Looking after 300 websites with AI in the room

People sometimes ask how a small team looks after roughly 300 websites. The honest answer is years of good systems, a lot of coffee, and lately, AI.

We want to be clear about what we mean by that, though. AI hasn’t replaced anyone here. It’s more like the colleague who never complains about the boring jobs.

Finding what’s broken, faster

When a client emails “the homepage looks weird”, the fix is rarely the hard part. Working out why is. Was it a plugin update? A theme change? Something clashing with the caching? AI is very good at reading through error logs and comparing code, so we get to the cause a lot quicker. We still fix it ourselves. We just spend less time hunting.

Small changes without the fuss

So much website work is little custom stuff. An extra field at checkout, a layout tweak, a rule on a form. AI writes a first draft of the code, one of us checks it, we test it on a staging site, and only then does it go live. We add it in a way that survives updates, because we’ve inherited enough sites held together with fragile hacks to last us a lifetime.

Make the team better, not make the team smaller.

The jobs nobody gets round to

Missing alt text. Thin meta descriptions. Broken links. That page about your 2021 special that’s still live. None of it ever feels urgent, so on most sites it never happens. With AI we can draft fixes across hundreds of pages in a fraction of the time, and someone approves each batch before it goes anywhere.

Better audits

We can now look at a site’s speed, structure, accessibility and AI-readiness far more thoroughly than was practical before, and boil it down to a short list of what actually matters.

What we won’t hand over

Your brand and design decisions stay with people. Nothing AI produces goes live without a check. Client data is handled carefully and isn’t used to train anything. And you still talk to a real person who knows your site. Usually one of us.

For our clients, what this really means is that problems get fixed faster, and small improvements that used to be “not worth the hours” now get done. That’s what we think AI should do for a small business. Make the team better, not make the team smaller.