The conversation usually starts the same way. A business owner — sharp, driven, clearly not someone who avoids hard problems — tells me they want to “get into AI.” Sometimes they mean they want a strategy. Sometimes they mean they tried ChatGPT once and weren’t sure what to make of it. Sometimes they mean they’ve been reading about competitors and they’re quietly worried they’re falling behind.

What they almost never mean is that they’re two tools away from running a fundamentally different business. But in most cases, that’s exactly what’s true.

I’ve spent years working with SMEs across a lot of different sectors. The pattern I keep seeing isn’t that businesses lack ambition, or resources, or even understanding. The pattern is that they’re trying to solve problems at the wrong level of abstraction. They’re either attempting to automate everything at once — which is expensive, slow, and almost always overly complicated — or they’re experimenting randomly with individual tools without a clear sense of what problem they’re actually trying to solve.

The businesses that move fastest aren’t the ones with the biggest budgets. They’re the ones that pick one painful problem and solve it completely before moving to the next.

The “two tools” framing comes from something I started noticing in my own work. When I do an initial discovery session with a new client, I’m looking for two things: the task that’s eating the most time, and the decision that’s creating the most friction. Almost always, both of those have an AI-based solution that’s affordable, implementable within weeks, and measurably better than what they’re doing now.

What “two tools” actually means in practice

The first tool is usually about information processing. Summarising, extracting, categorising, drafting. Most businesses have processes where a human is spending time reading something and then creating something else based on what they read. Client emails that need to become briefs. Invoices that need to become reports. Meeting notes that need to become actions. This is where AI replaces genuine cognitive effort with something faster and usually more consistent.

The second tool tends to be about workflow triggers. Something happens, and something else needs to happen because of it. A form gets submitted, a sale is made, a document is uploaded. Connecting those trigger points through automation — removing the human whose only job was to spot that it had happened and tell someone else — is often where the biggest efficiency gains are.

Two tools. One to process, one to connect. The specifics change with every business, but the structure is almost always the same.

The mistake most businesses make first

They start with the most visible problem rather than the most painful one. Visibility and pain aren’t the same thing. The task that looks messy isn’t always the task that’s actually costing the most. I’ve worked with businesses that wanted to automate their social media — which is visible and feels strategic — when the thing genuinely bleeding their capacity was manual data entry that nobody was even tracking as a cost.

The way I approach this is simple. Before recommending any tool, I ask the team: if you could wave a magic wand and never have to do one task again, what would it be? Then I ask: what’s the thing that most often causes delays, mistakes, or rework? The intersection of those two answers is usually where we start.

It’s not glamorous. It doesn’t require a six-month transformation roadmap. But it works — and it works fast enough that by the time most companies are still planning their AI strategy, my clients have already reclaimed weeks of collective time and redirected it into growth.

If you’re curious where the two tools would sit in your business, that’s exactly the kind of thing we dig into during an initial consultation. The answer is usually clearer than you’d expect.