Can AI Actually Read Your Website? The Design Choices That Decide Whether You Get Recommended
By Andrew Fenwick 5 November 2026 News
Your website might look brilliant to the people who visit it and still be half blank to the AI tools your customers are now asking for recommendations. An AI-friendly website is not a different kind of website. It is one where the words that sell your business are actually readable by the machines that summarise it.
We have written a lot recently about the content and trust signals that get a business recommended by ChatGPT, Gemini and Google’s AI Mode. This post is about something more basic, and more often overlooked: can those tools read your site in the first place?
The Crawler Problem Nobody Designs For
When a person visits your website, their browser downloads the page, runs the JavaScript, loads the sliders, tabs and pop-ups, and paints the finished design. Many AI crawlers do not do all of that.
Research from Vercel’s study of AI crawler behaviour found that none of the major AI crawlers it monitored rendered JavaScript, including OpenAI’s GPTBot and OAI-SearchBot, Anthropic’s ClaudeBot and PerplexityBot. They fetched the raw HTML and moved on. Google’s Gemini was the exception, because it uses Googlebot’s infrastructure, which does render pages.
That study is from late 2024 and crawlers will keep improving. But the safe assumption for any business is simple: if an important sentence only appears after JavaScript runs, some AI tools will never see it. Building an AI-friendly website starts with that assumption. It is a different problem from blocking AI crawlers altogether, which we covered in the setting on your website that can block ChatGPT. Access gets the crawler through the door. Readability decides what it finds once it is inside.
Five Design Choices That Hide Your Content From AI
None of these are mistakes in the traditional sense. Most of them were perfectly reasonable design decisions a few years ago. They just were not made with AI readers in mind.
- Content loaded after the page. Reviews widgets, “load more” buttons, tabs that fetch their content on click, and some page builder effects inject text after the initial load. Accordions and tabs are fine, as long as the text is already in the page’s HTML.
- Words baked into images. Hero banners with the headline in the graphic, service lists designed as a single image, prices in a JPEG. A person reads them instantly. A crawler reads nothing, unless the alt text happens to repeat it.
- Key details locked in PDFs. Price lists, brochures and specification sheets are often the most useful content a business has, and they are frequently only available as downloads.
- Vague headings. “Welcome”, “Our Story” and “Solutions” tell a machine nothing. AI tools lean heavily on headings to work out what each section of a page is about.
- One page doing everything. If six services are squeezed onto your homepage, no single page gives a clear, quotable answer about any of them.
Structured Data: Labelling Your Content for Machines
Structured data, often called schema markup, is a small block of code that tells search engines and AI tools exactly what is on a page: this is a business, this is its address, these are its opening hours, this is a service it offers, this is a review. Google’s structured data documentation explains how it helps search engines understand content.
It matters for AI too. Microsoft’s Fabrice Canel has said that schema markup helps Microsoft’s large language models understand content, as reported by Search Engine Roundtable. For most small businesses, the essentials are Organization or LocalBusiness markup, Service markup for key services, and FAQ content that matches what customers actually ask.
Schema is one of the quickest wins on the road to an AI-friendly website. It is also one of the best defences against AI getting your details wrong. If your address, phone number and opening hours are clearly labelled and consistent everywhere, there is far less room for guesswork. We explained why that guesswork costs money in AI is telling people the wrong things about your business.
A Two-Minute Test You Can Run Today
You do not need to be technical to get a rough idea of how AI-friendly your website is.
- View the source. Open one of your service pages, press Ctrl+U (Cmd+Option+U on a Mac) and search for a key sentence from the page. If it is not there, it is being loaded by JavaScript.
- Look for text in images. Right-click your main headline. If you can’t highlight the words, they are part of an image.
- Ask the AI. Paste the page address into ChatGPT and ask it to summarise what the page says about your service, location and prices. Compare its answer with what is actually on the page.
If the summary is thin, vague or wrong, your design is getting in the way of your message.
Good for People, Good for Machines
The encouraging part is that an AI-friendly website is mostly just a well-built website. Clean HTML, fast loading, real text instead of pictures of text, descriptive headings, one clear page per service and proper accessibility all help human visitors as much as they help AI. Screen readers and AI crawlers struggle with many of the same things.
That is why we treat AI readability as part of good web design rather than a bolt-on. Your website still has to do the selling once people arrive, as we covered in why your website is your best or worst salesperson. It now also has to make the case before they arrive, to the tools doing the shortlisting.
Where to Start
Run the three checks above on your homepage and your two most important service pages. If anything important is hidden, it is usually fixable without a full rebuild. If the site is due a refresh anyway, it is the right moment to build AI readability in from the start, and your next web design project could be up to 50% funded.
Our websites and digital platforms team builds sites that read clearly to people and machines alike. Book a discovery call or call 0191 259 1234, and we will tell you honestly what AI can and cannot see on your site today.
About the author
Andrew Fenwick
Head of Web and Digital, One Design Group
Andrew leads web, SEO and AI search visibility at One Design Group. He has spent a career building websites, shaping strategic content and getting businesses found online, from WordPress design and UX to the SEO and AI search work that decides who gets seen today. It is a craft he first learned in journalism, as a staff writer at Metro and a contributor to the Guardian, and later sharpened by retraining in artificial intelligence at Saïd Business School, University of Oxford. In short, he builds sites that work then makes sure people find them.