The agency of tomorrow runs on a different operating system

Most of the AI talk in our industry is about output: faster copy, cheaper production. That's the easy part. What I find far more interesting is what happens to the model underneath. How do you still deliver value as an agency, and how do you organise yourself?

For years, agencies sold time. More work meant more people and more hours. That model is tipping over. Work that used to take days now takes hours. Agents pick up the routine tasks, and a small team delivers what used to take a whole department.

So clients start asking something different. They care less about how many hours it takes and more about what it gets them. An agency that keeps billing by the hour ends up competing against its own efficiency.

The first instinct makes sense. You keep the tools you have and add AI on top, like a summary button in the CRM or an assistant in the planning tool. Useful, but the foundation is still built for the old way of working.

The real problem is fragmentation. Planning, sales, projects and client knowledge all live in separate systems, each holding its own piece of the truth. AI only ever sees part of the picture, so it can never really think along about the business.

Where I think this is heading is agencies where people and agents work from one shared base. Information is there when you need it, and planning moves with whatever is happening. AI stops being a feature and becomes the way the company runs. That takes something the market barely offers yet: an AI-driven operating system for the organisation itself, one layer where everything comes together.

This goes beyond agencies. I think we're at the start of a new era, the era of customized products. For years, off-the-shelf SaaS was the sensible choice because custom work was too expensive and too slow. You bent your processes to fit the software, and that was just how it went.

AI flips that maths. Building software that fits exactly how an organisation works has never been faster or more affordable. So why keep paying for ten subscriptions that each half-solve a piece of your work? For a lot of organisations, custom will become the default, simply because it gives you an edge.

For agencies, that's uncomfortable. Processes that grew up around old tools need rethinking from scratch, and digitising them as they are won't do it. That only works if your information lives in one place, otherwise AI has nothing to think with. And you have to go through it yourself. You can only advise a client properly on AI once you've rolled it out in your own organisation, awkward parts included.

Within United Playgrounds we went looking for a platform that supports this new model. We couldn't find one, so we started building it. Over the coming months our labels will move onto that AI-driven operating system of our own, and we'll drop a good chunk of the SaaS software we've run on for years.

Building is fun, but that's not why we're doing it. We believe this is where the whole industry is going, and we'd rather take that step now than in three years as a follower. That comes with lessons, including the hard ones. We take them straight into our client work. What we build for ourselves, we build for them too: customized products that fit how they work.

Sure, AI helps you write copy faster. Look only at that, though, and you miss the bigger shift: what your agency runs on underneath. We think more and more agencies and service businesses will face that choice. Facing it yourself, or already in the middle of it? I'd love to talk it through: eelco@livewallgroup.com.