Building an agency is a brutalist education in business. You learn about sales, operations, human psychology, and the crushing reality of cash flow, all at once. My journey with Creative Marketing Group has been exactly that. But if you asked me to do it all again in 2026, I wouldn’t build an agency. Not in the traditional sense.
The model is broken. It’s a relic of a pre-digital, pre-automation era. It’s built on a foundational lie: that the value you provide is proportionate to the hours you work. This is the billable hour, and it’s the original sin of the service industry. It rewards inefficiency, inflates headcount, and fundamentally misaligns the agency with the client. The client wants outcomes; the agency sells time. See the problem?
So, if I were forced to start over in 2026, my first principle would be to build an organisation that sells outcomes, not hours. An entity that looks less like an agency and more like a software company that happens to deliver marketing. Here’s what I’d do differently.
This is the root of all evil in the agency world. It’s a framework that encourages bloat and discourages true innovation. Your best-performing analyst who completes a task in one hour is, under this model, four times less valuable than the mediocre one who takes four. It’s insane. We’ve all seen it: the padded timesheets, the long-winded meetings, the endless reports that justify the retainer. It’s a dance of plausible deniability.
At Creative Marketing Group, moving away from this was painful. Clients are conditioned to ask "How many days do I get for my £10,000?" We had to re-educate them to ask, "What outcome do I get for my £10,000?" This meant defining outcomes with microscopic precision. Not "increase brand awareness," but "generate 150 qualified leads with a cost per acquisition below £85."
In 2026, I wouldn’t even let the billable hour get a foothold. The model would be value-based, tied to AI-driven performance. We’d use forecasting tools to model the likely outcome of a strategy, and we’d price based on a share of the upside. If our AI-driven system was predicted to generate £1M in new revenue for a client, we’d charge a percentage of that. Our success would be their success. No hours, no headcount, no fluff. Just results.
Your first ten hires define your company’s DNA. In a traditional agency, you hire account managers to handle clients, and maybe a few specialists (an SEO, a PPC manager). This is a recipe for building a slow-moving, human-dependent consultancy.
My first hire in 2026 would be a Python developer. My second would be a UX designer. My third would be a data scientist. I would be building a product, not a service team. The service is the product. The core mission would be to automate the work that traditional agencies pay people to do.
Why? Because the future of marketing isn’t about having the smartest Head of Performance; it’s about having the most effective automation stack. It’s about building proprietary tools that give you an unfair advantage. Your value is no longer in your people’s time, but in your system’s efficiency.
Instead of a ‘Head of Social’, I’d hire an ‘Automations Lead’ whose entire job is to identify and eliminate manual tasks within our marketing workflows. Their bible would be the API documentation for platforms like Meta, Google, and TikTok. Their team’s output wouldn’t be client reports; it would be internal software that gets results faster, cheaper, and more reliably than a human team ever could.