We ran VisiScan on VisiScan: our day-one AI visibility score

Every AI visibility report we build for a customer starts with the same blunt question: when someone asks ChatGPT, Claude, Perplexity, or Gemini about your business, does your name come back — or does a competitor's? This question is pivotal because it sets the stage for understanding your business's presence in the rapidly evolving digital landscape dominated by AI-driven search engines.

The result is an observation of a defined prompt set and measurement date, not a permanent ranking. This means that the visibility you have today can change tomorrow based on various factors, including content updates, backlink acquisitions, and shifts in user engagement. If you want to run the same kind of check on your own business, how to choose an AI visibility scanner explains what to look for in a tool, and how to check whether ChatGPT, Claude, Perplexity, and Gemini already mention your business walks through the manual version.

We hadn't asked that question about ourselves until recently.

So on July 24, 2026, we ran the same kind of audit VisiScan runs for customers — but pointed at visiscan.app itself, plus a full technical and off-page SEO crawl. This was a critical step in self-assessment, allowing us to understand what a brand-new domain with zero marketing history actually looks like when measured honestly.

Short answer: On July 24, 2026, our worldwide AI visibility score came back at 0 — zero AI mentions, zero cited pages, zero referring domains, zero backlinks. That is the honest day-one baseline for a brand-new domain, not a permanent rank; visibility is an observation of a defined prompt set and date.

The number: 0

The external launch baseline checked ChatGPT plus Google AI surfaces, while the VisiScan product measures ChatGPT, Claude, Perplexity, and Gemini. Across the completed launch checks, our worldwide visibility baseline came back at 0. We also found zero cited pages, referring domains, and backlinks in the accompanying web audit. These are different observations, so later comparisons must use the same engines and prompt set rather than treating them as one permanent rank. The measurement methodology defines the distinction between mentions, citations, recommendations, and web-audit findings.

This zero score is not a bug in the measurement. It's an accurate description of a domain that launched a few weeks ago with no off-page footprint yet. AI engines answer from what they've read and what other sites say about a business — and right now, nothing points at us. This highlights a crucial aspect of AI visibility: it is not just about having a website; it’s about having a presence that is recognized and referenced by other credible sources.

For example, if a business has an active blog, engages with its audience on social media, and collaborates with industry influencers, it is more likely to receive mentions in AI search results. Conversely, a new business without such activities may find itself invisible. To illustrate, consider a hypothetical startup in the tech industry that regularly publishes insightful articles and participates in relevant online discussions. This proactive approach not only helps in gaining visibility but also positions the startup as a thought leader, which can lead to increased mentions and citations in AI responses.

First-party benchmark (measured July 24, 2026): VisiScan ran 5 buyer-intent questions across 4 answer engines (ChatGPT, Claude, Perplexity, Gemini), 2 samples each = 40 total answer observations. Result: 0 mentions, 0 citations, 0 referring domains, 0 backlinks. On-page readiness scored 100/100 (Lighthouse SEO) with 3/3 agentic-browsing checks passing. This is a single-day snapshot; re-run with the same 5-question set is planned for Q4 2026.

Metric	July 24, 2026
Answer-engine prompts tested	5 questions × 4 engines × 2 samples = 40
AI mentions of visiscan.app	0
Cited pages	0
Referring domains	0
Backlinks	0
On-page readiness (Lighthouse SEO)	100/100
Agentic browsing checks	3/3

Reproducible measurement record

Field	Recorded value
Target	visiscan.app
Date	July 24, 2026
Questions	5 unbranded buyer-intent questions
Engines	ChatGPT, Claude, Perplexity, Gemini
Samples	2 per question and engine
Total observations	40
Interpretation	Dated snapshot, not a permanent rank

Representative prompt format:

For a buyer in [city] looking for [service], which businesses should they consider? Name the strongest options, explain why, and cite the sources you used. Do not assume the business in the prompt is reputable.

The measurement methodology explains the recording rules. The citation source maps show which independent domains appear in category-level observations.

What did pass

While the audit revealed a stark absence of external visibility, it wasn't all zeros. Our on-page fundamentals scored well: a 100/100 Lighthouse SEO score, complete Open Graph and Twitter card metadata, a valid canonical tag, a published llms.txt file, and a perfect 3/3 on Lighthouse's newer "agentic browsing" checks — the ones that measure whether an AI agent can actually parse and act on a page, not just whether a human can read it.

These on-page factors are critical because they ensure that when users or AI systems do come across our site, they have a seamless experience. For instance, having complete Open Graph metadata means that when our pages are shared on social media, they display correctly, enticing users to click through. However, we also identified real, fixable gaps: a stale directory badge linking to someone else's product, FAQPage schema that wasn't marked up despite having 13 useful FAQ answers on the homepage, and a Content-Security-Policy header that blocked our own sign-in widget for some visitors. All three are now fixed.

Addressing these issues is essential for improving our visibility and user experience. For example, updating the directory badge to link to relevant and current resources not only enhances our credibility but also improves the chances of being cited by other sites. Ensuring that our FAQPage schema is correctly marked up allows search engines to better understand our content, potentially leading to rich snippets in search results. Rich snippets can significantly enhance click-through rates, making it an important aspect of SEO strategy.

Moreover, optimizing on-page elements like meta descriptions and header tags can further improve how search engines interpret our content. For instance, using relevant keywords in these areas can help align our pages with the queries users are making, thereby increasing the likelihood of being featured in AI-generated responses.

Why we're publishing our own bad score

Because the alternative — quietly fixing things and never mentioning the zero — would contradict the entire reason VisiScan exists. Our pitch to customers is "stop guessing what AI says about you, measure it." Measuring it and then hiding the number when it's unflattering would make us exactly the kind of company we built this tool to catch out.

Transparency is key in building trust with our customers. By openly sharing our initial score, we demonstrate our commitment to accountability and improvement. The plan from here is straightforward and public: publish real content (this post is part of that), earn genuine backlinks and citations instead of manufacturing them, and re-run this same audit in a few months to see the number move. We will publish that follow-up too, whatever it says.

This proactive approach not only helps us improve our visibility but also serves as a case study for our clients. They can see firsthand the steps we are taking to enhance our AI visibility and the results of those efforts. By sharing our journey, we aim to encourage other businesses to take similar steps towards improving their own visibility.

You can follow the measurement approach in our methodology, explore the AI visibility tools, or see how public scores are presented in the AI Visibility Index. The benchmark above and our industry citation-source maps live in the public research hub.

If you want to see where your own business stands before we do, run a free scan — it takes about 60 seconds, and it's the same honest measurement we just pointed at ourselves. The AI visibility benchmark guide explains how to repeat and publish a comparable result.

What we're doing about the zero

Publishing this post was step one. Since then, we've published research on AI visibility patterns, built the AI Visibility Index with live leaderboards, and opened the public research hub with citation-source maps per industry. Each piece of content is citable on its own terms — original data, described transparently, with the methodology published alongside it — because that is how you earn the citations that move the number. We'll re-run the same 5-question audit and publish the follow-up regardless of whether the score moves. This is an excellent opportunity for businesses to gain insights into their own AI visibility and take actionable steps to improve it. By taking this step, you can begin to understand how your business is perceived in the AI landscape and what you can do to enhance that perception.

Benchmark FAQ

What does the zero score mean?

It means the defined five-question, four-engine, two-sample baseline recorded no VisiScan mentions on July 24, 2026. It is a dated observation, not proof that the domain can never be retrieved or recommended.

Can this benchmark be compared with a later scan?

Yes, if the later run uses the same engines, questions, sample count, and interpretation rules. Record provider failures and measurement mode separately so a changed denominator is not mistaken for a visibility change.

Sources

• VisiScan methodology — scoring, sampling, provenance, and limitations.
• VisiScan research hub — citation-source observations and industry context.
• VisiScan AI Visibility Index — public score presentation and grouping.