Squarespace AI Visibility Explained, and the Schema It Still Doesn't Add

In July 2026 Squarespace released AI Visibility, a panel that tells you whether ChatGPT and Gemini mention your business when people ask them questions in your category. It's a useful dashboard. It is not an optimisation tool, and the part that most affects whether you get cited is something Squarespace still leaves to you.

This is what the tool shows, what to do with it, and the structured data we add to every site because Squarespace doesn't.

What AI Visibility does

You'll find it under SEO & AIO in your site dashboard, next to the SEO report. It's included in your plan but runs on AI credits, which are allocated by plan tier; check your account for the current allowance. It works on English-language 7.1 sites and currently covers ChatGPT and Gemini only.

The panel shows four things:

Prompts and whether you're mentioned. Squarespace generates a set of questions a customer might ask an assistant ("best interior architect in Manchester", "dental clinic that offers same-day crowns near me") and reports whether your business appeared in the answer.

Who appeared instead. The competitors named in those answers. This is the most useful part of the tool, because it tells you who the models currently trust in your category.

Sources cited. The URLs the assistants pulled from when composing their answers. Often these aren't the competitors' own sites; they're directories, review platforms, and local news.

A visibility score over time. A single number that moves as mentions change. Squarespace suggests checking monthly and allowing a week or two after changes before expecting movement.

What it doesn't do

It doesn't change anything on your site. It doesn't tell you why a competitor was mentioned and you weren't. It doesn't cover Perplexity, Claude, Copilot or Google's AI Overviews, which is where a large share of AI-assisted searching actually happens. And it measures a fixed set of prompts that may or may not match what your customers ask.

Treat it as a monitoring tool, the way you'd treat rank tracking. Useful for noticing change, not for causing it.

Why sites get cited by AI assistants

The models cite pages they can read clearly and trust. Clarity comes from structure: plain HTML, headings that state the topic, direct answers near the top, and structured data that tells a machine unambiguously what the page is about, who's behind it, and where they are. Trust comes from the same things that have always driven SEO, mainly being referenced elsewhere.

Squarespace does the first half well. Pages are server-rendered, the sitemap is automatic, titles and descriptions are editable, and blog posts and products get basic Article and Product schema.

What it does not add, on any plan, is the structured data assistants lean on most for a business:

• LocalBusiness (or a subtype like Dentist, Restaurant, LegalService) with address, phone, opening hours, price range and geo coordinates.

• Service for each thing you sell, with a description and area served.

• FAQPage for your FAQ sections. Squarespace's accordion block looks like an FAQ to a human and like a list of paragraphs to a machine.

• Person for the founder or practitioner, linked to their credentials and profiles.

• Organization with your logo, founding date and social profiles, done properly rather than the minimal version Squarespace emits.

• Review and AggregateRating, where you have real reviews to mark up.

Adding the schema Squarespace leaves out

All of this goes into header code injection (Settings, then Developer Tools, then Code Injection, on Core plan and above) or into per-page header injection for page-specific schema. It's JSON-LD, which means a <script type="application/ld+json"> block; nothing on the visible page changes.

Here's a LocalBusiness block we'd put on the home page of a dental clinic. Replace every value.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Dentist",
  "name": "Avira Dental",
  "url": "https://www.example.com/",
  "logo": "https://www.example.com/logo.png",
  "image": "https://www.example.com/clinic.jpg",
  "telephone": "+1-555-010-0100",
  "priceRange": "$$",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "12 High Street",
    "addressLocality": "Springfield",
    "addressRegion": "IL",
    "postalCode": "62701",
    "addressCountry": "US"
  },
  "geo": { "@type": "GeoCoordinates", "latitude": 39.7817, "longitude": -89.6501 },
  "openingHoursSpecification": [
    { "@type": "OpeningHoursSpecification",
      "dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday"],
      "opens": "08:00", "closes": "18:00" }
  ],
  "sameAs": [
    "https://www.instagram.com/example",
    "https://www.linkedin.com/company/example",
    "https://g.page/example"
  ]
}
</script>

For a FAQ section, per-page injection on that page:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    { "@type": "Question",
      "name": "Do you offer same-day crowns?",
      "acceptedAnswer": { "@type": "Answer",
        "text": "Yes. We mill crowns in-house and fit most in a single two-hour appointment." } },
    { "@type": "Question",
      "name": "Do you accept new patients?",
      "acceptedAnswer": { "@type": "Answer",
        "text": "Yes. Book online or call; new patient appointments are usually available within a week." } }
  ]
}
</script>

The answers in the schema must match the answers on the page word for word, or close to it. Google penalises FAQ markup that doesn't correspond to visible content, and assistants ignore it.

Validate everything at validator.schema.org and in Google's Rich Results Test before you move on. One missing comma and the whole block is ignored silently.

The rest of the checklist

Schema gets you read correctly. The following gets you cited.

Answer the question in the first hundred words. An assistant summarising your service page wants the answer, not the story. Put "We are a dental clinic in Springfield offering X, Y and Z, open weekdays 8 to 6" at the top. The narrative can follow.

One page per service. Assistants match a query to a page. A single "Services" page with nine services on it competes with nine competitors' dedicated pages and loses.

Write headings as questions where the content answers one. "How much does a crown cost?" as an H2 with the answer beneath it is easier for a model to extract than "Pricing".

Make sure you exist elsewhere. Look at the sources in the AI Visibility panel. If assistants are citing a directory or a review site for your competitors, you need to be on it, with the same name, address and phone number as your site. This matters more than anything on the site itself.

Original images with descriptive filenames and alt text. Stock images tell a model nothing. A photo named avira-dental-reception-springfield.jpg with matching alt text is a signal.

Check robots.txt and the "Hide from search engines" setting. We still find live sites with search hidden, months after launch. Squarespace's robots file allows the major AI crawlers by default; if you've added a custom disallow, review it.

Turn off nothing. Some advice recommends blocking GPTBot to protect content. If you want to be cited, don't.

How we run this for clients

An AI visibility set-up on an existing Squarespace site takes us four to eight hours: schema for the organisation, every location and every service, FAQ markup where FAQs exist, page rewrites of the opening paragraph on the top ten pages, and a citation audit against the sources the panel reports. Then a monthly check of the panel as part of the support plan.

It's part of our SEO & AI Optimisation service, and it's included in every new site we build, because there's no reason to launch without it.

Questions we get

Does adding schema guarantee AI assistants will cite me? No. It makes you readable. Being cited also depends on whether the assistant finds you on the sources it already trusts, which is why the sources list in the panel matters more than the score.

Do I need a separate page for every question in my FAQ schema? No. One FAQ page or section with FAQPage markup covering its questions is correct. Don't mark up the same question on ten pages.

Will Squarespace add these schema types eventually? Possibly. They've added Article and Product; LocalBusiness would be the obvious next step. Until then, injection is the only route, and the code above works regardless of what Squarespace later adds, as long as you don't duplicate the same type twice.

What about Google's AI Overviews? Not tracked by the panel, but the same work applies: clear answers near the top, structured data, and citations elsewhere. Pages that rank well in normal search are what AI Overviews draw from.

Should I block AI crawlers to protect my content? If the site exists to bring you customers, no. Blocking GPTBot and similar removes you from the answers your customers are increasingly asking.

How long until the panel shows a change? Squarespace suggests a week or two. In our experience, a month is a fairer horizon, and citation changes tend to follow directory and review changes, not on-site changes alone.