AI isn't what's going to kill your business.

This isn't a prediction. It's what I see working with companies across every sector, on every kind of software. Always the same pattern, and always the same mistaken diagnosis.

It's not AI. It's the one across the street.

Technology has never destroyed a business. It raises the level everyone plays at. What puts you in difficulty isn't the tool, it's someone else who picked it up before you and now runs on costs you can no longer match.

Every company is chasing the same thing: lower costs. That's the heart of the matter, and always has been. You cut a budget, trim an expense, renegotiate with a supplier. What's changed is that an enormous lever has just opened up, and most companies haven't touched it yet. Repetitive tasks, done by hand, every day, for years, that could take ten times less time.

The competitor who worries you in two years won't be better than you at your trade. He'll simply have removed half his friction.

One word before going further, and I mean it. If your business suits you as it is, if you have no growth plans and organisation isn't a concern, close this page. You won't miss anything. This isn't a provocation and there's no judgement in it: everything that follows only serves people who want to grow. For everyone else, it's noise.

Excel never killed anyone

There used to be a job called computer, in the original sense: people whose work was to run calculations by hand, in columns, all day long. It was a real job, with real skills and capable people in it.

Then the spreadsheet arrived. The job disappeared within a few years. The people didn't. They became accountants, analysts, controllers. They did the same work better, on volumes nobody could have handled by hand. The value wasn't destroyed. It moved towards what needs a brain.

That's exactly what's happening today, faster and across far more trades at once. Here's an example everyone knows. Some medical practices still employ someone to copy out reports and prescriptions by hand. In 2026. When you can dictate out loud and have the note write itself, cleanly, in seconds.

That person doesn't disappear. She welcomes patients, runs the practice, handles emergencies, does the work a human is irreplaceable for. Nobody took her job away. They took away the mindless part of it.

Will some roles disappear along the way? Yes. I'm not going to claim otherwise to make this sound nicer. But that's how the world has always moved, and nobody has ever managed to put it on pause.

Why chains win

Look at why chains are spreading everywhere, in food, in services, in retail. It's not because their product is better. Often it's worse. It's because every one of their costs is divided.

One piece of software for a hundred locations. One website. One marketing operation. The same processes, written once, applied everywhere. A new hire is trained in three days because everything is documented. That's the real strength of a chain, and it has nothing to do with the quality of what it sells.

The key point, the one that changes everything, is that this advantage is no longer reserved for the big players. A five-person company can now run on operating costs comparable to a chain's. Tools that cost hundreds of thousands ten years ago are within a small company's budget.

Which means two things. First, you can finally fight the big players on equal terms. Second, less pleasantly: the newcomer who opens across the street won't have your twenty years of trade, but he won't have your twenty years of habits either. He'll start clean. No double entry, no lost file, no three hours a week on a schedule. And on the same margin as you, he'll be able to sell cheaper or serve better.

AI can't do anything with a messy drawer

People talk to me about artificial intelligence as if it were a switch. Flip it and the business runs itself. That's not remotely how it works.

AI needs material. Your data, your rules, your way of working. If your information sits on paper, spread across three Excel files that don't talk to each other, inside two people's heads and in a third person's inbox, there's nothing to automate. There's nothing even to plug into.

The starting point isn't the tool. It's knowing precisely how you work.

So you begin with the thankless part. Writing down what you do, in what order, under what rules. Bringing the information into one place, in a clean format. Naming the steps. It doesn't show, nobody finds it exciting, and it's what decides everything else. Without it no automation will hold, because there'll be nothing solid to automate.

The good news is you don't do this in one block. You take one task. Just one. The one that annoys you most, the one you keep pushing to Friday evening. You tidy it, you simplify it, you give it a tool. Then you move to the next. It's common sense: build only what's useful, and judge it on the result.

The software you rent isn't yours

Let's say you get organised. Good. That leaves a question almost nobody asks: on top of what?

Take a scheduling tool like Agendrix. It's a good product, and I'm not saying otherwise. But your schedules, your hours, your employees, your history all live at someone else's place. You're patiently building your organisation on ground that isn't yours.

People always give me the same answer: I can export. On paper, yes. In practice you export what the vendor decided to hand back, in their format, with their gaps. Rarely what you'd actually need, and almost never in a reusable state. I've run that exercise often enough to know how it ends.

The real wall comes later. The day you want to cross-reference your schedules with your sales. The day you want to connect two tools together. The day you want the whole thing analysed to understand where your money goes. That's when you find out you can't. You're held by the limits of the tool, and those limits won't move because you asked.

Subscription software is fast food. That's not an insult, it's its very nature: standardised, designed for the largest number, and therefore designed for nobody in particular. For plenty of companies that's more than enough, and good. The signal that it's no longer enough is the patches: the Excel files sitting next to the software to compensate for what it can't do.

What has changed is the price of the alternative. Custom software used to be out of reach. It isn't anymore, and over three years it often costs less than the pile of subscriptions it replaces. I wrote a whole article about it, with the numbers.

You don't swim against the wave

You can decide none of this concerns you. That's a perfectly respectable position, and I know very good people who hold it. They like working by hand, they want to keep things craft-made, they have no desire to file their business into boxes. I understand them.

But let's call it what it is: a choice, not a protection. Going against the current has never worked for anyone. Not for the human calculators, not for travel agencies, not for video rental shops. The wave asks nobody's permission.

What does work is catching it early and small. One task. One tool. Then another. You measure what it returns, and you keep going. Two years later you have a business that runs without you holding it up, costs that no longer depend on the hours you pour in, and something with real value the day someone wants to buy it.

So no, AI isn't going to destroy your business. It's simply going to make visible the gap between those who got organised and those who waited. If you want to work out what, in your case, should be tidied first, write to me. We look at what's costing you time, and we start there.