AI consultancy for SMEs: from advice to working systems

Most AI projects at SMEs end up in the same place: a polished presentation. An advisory firm delivers a strategy, a roadmap and a stack of slides, and after that little happens. The problem is rarely the technology, it’s the gap between advice and execution. Research by RAND shows that more than 80% of AI projects fail, twice the rate of regular IT projects, and that the main cause lies with leadership and execution, not with the model. That’s why I approach AI consultancy at ClickForest differently: no report, but working systems you manage yourself afterwards. In this article you’ll read what AI consultancy concretely means for an SME, why so many projects stall, and how to make it pay off.

What is AI consultancy and what do you really get from it?

AI consultancy is guidance to deploy AI purposefully in your business, from choosing the right use case to measuring the result. At ClickForest it doesn’t stop at advice. AI consultancy here means the solution is also actually built and implemented, so an SME ends up not with a plan but with a working system.

The difference is where the project ends. A classic advisory project gives you insight and a direction. Valuable, but if no one builds it afterwards, it stays good intentions. An implementation partner closes that gap itself. If you want the broader context around AI in your business, you’ll find it on AI for growth and in my article on how AI transforms online marketing.

Why do so many AI projects stall at a report?

The core isn’t the technology but the execution. RAND found that more than 80% of AI projects fail, and MIT that 95% of generative AI pilots deliver no measurable return. Each time it trips over the same thing: a plan no one puts into production, a use case with no owner, or execution dropped entirely on the IT department.

The figures all point in the same direction, even though each measures something different. Gartner forecast that at least 30% of generative AI projects would be abandoned after the proof-of-concept stage by the end of 2025. A survey by S&P Global Market Intelligence found that 42% of companies halted most of their AI initiatives in 2025, compared with 17% a year earlier. And McKinsey reported that more than 80% of organisations see no tangible bottom-line (EBIT) impact from generative AI.

Tellingly, the cause is almost never the model. It’s how the project is steered.

“Many companies’ instinct is to delegate implementation to the IT or digital department, but over and over again, this turns out to be a recipe for failure.”

— Alexander Sukharevsky, global coleader of QuantumBlack (McKinsey)

An honest baseline measurement stops you from falling into that same trap. That’s what an AI audit is for, or more broadly an audit of where you stand today, before a single line of code is written.

AI strategy or implementation first?

Both, but linked. A strategy without execution changes nothing, and building without direction wastes budget. McKinsey found that of all the factors studied, the redesign of workflows correlates most strongly with real bottom-line impact from AI. So the gain isn’t in the model, but in how you weave it into your daily workflow.

That doesn’t mean you have to plan for months. Setting a short direction and immediately implementing one process for real delivers something faster than an elaborate strategy project that never touches practice.

“The redesign of workflows has the biggest effect on an organization’s ability to see EBIT impact from its use of gen AI.”

— McKinsey, The state of AI (2025)

How I approach AI consultancy

In four steps: first an audit of your processes and data, then prioritising on return, then building, and finally handing over. ClickForest works founder-led: you deal directly with me, not with an account manager. The building is done with Claude Code and, where needed, Make.com, always with the human in the loop, so you keep control afterwards.

That order is deliberate. First look at where the hours and errors disappear, then pick the process with the best ratio of effort to result, and only then build. Not everything at once, but one pain point solved well. That matches what the researchers behind the MIT report saw at the companies that did succeed.

“Some large companies’ pilots and younger startups are really excelling with generative AI. It’s because they pick one pain point, execute well, and partner smartly with companies who use their tools.”

— Aditya Challapally, lead author of MIT Project NANDA

That partnering is no detail. According to that same MIT research, projects that combine internal knowledge with an external partner reach a success rate of 67%, compared with roughly one third for what’s built purely internally or by IT alone. How I do that in practice is on building with AI, and why the human stays in the loop is in AI copywriting versus the human.

Curious which AI process would pay off fastest for you? Discover my full approach on AI for growth or schedule an introduction via the contact page.

What does AI consultancy concretely deliver for an SME?

Working systems that save time and reduce errors. Research by the Federal Reserve of St. Louis found that a fifth of frequent AI users save four hours or more per week, and for daily users that ran even higher. That’s time saved per employee, not a company-wide profit guarantee, but it shows where the value sits: less repetitive work.

Which system delivers the most depends on your business. These are the building blocks I most often set up for SMEs:

• AI agents that handle multiple steps themselves, from lead follow-up to data processing. You’ll find more context in my piece on AI agents in marketing.
• Marketing automation that connects your tools and eliminates repetitive work, often via Make.com.
• Content production for scalable SEO content and social posts while keeping your own voice.
• AI chatbot that handles first-line customer questions, complementing the approach in AI chatbot for customer service.
• Building with AI: custom software and internal tools, built with Claude Code instead of expensive off-the-shelf packages.
• AI training so your team learns to manage and extend the systems itself.

Which SMEs is AI consultancy worthwhile for, and which not?

AI consultancy pays off as soon as you have repeatable processes where time or errors are lost, and you’re willing to rethink those processes along the way. It’s less worthwhile if you’re looking for a magic bullet that delivers revenue without any change, or if you don’t have a single process structured yet today.

Being honest is part of it. Not every business is ready, and not every task is worth automating. Sometimes the answer after an audit is that you’d be better off getting your data or your process in order first. I’ll give that advice too, because an AI system on a messy process just makes the chaos faster.

“After last year’s hype, executives are impatient to see returns on their GenAI investments, yet organizations are struggling to prove and realize value.”

— Rita Sallam, Distinguished VP Analyst, Gartner

How do you choose an AI partner for your business?

Choose a partner who builds, not just advises. Watch three things: do they deliver a working system or only a report, does the solution stay your property without lock-in, and do they keep the human in the loop. An SME gets more from someone who works directly alongside you than from a layered structure of account managers.

The context helps frame that choice. According to the Belgian FPS Economy, more than a third of Belgian companies now use at least one AI technology, but among small companies that’s markedly lower than among large ones. That’s exactly where the opportunity lies: an SME that implements AI thoughtfully now builds a lead while most peers are still hesitating. ClickForest guides that step for SMEs from the Mechelen and Antwerp region, founder-led and implementation-first.

Conclusion: choose execution, not a report

AI consultancy doesn’t fail for lack of good ideas, but for lack of execution. The figures from RAND, MIT, Gartner and McKinsey all point to the same conclusion: strategy without implementation changes nothing. For an SME the smartest approach is to start small with one validated use case, actually build it, and then take the system into your own hands. No slides gathering dust, but a working system that saves time every week. That’s how I see AI consultancy, and that’s how I deliver it.