For most of my working life, looking good was a decent shortcut for being good. A smart website, a well-written brochure and a few glowing reviews told you somebody had made an effort. That shortcut has gone. When a machine can write anything, polish proves nothing, and the only thing left that can’t be conjured up in seconds is the real thing: trust and reputation.
01The shop window is free now
AI can write a convincing web page, a hundred product descriptions, a heartfelt “about us” and a stack of five-star reviews before the kettle boils. So can everybody else. When everyone can look professional at no cost, looking professional stops meaning anything.
That’s uncomfortable for anyone who has relied on presentation. But I think it’s good news for anyone who is actually good at what they do.
02Reputation is the one thing you can’t generate
Reputation doesn’t live on your website. It lives in other people: customers who come back, suppliers who vouch for you, the local paper that wrote about you, the neighbour who says “use them, they’re sound”. It builds slowly, it’s spread across many places, and it’s very hard to fake at scale.
In a small town everyone knew who did good work and who didn’t. Nobody needed a marketing department to find out. In a strange way, AI is taking us back there. The difference is that the person asking around is now a machine.
When everything can be faked, the only thing worth having is the thing that can’t.
03The machine is asking the same question we always did
Ask an AI assistant for a good accountant in Llandudno or a decent B&B near Snowdon and it doesn’t hand you ten links and wish you luck. It gives you an answer. To do that, it has to decide who to believe.
It does roughly what a sensible person would do. It looks for evidence, checks whether the evidence agrees with itself, and leans towards the businesses that lots of independent sources describe in the same way. It’s not judging how shiny you are. It’s judging how sure it can be about you.
04Don’t game it. Feed it.
There will always be people trying to manipulate the machines, just as there were people stuffing keywords into web pages twenty years ago. Fake reviews, invented credentials, pages written for robots rather than people. Some of it will work for a while. Then the systems get better at cross-checking, and the tricks become liabilities.
The better strategy is the boring, honest one: make your real reputation easy to see. The AI can only recommend what it can find and verify, so give it the tools. At Livetech we call this Trust Density: the amount of clear, repeated, real-world evidence AI can find about a business. It comes down to five signals:
None of that is a trick. It’s just the truth about your business, written down clearly in enough places that a machine can be confident about it.
05What it means for the rest of us
There’s a responsibility on the other side too. AI answers arrive fast and sound sure of themselves, and sounding sure isn’t the same as being right. If a decision matters, ask where the answer came from and look at the sources yourself. Treat the machine like a well-read friend: useful, often right, and never the last word.
I keep coming back to a simple idea. When decisions carry long-term consequences, decision quality matters more than speed. That’s true for the businesses deciding how to present themselves, and for the people, and now the machines, deciding who to believe.
The future doesn’t belong to whoever shouts loudest or polishes hardest. It belongs to the people who are genuinely good, and who make it easy for both humans and machines to see it.
Paul Levy is the founder of Livetech, a digital agency in Llandudno, and the creator of the Trust Density framework for how AI decides which businesses to recommend. The views are his own.
![]()