Most of the advice on AI for marketing is utter garbage.
There, I said it. We're drowning in a sea of generic blog posts written by content-mill AIs, promising a revolution that never seems to arrive for the average small-to-medium-sized enterprise. The tech bros in Silicon Valley talk about Large Language Models and neural networks, while the owner of a plumbing supplies firm in Bolton just wants to know if they'll make more money.
This is the great disconnect. I see it every day. After selling my first marketing agency for a sum that made my accountant smile, I’ve spent the past two years in the trenches with over 300 UK SMEs. I’m not just an investor; I’m a founder again, building agentic AI tools like Competable because I saw this gap firsthand. Clients don’t want AI. They don't care about your tech stack. They want outcomes. And after hundreds of brutally honest conversations, I can tell you they want three things above all else: settled revenue, de-risked growth, and a human to call when things get choppy.
The biggest lie the marketing industry ever told itself was that 'more leads' was the goal. For decades, agencies have sold this vanity metric, delivering a CSV file of questionable email addresses and patting themselves on the back. The client is then left with the actual hard work: converting those so-called leads into actual, spendable cash.
Our own data paints a grim picture. We analysed the sales funnel of a Coventry-based manufacturing client. They were proud of their £1.2M pipeline, generated by a previous agency. The problem? Their conversion rate from lead to sale was a miserable 8%. They weren't celebrating a pipeline; they were drowning in follow-up emails and wasted sales time. They had a lead problem, but it wasn't a lack of them.
This is the core philosophy behind agentic AI. It’s not about using AI to generate 10,000 more low-intent leads. It is about deploying an autonomous agent that intelligently qualifies, nurtures, and settles opportunities into your calendar. Think of it as a shift from 'cost per lead' to 'cost per settled pound'. Our own tool, AutoSettle, was born from this exact frustration. It’s an agent that doesn’t just identify a potential customer; it engages them, handles the initial back-and-forth, and works to place a confirmed, qualified meeting directly into a sales rep's diary.
For the SME owner, this is transformative. It removes the largest and most painful variable in the sales process. It delivers not just a possibility of revenue, but a tangible, scheduled conversation with a high-intent prospect. That’s not a lead. That’s a settled opportunity. And in 2026, it will be the only metric that matters.
The second thing SMEs crave is stability. The venture-capital-fuelled ‘growth at all costs’ mindset that infects so much of the tech world is utterly toxic to a bootstrapped, family-run business. These companies cannot afford to burn their marketing budget on a high-risk ‘moonshot’ campaign that a fresh-faced agency account manager thinks is a brilliant idea.
Yet, this is precisely what most agencies sell. The ‘big bang’ launch. The expensive rebrand. The all-or-nothing digital PR push. When it works, the agency principal gets a shiny case study for their website. When it fails, the client is left picking up the pieces, often with a hole in their finances that takes years to repair.
Here’s a contrarian take for you: the modern obsession with constant A/B testing is a trap. It encourages marketers to focus on localised, incremental gains — changing a button colour here, tweaking a headline there — while ignoring the big strategic picture. You can’t A/B test your way out of a flawed strategy. True, sustainable growth comes from making better macro-level bets.
This is where agentic AI provides the second piece of the puzzle. By using AI to constantly model the competitive landscape, we can identify low-risk, high-probability avenues for growth. Our platform, Competable, was designed for this. It ingests millions of data points about your competitors — their ad spend, their content strategy, their pricing, their hiring — and uses that to model what strategies are working in your specific market. It turns the entire market into a live laboratory, allowing you to learn from your competitors' expensive mistakes and high-risk experiments without spending a penny.
A great example of this thinking, albeit pre-agentic AI, is Gymshark. Ben Francis didn't build that empire by A/B testing checkout buttons. He made a huge strategic bet on influencer marketing and community before anyone else in his sector. It was a macro-level decision. Today, an agentic AI could have modelled the rising influence of fitness creators, quantified the engagement they were generating versus traditional ad channels, and flagged it as a high-probability, asymmetric bet. This is what de-risking growth looks like. It’s not about finding a silver bullet; it’s about using data to systematically load the dice in your favour. We even built an AI marketing ROI calculator to help clients begin to frame these decisions financially.