Most small business websites don't need a web developer

This is an odd thing to hear from someone who spends his week telling people to build on Astro or Next.js with Payload or Sanity behind it, but most small business websites don't need one of us. Every week a developer posts that Lovable or Replit is coming for their clients, and every week I want to reply that they should be. I'm tired of the charade. If a website only has to prove the business exists, the builder is fine.

What I mean by small

Not what the Australian government means. Fair Work says under 15 staff and the Bureau of Statistics says under 20. The Tax Office doesn't care how many staff you have, it goes off turnover. A plumber with one ute is small by every definition they've got, and so is a brand with twelve people that takes every order through its website, which is a strange pair to lump together.

The plumber needs a page with his number on it and enough to show he's real when someone looks him up at 9pm. If the site went down for a week he'd hear about it from a mate, if at all. The brand's customers log in and every order comes through the site, so when it breaks the business stops until someone fixes it.

Let the marketing person build it

If the site is a business card, the SEO guy or the part-time marketing person can sit down with Lovable, or whatever Wix has bolted on this year, and have something decent by the end of the afternoon. It'll be better than anything the business has had before. A developer would charge thousands for the same result, and an honest one would tell them to spend less and turn the job down.

None of this is new. Before the AI it was a Wix template, and before that it was a nephew who'd done a course.

The builder only knows what you told it

The plumber's site works because there's almost nothing to get wrong. Ask for a site for a plumbing business and you'll get one. Once the site has a job, the builder can only work with what it's given, and the person typing usually can't tell what's missing.

Tell it you want a booking form that drops straight into the calendar and it'll make you one. Nobody asks what happens when two people book the same slot, or when the calendar app changes something on its end, because you'd have to have built one before to think of it. Then it breaks, and the marketing person who built it can't fix it.

Then there's ownership. Wix and Squarespace don't hand you the code, and the ones that do hand it to someone who can't read it. Fine for the plumber, a real problem for the brand the day they want to leave.

It isn't a threat

Photoshop showed up in 1990 and the people retouching photos by hand had every reason to be worried. Thirty-odd years on nobody calls it a threat to photographers. Same thing is happening in every dev shop I know. Claude Code or Cursor is open all day and does a fair bit of the typing. It's a tool. Knowing what to build is still on us.

My clients are closer to the brand than the plumber. They care where the content lives and where it has to end up, and the person editing it every morning has to be able to do that without me. For them, building it badly costs a lot more than building it properly, so they aren't going to bet the business on Replit. If the plumber wants to and it isn't costing him anything, let him get on with it.