Copy article

Claude will build me a website

ended 11. August 2026

Looking for non-technical business owners who have tried to use AI to build them a new website, create a marketing strategy or do some other clever commercial back end stuff for their business  — but then, when the wheels have come off or it doesn't work, have ended up doing it themselves or shelling out some proper coin for a human to do it after all. Translate: AI can do come things brilliantly, but are we being complacent to believe it can do complex things without risk or opportunity cost? Any thoughts, by tomorrow COB please. And if AI has been brilliant and turned your business into a money-printing machine, send them across, too.

4 responses from the Newspage community

Copy all

Star Quote
Copy

Claude can absolutely build a website. For a small or new business where every penny counts, it can be brilliant if the owner will learn, test and fix things. GoDaddy and Squarespace have always offered this trade: spend time instead of paying a specialist. AI makes it easier.

The real cost is often verification, not generation. Experienced developers know what to reject, test and rewrite. We have seen vibe-coded sites damage SEO and drift off-brand with each new page.

The same applies to design. AI can execute a creative direction, but if also asked to invent that direction, it defaults to the same layouts and trends as everyone else. A human must understand the audience, design something distinctive and maintain the system consistently.

AI excels when the task is defined and a specialist is in the driving seat. It is less dependable when asked to discover requirements and make commercial trade-offs.

Use AI to accelerate expertise, not pretend expertise is unnecessary.
Copy

I am not a technical person. I built ValuQ, a UK property platform, and I have leaned on AI across the business from the start. So I will answer this from the inside, not the sidelines.

AI is genuinely brilliant at the first 80%. The draft, the layout, the copy, the boilerplate that used to mean a week and a freelancer's invoice, it does in an afternoon. For a non-technical founder that is real leverage.

The catch is the last 20%, and that is the part that carries the risk. AI is fluent and confident even when it is wrong. If you cannot spot the difference, the opportunity cost is not the hour you saved. It is the mistake you shipped without knowing.

That lesson runs so deep that we took AI out of the one thing that matters most in our business: valuing someone's home. AI can build the shop. It should not be trusted to price the house. Use it for the scaffolding. Keep a human on anything where being confidently wrong costs someone real money.
Copy

For anyone starting out the saving is real and it's big. A website, a logo, the first marketing copy: that used to be a founder's opening cheque, and it's now close to free. I see people trading within weeks who would once have spent a year saving up to begin. But AI is cheap to start and expensive to abandon, and the reason is timing. You rarely find out it hasn't worked until you've built the business around it, and by then a developer isn't starting fresh, they're unpicking what's already there. That costs more than doing it properly at the outset would have. So use it for the cheap, repeatable work, bank the saving, and pay a human for the thing the business actually runs on.
Copy

Claude and other AI is great at strategy and copy, if you supervise it well. Left to its own devices it sounds like it thinks it should which is the same as everyone else.

For AI to have soul you have to be in tune with yours and what really matters to you. Beyond the profit and KPIs. Why are you building this strategy or website? That's the question which frequently remains unanswered but is at the core of success. Know yourself and tell the AI, great things can happen very quickly.