# Does our website platform affect whether ChatGPT can read our site? Canonical: https://fortitudemedia.ai/insights/cms-choice-affects-ai-visibility Author: David Adams Published: 2026-03-31 Last reviewed: 2026-09-24 Reviewed by: David Adams Pillar: Websites that bring work in Summary: Can ChatGPT read your site? How WordPress, Webflow, HubSpot, Wix and headless builds affect AI visibility, how to test yours and when to switch. Keywords: cms ai visibility, can chatgpt read my website, webflow vs wordpress ai search, is my react site crawlable, lovable seo, hubspot cms schema markup Short answer: Somewhat, but less than how the site is built. GPTBot, ClaudeBot and PerplexityBot do not run JavaScript, so your text must be in the page's HTML when it is delivered. WordPress, Webflow, HubSpot, Wix and Squarespace can all do that; sites built with JavaScript frameworks or AI builders need server rendering or prerendering. Check robots.txt, your platform's AI crawler switch and Cloudflare, and test what crawlers see before you switch. What changed: Added a platform comparison table, AI site builders and React checks, the Vercel crawler finding, Cloudflare's default block, verified Wix, Webflow and Squarespace settings, and ID-Pal and arum. Publisher: Fortitude Media Limited When quoting, attribute to Fortitude Media and link to the canonical URL above. --- At some point in the life of every website, someone suggests changing platform. Your developer says WordPress has become a headache. A new agency swears by Webflow. Your sales director wants everything in HubSpot. Someone has built a prototype in an afternoon with an AI site builder. And now there is a new question in the mix: will the choice affect whether ChatGPT, Claude, Gemini and Perplexity can read our site and recommend us? It is a sensible question, because a migration is expensive and disruptive, and you only want to do it once. The platform matters less than most agencies suggest and more than most founders assume. The [HTTP Archive's 2025 Web Almanac](https://almanac.httparchive.org/en/2025/cms) found WordPress runs roughly 64% of sites built on a CMS, yet only 45% of WordPress sites passed Google's Core Web Vitals on mobile, against 74% for Wix. The same platform can be fast or slow, readable or not, depending on how it is built. This guide explains what AI assistants need from your website, how the common platforms compare, what to check on sites made with AI builders, how to test whether your current site is holding you back, and when switching is worth it. ## Key takeaways - Your website platform matters less than how the site is built, but a poor set-up can hide your content from the crawlers AI assistants rely on. - GPTBot, ClaudeBot and PerplexityBot do not run JavaScript, so your main text must be in the page's HTML when it is first delivered; Googlebot does run it, which helps Gemini and Google's AI features. - Since 1 July 2025 Cloudflare has blocked AI crawlers by default on new domains, and Webflow and Squarespace have their own AI crawler switches, so check the settings. - WordPress, Webflow, HubSpot, Wix and Squarespace can all deliver readable pages; they differ in control over robots.txt, structured data and code. - Sites made with AI builders or modern JavaScript frameworks need server rendering or prerendering, and you should test what a crawler receives before launch. - Before switching platforms, check what your current site shows crawlers and time a simple edit, because many problems can be fixed without moving. ## What is a CMS, and why does it matter for AI search? A CMS, or content management system, is the software your team uses to create and edit website pages without writing code. WordPress, Webflow, HubSpot, Framer, Squarespace and Wix are all examples. It matters for AI search in two ways. First, the CMS, together with how your developer set it up, decides what code a crawler receives when it asks for a page. If that code contains your words, headings and key facts, the assistants can use them. If the words only appear after scripts have run, or are locked inside images, the main AI crawlers see very little. Second, the CMS decides how easy it is for your team to keep the site current. A platform that makes updates slow tends to produce an out-of-date site, and out-of-date sites produce out-of-date AI answers. ## What do AI assistants need from your website? Before comparing platforms, be clear about what you are asking the platform to deliver. ### Your words in the page itself Your main text, headings and key facts should be in the page's code when it is first delivered, not added afterwards by scripts in the visitor's browser. [Vercel and MERJ found in December 2024](https://vercel.com/blog/the-rise-of-the-ai-crawler) that GPTBot, ClaudeBot and PerplexityBot fetch JavaScript files but do not run them, so content that depends on scripts is invisible to them. Googlebot renders JavaScript, which helps Gemini and Google's AI features, but that does not help with ChatGPT, Claude or Perplexity. We explain how the crawlers differ in [how AI crawlers read your website](/insights/ai-crawlers-vs-google-spiders). ### A clear structure One main heading per page, sensible subheadings beneath it, and one clear topic per page. Headings should be real headings in the code, not ordinary text styled to look large. ### Accurate page titles and descriptions Every important page should have its own title and short description that say what the page is about. ### Structured data Structured data is a small block of labelled information in a page's code that tells machines facts such as your firm's name, address, services and the author of an article. It helps search engines read your facts correctly, and you want a platform that lets you add it. Our guide to [whether you need schema markup](/insights/structured-data-schema-markup-ai) explains what is worth adding. ### Access for AI crawlers Your robots.txt file, a short text file that tells crawlers which parts of the site they may visit, should not block the crawlers of the assistants you want to be named by. Check your security and hosting settings too. [Cloudflare changed its default on 1 July 2025](https://www.cloudflare.com/press-releases/2025/cloudflare-just-changed-how-ai-crawlers-scrape-the-internet-at-large/): new domains now start with AI crawlers blocked unless the owner chooses to allow them, and many firms never realise the setting is on. ### Easy editing Your team should be able to update a service page, add a case study or change a team profile in minutes, without calling a developer. This is the requirement most often forgotten when platforms are chosen. ## How do the common platforms compare? Each of the main platforms can meet those needs. They differ in how much control you have and where the risks lie. The table reflects each platform's help pages where we could check them in September 2026; features change often, and a lot depends on how your particular site is built. For running and build costs, see [how much a B2B website costs in the UK](/insights/real-cost-of-a-website). | Platform | Text in the page by default | Edit robots.txt | Custom structured data | AI crawler switch | |---|---|---|---|---| | WordPress | Yes, pages are built on the server | Yes, by file or plugin | Yes, by plugin or by hand | Through robots.txt or plugins | | Webflow | Yes | Yes, in site settings | Yes, through custom code on a paid plan | Yes, traffic control toggles for named AI bots | | HubSpot CMS | Yes, depending on the template | Yes, in settings | Yes, through head code or templates | Through robots.txt | | Framer | Generally yes, but test yours | Check current settings | Yes, through custom code | Check current settings | | Squarespace | Yes | No direct editing | Limited, through code injection | Yes, "Block known artificial intelligence crawlers" | | Wix | Yes | Yes, as an advanced feature | Yes, up to five JSON-LD blocks per page | Through robots.txt | | Headless or custom build | Only if pages are server-rendered or generated in advance | Yes | Yes | Yes | ### WordPress Self-hosted WordPress gives you the most control. You can edit almost anything, add structured data through plugins such as Yoast SEO or Rank Math or by hand, and move your content elsewhere later. The risks come from that same freedom: heavy page builders, too many plugins and poor hosting make many WordPress sites slow, and someone has to keep the software updated for security. ### Webflow Webflow is a hosted visual builder popular with design-led agencies. It handles hosting and security, lets editors manage articles and case studies through content collections, and its [site settings](https://help.webflow.com/hc/en-us/articles/33961368603539-Disable-search-engine-indexing) let you edit robots.txt and switch named AI bots such as GPTBot, ClaudeBot and PerplexityBot on or off. The trade-offs are cost, reliance on one company and the work involved in moving away later. ### HubSpot CMS Many B2B firms already use HubSpot for contacts and pipeline, and its website platform connects directly to that, which makes it easier to see which pages lead to enquiries. It is hosted, so updates and security are handled for you. How well a HubSpot site reads to crawlers depends on the template, so run the checks below. ### Squarespace and Wix These are the simplest to use, and they handle hosting, security and mobile layouts for you. They are more capable than they were: [Wix lets you edit robots.txt](https://support.wix.com/en/article/editing-your-sites-robotstxt-file) and [add your own JSON-LD structured data](https://support.wix.com/en/article/adding-structured-data-markup-to-your-sites-pages), and [Squarespace has a setting to block known AI crawlers](https://support.squarespace.com/hc/en-us/articles/360022347072), which you should check is off if you want to be read. The limits appear as the site grows: less control over code, and templates that make detailed service and sector pages harder. We look at the wider trade-offs in [template or custom website: which does a B2B firm need?](/insights/template-vs-custom-website-b2b). ### Headless and custom builds A headless set-up stores content in one system and has developers build the website separately, often with a JavaScript framework such as React. It can be fast and flexible. The main risk is a site that assembles its pages in the visitor's browser, which leaves the main AI crawlers looking at an almost empty page. Insist that pages are rendered on the server or generated in advance. ## Can ChatGPT read a website built with Lovable, Framer or another AI site builder? Usually, if the builder delivers full HTML, and you should check before launch. Framer sites generally arrive as HTML, but test yours. Lovable's [documentation](https://docs.lovable.dev/tips-tricks/seo-geo) says apps created from 13 May 2026 use server-side rendering, while older React and Vite apps get pre-rendered pages served only to verified search and AI crawlers, and other tools see the ordinary single-page app. Other AI builders, such as Bolt and v0, generate code in various frameworks, and whether a page arrives as full HTML depends on the framework and settings used. Treat any site made this way like a custom React build: confirm it is server-rendered or prerendered, then test what a crawler receives, as below. ## How do I check whether my React website is readable by AI crawlers? Look at the HTML the server sends before any script runs. If your headings and body text are in it, AI crawlers can read the page; if you see an almost empty page with a script tag, they cannot. Three simple checks: open the page, right-click and choose "View page source", then search for a sentence from the page; turn off JavaScript in your browser's developer tools and reload; or ask your developer to fetch the page from the command line with a crawler's user agent. Do this for the homepage, a service page and an article, because different templates can behave differently. ## How to check whether your platform is holding you back You do not need to switch platforms to find out whether yours is a problem. These checks take a morning. ### Step 1: See what a crawler sees Use the page source check above on your homepage and one service page. If the source is mostly code with little of your text in it, that is the first thing to fix. ### Step 2: Check your crawler settings Open your robots.txt file by adding /robots.txt to your web address, and look for lines that block GPTBot, OAI-SearchBot, ClaudeBot, Claude-SearchBot or PerplexityBot. Then check your platform's AI crawler setting and any Cloudflare or hosting setting. Decide deliberately which crawlers you allow. ### Step 3: Check headings and titles On your five most important pages, check that each has one main heading, clear subheadings and its own page title and description. ### Step 4: Check whether you can add structured data Ask your developer whether the platform lets you add structured data for your organisation, services and articles, and whether any is there already. ### Step 5: Time a simple edit Ask the person who manages the site to update a sentence on a service page and add a new team member. If it takes more than a few minutes or needs a developer, the site will struggle to stay current. For a fuller list, see [the technical checklist for an AI-ready website](/insights/technical-checklist-ai-ready-website). ## When is it worth switching platforms? Switch when the platform stops you fixing problems, not when a new agency prefers a different tool. Moving platform almost always changes page addresses and puts search history, content and senior time at risk. Good reasons include: - Your team cannot update pages without a developer, and the platform cannot be set up to change that. - Your main content cannot be delivered in the page itself on the current set-up. - You cannot control crawler access, page titles or structured data at all. - The platform, or your version of it, is no longer supported, or the site is too slow to fix without rebuilding. If you do switch, list every existing page, redirect every old address to its new one, and check afterwards that nothing was dropped. When we rebuilt [arum](/work/arum)'s site, "Every old address redirected, so years of rankings carried over." We discuss when a rebuild is justified in [whether to redesign your website or keep improving it](/insights/website-redesign-vs-continuous-iteration). ## What matters more than the platform? What your site says, and whether the rest of the web agrees with it. A perfectly built site will not be recommended if it describes your firm vaguely or contradicts your LinkedIn page and directory listings. Many fixes need no migration at all. For [ID-Pal](/work/id-pal), the fix was "About four hours of structured data work declaring NorthRow part of ID-Pal on both domains. Live in week one, with no copy changes and no risk to the redirect." In our experience, most firms that are not named by AI assistants have a content and clarity problem, not a platform problem. We look at how structure affects visitors and AI in [how to organise your website so buyers and AI understand it](/insights/website-architecture-ux-and-ai). ## What to do this month 1. Check the page source of your homepage and main service page for your actual text. 2. Check your robots.txt file, your platform's AI crawler setting and any Cloudflare setting, and allow the crawlers you want to be read by. 3. Fix headings and titles on your top five pages. 4. Time a simple edit, and if updating a page needs a developer, fix that before anything else. 5. If your site was made with an AI builder or a JavaScript framework, confirm it is server-rendered or prerendered. 6. Decide on switching only after these checks. ## Frequently asked questions ### Is Webflow or WordPress better for SEO and AI search in 2026? Neither is better in itself. WordPress gives more control and more structured data options but depends heavily on how well it is built and maintained. Webflow needs less upkeep and has built-in AI crawler controls, but ties you to one company. Both deliver pages as HTML that AI crawlers can read, so choose the one your team can edit easily and your developer can support properly. ### Does HubSpot CMS let us add custom schema markup? Yes. Structured data in JSON-LD format can be added to a page's head code or built into the templates your developer creates, so organisation, service and article markup are all possible. Check what the template already outputs first, so you do not end up with two conflicting versions. ### Can a Squarespace or Wix site be recommended by ChatGPT? Yes. What an assistant says about you depends mainly on what your site says and what other sources say about you, and a clear, accurate Squarespace or Wix site can be read well. For a small, simple firm, it may be enough. ### How much does it cost to move a website from WordPress to Webflow? It depends far more on the site than on the platforms. The main drivers are the number of pages and templates, how many articles and case studies must move, redirects for every old address, forms and integrations, and who rewrites the content. We do not quote prices here: ask for a fixed quote that lists each of those items, and see [how much a B2B website costs in the UK](/insights/real-cost-of-a-website) for published market ranges. ### Should we block AI crawlers to protect our content? That is a business decision, but for most B2B service firms the answer is no. If you block the crawlers ChatGPT, Claude and Perplexity use, those assistants have less of your own material to work from. Some firms block training crawlers such as GPTBot and ClaudeBot but allow search crawlers such as OAI-SearchBot and Claude-SearchBot. Whatever you choose, choose it deliberately. The right platform is the one that shows your words clearly to crawlers and lets your team keep them current without a fight. To have a site built with both in mind, see [Website That Sells](/website-that-sells), or book a free [Growth Review](/growth-review) to find out what crawlers and assistants see today. ## Sources - [HTTP Archive, Web Almanac 2025: CMS (January 2026)](https://almanac.httparchive.org/en/2025/cms) - [Vercel, The rise of the AI crawler (December 2024)](https://vercel.com/blog/the-rise-of-the-ai-crawler) - [Cloudflare, Cloudflare just changed how AI crawlers scrape the internet at large (July 2025)](https://www.cloudflare.com/press-releases/2025/cloudflare-just-changed-how-ai-crawlers-scrape-the-internet-at-large/) - [Lovable documentation, search and AI visibility best practices (accessed September 2026)](https://docs.lovable.dev/tips-tricks/seo-geo) - [Webflow Help Center, Disable search engine indexing (accessed September 2026)](https://help.webflow.com/hc/en-us/articles/33961368603539-Disable-search-engine-indexing) - [Wix Help Center, Editing your site's robots.txt file (accessed September 2026)](https://support.wix.com/en/article/editing-your-sites-robotstxt-file) - [Wix Help Center, Adding structured data markup to your site's pages (accessed September 2026)](https://support.wix.com/en/article/adding-structured-data-markup-to-your-sites-pages) - [Squarespace Help Center, Request that AI models exclude your site (accessed September 2026)](https://support.squarespace.com/hc/en-us/articles/360022347072)