Execution is too cheap to sell: an agency crisis.

Six years on from writing this original article, the same argument is being made again, just with a different trigger.

Through 2025 the holding companies went through very public restructuring: WPP’s leadership changes and Omnicom’s merger with IPG both landed against a backdrop of one shared pressure, proving the traditional agency model can still justify itself when AI tools have made so much of the actual production cheap and fast. In-housing, already a decade-long trend before Generative AI arrived, has accelerated further as AI tooling lets client marketing teams produce more of their own content without an agency’s full execution machine behind them. Alongside all this, a new wave of AI-native and AI-augmented agencies has emerged, built from the ground up around these tools rather than bolting them onto an existing studio model.

So what does this mean for the industry? We take another look, six years and one AI revolution on from when we first asked.

The cause of this shift in agency model has been driven by the demands of clients. While the importance of quality creative and brand presence continues to be paramount, the increase in the number of platforms and iterations for personalisation seems to be directionally proportional. Agencies have to deliver a production line of quality creative, the likes of which the industry has never seen before. TSB Chief Marketing Officer Peter Markey says:

“Keeping in tune with a client’s changing needs means adopting the mindset of a management consultancy, identifying client needs, providing solutions to fix problems and specialist skills where needed. Producing such a breadth of content requires much tighter control over a brand, requiring agencies to have a much broader understanding of the client’s wider business goals, ultimately having to advise on approach and delivery.”

But there’s a second, more structural force at work now, separate from the sheer proliferation of content. Generative AI is doing to production and execution what production and execution used to do for creative concepts: making them fast, cheap and abundant. A shot list that took a studio days can be drafted and roughed out in minutes. Copy variations for a hundred audience segments can be generated before lunch. That’s a genuine gift for speed, and it’s exactly the part of the agency model that AI is best at replacing.

What it can’t replace is the judgement that decides which of those hundred variations is actually right, for this client, this audience, this moment, and why. As execution gets commoditised, the value in the agency relationship gets pushed further up the chain, into strategy, into the client relationship, into the accumulated understanding of a business that lets someone say no to ninety-nine AI-generated options and yes to the one that matters. That’s not a hypothetical shift. It’s exactly the consultancy positioning we describe below, and it’s the reason the argument in this piece is more true in 2026 than when we first made it.

According to the Cambridge Dictionary, the word ‘agency is defined as ‘a business that provides a service to other people or organisations.’ There has always been a wide range of models for interpreting this. At Studiomade, for example, we have always filled more of a consultancy role for our clients, this is where we believe our value to be, offering insight and guidance to help resolve a commercial challenge. We always ask the driving questions behind a problem to ensure that a meaningful result is delivered, no matter the output.

This isn’t a hypothetical position for us, either. Broadridge, Motics, City & Guilds and Confluence are among the clients we’ve worked with recently in exactly this capacity: brought in for strategic thinking and judgement, not just production.

As a business, our move towards this model has been a result of both client needs and our own experience growth in identifying opportunities and solutions at a strategic level. In a climate where delivering impact and measuring a return on investment is vital, we believe that the agency model must alter.

So what does the future of an agency look like?

In an increasingly competitive and AI-saturated landscape, there’s no doubt that there is a large opportunity for clever, inspiring agencies to exist. As Elmwood identify, ‘It’s no longer good enough to be the best of the best, you have to be the only ones who do what you do.’

Accenture Song is the clearest large-scale proof that this pivot is possible, not just theoretical. Built explicitly as a strategy-and-creative consultancy rather than a production shop, it has grown into one of the biggest names in the industry precisely by selling judgement and business outcomes rather than execution alone. It won’t be the only model that works, but it settles the question the industry was still asking in 2019: agencies that make the pivot to strategic partner don’t just survive the disruption, they can end up leading the market that disruption creates.

Originally published 20 November 2019 · Updated for 2026