharper strategy, deeper decision-making, faster learning loops, stronger collaboration, clearer accountability.

The Marketing Department Has to Change 

You can't redesign the agency model and leave the client side untouched. If AI changes how agencies work, it changes what marketers need to lead and govern.

Expect fewer traditional coordination roles and more strength in commercial translation, data-informed decision-making, partner governance, customer understanding and change leadership. Middle managers will absorb most of the pressure, sitting between executive ambition and daily reality, expected to translate strategy, guide teams, manage partners and decide faster with less certainty than ever.

The answer is deliberate operating model design, backed by coaching, change management and genuinely practical ways of working. People need to understand not just what's changing, but why, what's expected of them and how success will be measured.

A Prediction, For What It's Worth

Marketing departments will become more focused on what customers really want, on overseeing automation and on building great partnerships. Agencies will lean harder into groundbreaking ideas and less into execution. As an industry, expect fewer suppliers, far more technology and much greater integration across the ecosystem. Success will depend less on managing a stable of specialist partners and more on orchestrating a connected system of capabilities.

Beyond the Recommendation

This is a performance system question and most organisations stop at the recommendation stage, when the real work is in the operationalising: designing the future-state model, aligning leadership around it, clarifying decision rights and governance, building capability, driving adoption and measuring whether it's actually delivering value. Strategy without execution is just an expensive opinion.

The outcome worth aiming for isn't an "AI-enabled marketing model" and a fresh organogram. It's a practical blueprint: the future agency ecosystem, the internal capability required, the split between in-house, embedded and external expertise, the governance AI use demands, the operating rhythms, the capability gaps and the measures that prove the model is working.

Most importantly, it gives agencies and clients something they usually lack, which is a shared operating language with less territorial protection and fewer hidden assumptions. That makes for faster decisions and better work.

The question keeping marketing leaders awake isn't whether they have the right agency. It's whether they've built an operating system capable of extracting value from AI, technology, data and partners.

Every organisation will have access to the same tools. The winners will be the ones who redesign how strategy, people, process, systems and partners work together.

For more information, visit www.euphoria.co.za. You can also follow Euphoria Telecom on Facebook, X, or on Instagram.

*Image courtesy of contributor