As AI forces agencies to question not only how they make the work but where their value really lies, few leaders are pretending to have all the answers. Leith Managing Partner Richard Marsham has been thinking about what that uncertainty means in practical terms: for the agency model, for the relationship between clients and their creative partners, and for the kind of thinking businesses will still be willing to pay for when execution becomes increasingly accessible.
Here, he considers why the future of agencies might depend less on the tools they use and more on the judgement, craft and strategic value they can bring to them.
I was reading recently about Amara's Law: we overestimate the impact of new technology in the short run and underestimate it in the long run. It's a simple idea. It's also stuck with me, because it names something about the moment we're in right now. Change is happening. We're continually reviewing our business model so we're fit for the future, but nobody has a settled view of what that future looks like.
What counts as short term here, and what counts as long term? I don't think anyone can answer that with confidence, and I'd be wary of anyone who claims they can. AI is easy to talk about in sweeping terms and much harder to plan around when you're the one deciding where to put your money and your people this quarter. Is the current disruption a phase we ride out, or the start of something that changes what an agency is for? Nobody knows yet. That's the honest position. The only thing we do know is that change is coming and it's coming fast.
The more useful question sits underneath: what will clients still want agencies for once AI lets them do so much themselves, and what will they actually pay for? That's the question that should shape the business, not a guess at how the next decade unfolds. Clients are getting more capable with these tools, more self-sufficient, more comfortable doing in-house what used to sit with agencies by default. That's not something to get defensive about. It's something to embrace and plan around.
For us, that means shifting Leith further towards a creative consultancy. We want our creative thinking applied to a wider range of business problems, not just campaigns and assets. Creativity at its best isn't a service line. It's a way of working through a problem. If clients can execute the basics themselves using AI, our value has to move upstream, into the thinking that shapes what gets made and why. That means sitting in conversations about strategy, positioning, customer behaviour and innovation, not just producing the output at the end.
And here's the part I keep circling back to. As AI tools get better and cheaper, we're heading into a flood of moderately good creative work. Producing something decent will get easy for almost anyone. That's useful in plenty of ways. It also means decent stops being worth anything as a differentiator. So the real question is how we use judgement and creative skill to pull work up from moderately good to genuinely excellent. That gap, between competent and brilliant, is where the value will sit.
Everyone will have access to the same AI tools. Having them won't be what wins. Knowing how to push them past their obvious output will. So will judgement about what's actually right for a brand, a moment, an audience. AI can generate ten options in seconds. It can't tell you which one is good, and it can't take a good option and make it exceptional. That's still a human job, built on taste and experience and a feel for what actually lands. And that's where we need to put our value. There will still be craft involved in using these tools, as there is in most skilled professions. And we need to link our remuneration away from hours spent on a task through to the real value that we add to a brand or business. Value-based pricing is another huge challenge we face in terms of how it works and how much clients will embrace it in a cut-throat agency world where there's always someone prepared to cut a deal.
I can't tell you whether we're in the short-term overestimation phase of AI or the long-term one where its effects finally show up properly. But I know what we're betting on. Judgement, creative craft, and strategic thinking get more valuable as AI pushes the baseline up and the average up with it. The tools will be shared by everyone. The agencies that lift those tools further than everyone else, that's who wins.
