Try something before you read any further. Open an AI assistant, ChatGPT or whichever one you use, and ask it to recommend a business like yours in your city. Then see what it says.

If your business comes up, with the right details, great, skip to the end. If it doesn’t, or if the details are wrong or outdated, keep reading. That gap is what this piece is about.

The question changed before the website did

For years, getting found online meant ranking on a search results page. A list of blue links, competing for position, and a website built to win a click. Most founder-led businesses still build for that world, because it’s the world their website was built in.

Increasingly, that’s not how people are finding businesses. They’re asking a single question and getting a single answer, one business named, sometimes two, not ten. The competition isn’t for position on a page anymore. It’s for being the one thing an AI assistant decides to mention by name.

That’s a completely different kind of competition, and most websites haven’t caught up to it.

These systems don’t read the way people do

A person lands on your homepage and takes in a feeling: polish, colour, tone, a hero image that sets the mood. That still matters for the person who gets to your site. But the systems generating AI answers aren’t looking at your homepage the way a person does. They’re reading structure: specific pages that name specific services, specific outcomes, and specific locations, in plain language that doesn’t require guesswork.

A site that says “premium event solutions” and leaves it there gives one of these systems very little to work with. A site that names the service, the audience, and the outcome specifically gives the AI system something concrete to repeat. We’ve watched this play out directly with a client whose entire commercial conversation changed once their site started naming outcomes instead of just listing products. That same shift is exactly what determines whether an AI system can say anything useful about a business at all.

This is sometimes called generative engine optimisation, or answer engine optimisation. The acronyms matter less than the underlying shift: the readership of your website now includes machines deciding, on someone else’s behalf, whether you’re worth mentioning.

Nothing about your business changed. The ground did.

This is the uncomfortable part. A site that was performing perfectly well eighteen months ago can quietly lose ground without a single line of text changing, simply because the way people search shifted and the site didn’t move with this shift. It’s the same mechanism as a print ad that stops working, not because the ad got worse, but because the audience stopped looking in the direction of where it was placed.

A website is infrastructure, not a one-off purchase. Infrastructure that isn’t maintained doesn’t stay still. It silently falls behind the systems around it, while still looking exactly the same to you.

What this actually calls for

It’s not a redesign. Mostly, it’s specificity: pages that name what you do, who it’s for, and what changes for the client, in language a machine-based system can lift and repeat accurately. It’s consistency: the same facts about your business, stated the same way, across your site and the other places you’re listed. And it’s structure: content organised so a system can find the right answer to a specific question, rather than one long page trying to say everything at once.

If you’ve read this far, go back and try the test again with more specific language in mind. Search for exactly what you offer, not just your business name, and compare what comes back.

If the gap between what your business actually does and what an AI assistant currently says about it is wider than you’d like, that’s a clear gap. We’ll show you exactly what these systems currently say about your business, and what it would take to close that gap.