If the words on your website only appear after JavaScript runs in a browser, a crawler that does not run JavaScript receives a page with almost nothing on it, and an AI answer cannot quote what it never received. You see a finished site because your browser runs the code. Google's own documentation says not all bots can run JavaScript, and OpenAI's crawler page does not say whether its crawlers do. The test takes a minute: open your page's source and search for your phone number. If it is missing, ask your developer to pre-render the pages so the words arrive in the first file, then run the test again.
You paid for a website. You open it on your phone and it looks right: your logo, your services, your prices, the photo of the shop. You ask ChatGPT for a business like yours in your area, and it names three others. Your site is live, so the obvious explanation is that the engines have not got to it yet.
Sometimes that is the answer. Sometimes the answer is that the crawler did get to it, and found almost nothing there. This post explains how that happens, how to check your own site in a minute, and what to ask your developer for if the check fails.
Why can a website look finished to me and empty to a crawler?
Because your browser does work that some crawlers skip. Many modern websites send a short starter file and a bundle of JavaScript code. Your browser runs the code, and the code writes the words and menus onto the page. You never see the starter file. A crawler that does not run JavaScript sees only the starter file, and on these sites the starter file holds a title and no content.
Google describes this in its own documentation for developers. It processes JavaScript pages in three phases: crawling, rendering and indexing. Rendering is the step where a headless Chromium browser runs the code, and Google says a page can wait in a queue for that step for a few seconds or longer. The same page recommends server-side rendering or pre-rendering anyway, and gives the reason in plain words: not all bots can run JavaScript.
So the question for your business is which crawlers run the code and which do not.
Does ChatGPT run JavaScript when it reads a website?
OpenAI does not say. As of 13 October 2026, OpenAI's crawler documentation lists four crawlers and what each is for: OAI-SearchBot surfaces websites in ChatGPT's search features, OAI-AdsBot checks the landing pages of ads, GPTBot collects content that may be used to train its models, and ChatGPT-User visits a page when someone asks ChatGPT a question that needs it. The page says nothing about whether any of them run JavaScript. Perplexity's crawler page, read the same day, is also silent on it.
When the companies do not document it, you cannot know which version of your page their crawlers receive. You can remove the question entirely, by making sure the words are in the first file your server sends. A page like that is readable by every crawler, whether it runs scripts or not, and it loses nothing for your human visitors.
We have seen the problem in our own work. In September we audited a site for a small US agency, built with one of the AI website builders that are now common. In a browser it looked finished. Fetched by a crawler that does not run JavaScript, every page returned the same near-empty file, carrying the home page's title on every address. The owner had no way to see it from the front of the site.
The one-minute test: search your page source for your phone number
You do not need a developer to find out. You need a laptop and your phone number.
- Open your home page in Chrome, Edge or Firefox on a laptop or desktop.
- Press Ctrl+U (on a Mac, Cmd+Option+U). A new tab opens with the page source, which is the starter file your server sends before any code runs.
- Press Ctrl+F (Cmd+F on a Mac) and type your phone number, written exactly as it appears on the site.
- Repeat for your services page and your contact page.
If the number is found, that page sends its words in the first file. If the search finds nothing, and the source is mostly lines of code with very little readable text, that page depends on JavaScript to show its content.
When the test is done you will know, page by page, whether your words arrive before the code runs or only after.
A phone number works well for this because it is a fact that should be on every page, it is unique to you, and it is exactly the kind of fact an AI answer repeats. If your number is not in the source, your list of services is almost certainly missing too.
What do I ask my developer for?
Ask for the pages to be pre-rendered, so the server sends finished pages with the words already in them. Developers have three names for closely related fixes: server-side rendering, static generation and pre-rendering. Any of them solves the problem in this post. Which one fits depends on how the site was built, and that is your developer's call.
Send them this, word for word if it helps:
Our phone number is not in the page source on these pages: [list them]. Can you pre-render the pages, or switch to server-side rendering, so the text is in the HTML the server sends? Please also make sure each page sends its own title, not the home page's.
That last line matters because the agency site we audited had a second problem hiding behind the first. Every address sent the same title, so even a crawler that did read the starter file was told every page was the home page.
If the site was made on a website builder rather than by a developer, ask the builder's support team the same question. Some offer pre-rendering as a setting or an add-on.
When the fix is live you will see it the same way you saw the problem: run the phone number test again and the number is there.
What else should I check while the developer is in there?
Two things, both quick.
- Is the crawler allowed in? Your site has a file at yourdomain.com/robots.txt that tells crawlers what they may read. OpenAI's documentation says a site that blocks OAI-SearchBot will not appear in ChatGPT's search answers, and that changes to the file take about 24 hours to take effect. Perplexity's page says the same of PerplexityBot. Ask the developer to confirm both are allowed. When this is done, the file names no rule that blocks them.
- Can Google show the page at all? Google's documentation says a page has to be indexed and eligible to show a snippet in ordinary search before it can appear in an AI Overview or AI Mode. If a page is missing from Google entirely, start there. When this is done, a search for your exact business name on Google shows your own pages.
Why this decides who gets named
An AI answer is assembled from text the engine could read. If your site gives a crawler no text, your business can only appear through what other sites say about you, and for most small businesses that is very little. The competitor whose site sends its words in the first file gives the engine a page it can quote: who they are, what they do, where, and how to reach them.
The people asking are not few:
Being readable does not get you named on its own. It is the condition every other step depends on. A page that answers your customer's question well does nothing if the crawler receives an empty file in its place.
How long until it makes a difference?
Nobody can give you a date. OpenAI, Google and Perplexity publish no timeline for how quickly a changed page is picked up. What you can control is the order: fix readability first, because every improvement you make to your pages after that depends on the engine being able to read them. Then ask the same customer questions in ChatGPT next month and write down who is named.
Want to know what else the machines cannot read?
Our free AI Visibility checklist covers 12 checks, starting with whether the crawlers can read your pages at all, then what ChatGPT, Gemini and Google say when someone asks about a business like yours. Enter your email and we will send it over.
✓ Check your inbox, your checklist is on the way!
We respect your privacy. Unsubscribe at any time.
A website that only shows its words after JavaScript runs can look finished to you and arrive empty to a crawler. Search your page source for your phone number; if it is not there, ask your developer to pre-render the pages and test again.
One forward before you go: open your home page on a laptop, press Ctrl+U (Cmd+Option+U on a Mac), press Ctrl+F and type your phone number. If it is not found, forward this post to whoever built the site with one line: our phone number is not in the page source, can you pre-render the pages?
Frequently asked questions
Does ChatGPT run JavaScript when it reads my website?
OpenAI does not say. As of 13 October 2026, its crawler documentation describes what OAI-SearchBot, GPTBot and ChatGPT-User are for, and says nothing about whether they run JavaScript. When a company does not document it, the safe choice is a page whose words are in the first file the server sends, because that page works for every crawler whether it runs scripts or not.
Google can read JavaScript. Why does this still matter?
Google documents that it renders JavaScript pages with a headless Chromium browser, in a separate step after crawling, and that a page can wait in a queue before that happens. The same page says server-side rendering or pre-rendering is still a great idea, because not all bots can run JavaScript. Google is one engine; ChatGPT, Perplexity and the others that build AI answers are separate crawlers with their own rules.
What is pre-rendering, in plain words?
It means the server sends a finished page, with the words already in it, instead of an empty page plus code that writes the words in the browser. Visitors see the same site. A crawler that does not run scripts now receives the text too. Your developer may call it server-side rendering, static generation or pre-rendering; any of the three solves the problem this post describes.
My site was built with an AI website builder. Is it affected?
It might be, and the platform name does not tell you. Some builders produce pages that write their content with JavaScript after they load, which is the pattern this post describes. Run the phone number test on two or three pages. If your number is in the source, that page is readable; if it is missing, ask the builder's support or your developer whether pre-rendering is available.
Is this an SEO problem or an AI problem?
Both, and the fix is the same. In SEO terms it is a rendering issue that Google handles better than most crawlers. For AI visibility it decides whether an engine that does not run scripts receives any of your words at all. One change, pre-rendering, solves it for both.
I fixed it. How soon will ChatGPT mention my business?
Nobody can tell you honestly. OpenAI, Google and Perplexity publish no timeline for how quickly a changed page is picked up. Being readable is the condition for being named; it does not guarantee it. Re-run the questions your customers ask next month and record what changes.
References & sources
- Google Search Central, Understand the JavaScript SEO basics (last updated 4 March 2026, read 13 October 2026: crawling, rendering, indexing; rendering queue; not all bots can run JavaScript)
- OpenAI, OpenAI bots documentation (OAI-SearchBot, OAI-AdsBot, GPTBot, ChatGPT-User) (read 13 October 2026; no statement on JavaScript)
- Perplexity, Perplexity crawlers (PerplexityBot, Perplexity-User) (read 13 October 2026)
- Google Search Central, AI features and your website (a page must be indexed and eligible for a snippet)
- Pew Research Center, Americans and AI 2026 (5,119 US adults, February 2026)
- BrightLocal, Local Consumer Review Survey 2026 (1,002 US adults)