Why Most SMEs Shouldn't Start With AI

Every week another founder asks me the same question: “Where do we start with AI?” It’s the wrong first question — and asking it too early is the fastest way to waste money and lose trust in technology across your team.

The AI hype trap

AI is the most exciting technology of the decade, and also the most over-sold. Vendors promise transformation in a click. The reality for most small and mid-sized businesses is messier: the tools land on top of broken processes and disconnected data, and the “magic” never shows up.

The problem is rarely the model. It’s everything underneath it.

Why foundations matter more than models

A useful AI feature needs three things most SMEs haven’t sorted yet:

• Clean data it can actually read and trust
• Connected systems so information isn’t trapped in spreadsheets and inboxes
• Stable processes that are consistent enough to automate

Skip these and you get confident-sounding answers built on bad inputs — which is worse than no answer at all.

What to fix first

Before a single AI pilot, I look at the unglamorous layer:

1. Where does the same data get typed in twice?
2. Which decisions are slow because nobody can see the numbers?
3. What breaks every time a key person is on leave?

Fixing these is cheaper, faster, and delivers value on day one — and it’s exactly the groundwork that makes AI worth it later.

A better sequence

The order that actually works:

1. Clarity — understand the business and where the friction really is
2. Connect & clean — get systems and data talking
3. Automate — remove the repetitive manual work
4. Then AI — apply it where it creates real, measurable impact

Businesses don’t need more technology. They need clarity, the right priorities, and a partner who thinks like an owner — not a vendor.

Start with the foundation. The AI will be far more valuable when you get there — and you’ll actually trust what it tells you.