TL;DR: Skip the demo and ask five questions — what problem it solves, where your data goes, who stays in control, what it really costs, and whether you can leave. A tool that answers all five cleanly is worth a look; one that dodges any of them is selling hype, not help.
If you’re trying to choose an AI tool for your business, here’s the honest shortcut: ignore the slick demo and ask five plain questions — what specific problem does it solve, where does my data go, does a human stay in control, what does it really cost, and can I leave without losing everything? A tool that answers all five cleanly is worth a look. A tool that dodges any of them is selling you the hype, not the help.
Why the hype works on smart people
The AI hype machine is good. The demos are dazzling, the language is confident, and there’s a real fear of being left behind. Smart, busy business owners fall for it not because they’re naive, but because they don’t have time to look under the bonnet — and the vendors count on that. So here’s the look-under-the-bonnet checklist I’d use before spending a cent.
The five questions
1. What exact problem does this solve for me? Not “it uses AI”. What task does it take off my plate, and is that a task I actually have? If you can’t name the problem in one sentence, you’re buying a solution looking for a problem. The best tool for a job you don’t have is no tool.
2. Where does my data go? This is the one most people skip and most regret. Does your customer data leave your control? Where is it stored? Is it used to train someone else’s model? If the answer is vague, treat it as a no — especially if you handle anything sensitive.
3. Does a human stay in control of anything that matters? A good tool keeps a person in the loop on decisions, money, and anything customer-facing that could go wrong. A tool that wants to act fully autonomously on important things is a risk dressed as a convenience.
4. What does it really cost — including the hidden bits? The headline price is rarely the real one. Ask about per-use charges, the cost as you grow, and the time your team will spend setting it up and babysitting it. A “cheap” tool that eats ten hours a month isn’t cheap.
5. Can I leave? If you walked away in a year, do you keep your data and your workflow, or are you stuck? Tools designed to make you dependent forever are the ones to avoid. The right tool leaves you more capable, not more trapped.
The question isn’t “is this AI impressive?” It’s “does this make my team more capable, with my data safe and my hand still on the wheel?”
The opinion part — because the opinion is the moat
Here’s what fifteen years in Australian tech taught me: most businesses need less tooling than they’re sold, not more. The right move is usually one well-chosen tool that solves one real problem, run by people who understand it — not a stack of subscriptions nobody fully uses. If a tool makes you feel like you should understand it but don’t, that confusion is a feature of the sale, not a failing of yours.
And be sceptical of urgency. “Everyone’s doing this, you’ll fall behind” is a sales tactic, not advice. The businesses that win with AI aren’t the ones who adopted fastest — they’re the ones who adopted deliberately.
What I’d do
Write your five answers down before any demo. Make the vendor fill the gaps in plain English. If they can’t, that’s your answer. Choosing well isn’t about being an AI expert — it’s about refusing to be rushed past the basic questions.
If you want a straight, no-pitch second opinion on a tool you’re weighing up — including an honest “you don’t need this” — that’s exactly the kind of thing worth asking a real person.