Purcell VenturesStart a project

Free, no account

AI visibility check

Enter your business website. The check reads your home page the way an AI crawler does and lists what to fix, with the reason for each.

The check takes up to 20 seconds. It reads three public files and stores nothing.

What the check looks at

The check fetches three addresses on your site: the home page, /robots.txt and /llms.txt. It reads the home page as the server sends it, before any JavaScript runs, because many AI crawlers read pages that way.

Secure connection

Browsers mark a site without https as not secure and many visitors leave, so ask your host to switch on a free certificate and send every http address to the https version. Fixing the not secure label.

AI crawlers in robots.txt

An assistant that cannot read your site describes you from directories and reviews instead, so remove the robots.txt lines that block the search crawlers unless you sell the writing on your site. Should I block AI crawlers.

llms.txt file

An llms.txt file lists your main pages for AI systems in plain text, and AI companies do not all read it yet, so treat it as a twenty-minute extra after the other checks pass. What an llms.txt file is.

Page title

Search results and AI answers often show the title as the name of your page, so give the home page a title with the business name, what it does and the town. How to write a page an AI will quote.

Meta description

The description is the summary a search engine shows under your link, so write one or two sentences that say who you serve, what you do and where. How to write a page an AI will quote.

Business structured data

LocalBusiness or Organization markup labels your name, address, phone and hours for machines, so add a JSON-LD block with the same facts the page shows. Does schema markup help AI find my business.

Phone number in the page text

Assistants repeat the phone number they can read, so print it as text on the home page and keep it identical to your Google listing. Claiming your own Google listing.

Address in the page text

A street address or a city, state and ZIP in the page text tells an assistant where you work, so print it as text and match your Google listing, or name your service area if you work from home. Claiming your own Google listing.

Opening hours

People ask assistants whether a business is open now, so print your hours as text on the home page and keep them the same as your Google listing. Claiming your own Google listing.

Readable text without JavaScript

Many AI crawlers read the raw page without running scripts, so a site that draws its text with JavaScript looks empty to them, and the fix is server-rendered or static pages. Can an AI find your business.

The five crawlers

  • GPTBot, from OpenAI, collects pages OpenAI may use to train its models.
  • OAI-SearchBot, from OpenAI, finds pages to show in ChatGPT search answers.
  • ClaudeBot, from Anthropic, collects pages Anthropic may use to train Claude.
  • PerplexityBot, from Perplexity, finds pages to show in Perplexity answers.
  • Google-Extended, from Google, controls whether Google may use pages for its Gemini models, and has no effect on Google Search.

The check reads your robots.txt the way the crawlers do. A rule that names the crawler applies first, and a rule for every crawler applies when nothing names it.

Common questions

What does the AI visibility check look at?

It reads your home page, your robots.txt file and your llms.txt file. It reports whether five AI crawlers may read the site, whether the page has a title, a meta description and business structured data, whether a phone number, an address and opening hours appear as text, how many words the page shows before JavaScript runs, and whether the site uses https.

Does the check store my website address?

No. The server fetches the three files, sends the findings back to your browser and keeps nothing. There is no account and no email step.

Why does the check count so few words on my home page?

The check reads the page the way the server sends it, before any JavaScript runs. Some site builders and app frameworks send a nearly empty page and draw the text in the browser. A crawler that skips scripts sees that empty page.

Should I block GPTBot and ClaudeBot?

Those two crawlers collect pages for model training. A business that sells its writing or research may reasonably block them. OpenAI says ChatGPT search uses OAI-SearchBot, so blocking GPTBot alone leaves a site eligible for ChatGPT search answers. Most local businesses leave all five open.

If every check passes, will ChatGPT recommend my business?

Nobody can promise that. The check covers the parts of your own site that you control. Assistants also read your Google Business Profile, directories and reviews, and none of them publish how they pick a business.

Why does it say there is no phone number when my number is on the site?

The check looks for a US-format number in the readable text. A number inside an image, a number that appears only in a click-to-call button, or a number drawn by JavaScript does not count, and those are the same cases a crawler misses.

Who built it

I am Elijah Purcell, Founder/CEO of Purcell Ventures LLC in Tuscaloosa, Alabama. I built it to run the checks from my AI websites guide on any site in one pass. If you want the fixes done for you, the websites page lists the build tiers and the care plans.

Updated