Why Zygur Scores 0/100 on Its Own AI Visibility Tool

If you run the Zygur URL through Zygur's own AI visibility scoring tool, you get a 0 out of 100. No partial credit. No sympathy points. Zero. For a company whose entire product is helping businesses get found by ChatGPT, Claude, Perplexity, and Gemini, that number looks like a punchline. It isn't. It's actually the most honest result the tool could return — and understanding why tells you more about how AI visibility works than a perfect score ever would.

What the Score Actually Measures

The Zygur AI visibility score measures how well a business's web presence is structured, discoverable, and trustworthy to AI systems that are actively recommending businesses to users. That includes large language models like ChatGPT and Gemini, answer engines like Perplexity, and the growing ecosystem of AI agents that search and summarize on someone's behalf. The score is built on real signals: structured data, content clarity, entity recognition, citation potential, and whether the site gives AI systems enough to work with to form a confident recommendation.

Why a New Site Scores Zero

Zygur.com launched recently. The site is sparse by design right now — the free scoring tool is live, the product direction is documented, but the kind of content depth that earns AI citations simply doesn't exist yet. Perplexity can't cite what isn't there. Claude can't confidently recommend a business it has never encountered in training data or live retrieval. ChatGPT won't surface a company name in response to a user query if there's no credible external signal that the company is real, established, and relevant. A score of zero means: the signals aren't there. That's it.

The Counterintuitive Part

Here's the thing most people get wrong about AI visibility: a zero score is not failure. It's a baseline. The tool is working exactly as designed when it returns a 0 for a site that genuinely has no AI-readable signals. The failure mode would be if the tool returned a 60 for a site with nothing to show — that would be a broken tool. What makes this useful is that the score is diagnostic. A 0 tells you where to start. A 43 tells you what's partially working and what isn't. The number is only meaningful if it's honest.

What Would Move the Score

To move Zygur's score from 0, the same fixes apply here that apply to any business. The AI generates these automatically for paid subscribers, but the categories are straightforward:

• —Structured data markup (schema.org Organization, Product, FAQ) that gives AI systems machine-readable facts about the company
• —An llms.txt file — a plain-language summary of what the company does, written for AI crawlers and agents, not humans
• —Clear entity definition: a named company, named product, named founder — enough for Gemini or Perplexity to form a coherent entity in their knowledge graph
• —Published content that answers specific questions a prospective customer might ask an AI — not blog posts for SEO, but answers for AI citation
• —Third-party mentions: press, directories, or other sites that corroborate the company exists and does what it says it does

Building in Public Means Scoring in Public

The decision to publish this score openly was deliberate. Zygur is building in public. That means the site will score 0 for a while, then 12, then 34, then higher — and we'll document what changed and why at each step. That progression is the most useful thing we can show anyone who is trying to understand AI visibility: a real company, starting from nothing, getting found by AI systems over time. Not a case study. An ongoing record.

What This Means for Your Business

Most businesses that run their URL through the Zygur tool are not starting from zero. They have websites, content, some existing structure — and they're scoring 20, 35, 51. That range is actually harder to diagnose than zero, because something is working and something isn't, and it's not obvious which is which. That's where the automated fixes earn their place. The AI audits the specific gaps — missing schema, weak entity signals, no llms.txt, thin FAQ content — and generates the exact fixes needed. No consultant. No guesswork.

FAQ

Does a 0/100 score mean a business is completely invisible to AI?

Not necessarily invisible — but it means there are no reliable signals for AI systems to cite, recommend, or confidently surface the business. ChatGPT might still mention a company by name if it appeared in training data, but that's different from being actively recommended in response to a user query. A 0 means you're not set up to be recommended.

How is the Zygur AI visibility score different from an SEO score?

SEO scores measure how well a site is structured for search engine crawlers and ranking algorithms. AI visibility scores measure something different: how well the site communicates to language models and answer engines that the business is a real, credible entity worth recommending. The two overlap in some areas — structured data, content clarity — but the intent and measurement criteria are distinct.

Which AI systems does the score account for?

The score is built against the behavior of ChatGPT, Claude, Perplexity, and Gemini — the four major AI platforms currently being used to research and recommend businesses. It also accounts for AI agent behavior, since agents increasingly conduct multi-step web research on behalf of users and need clear, machine-readable signals to work with.

What is an llms.txt file and why does it matter?

An llms.txt file is a plain-text file placed at the root of a website (yoursite.com/llms.txt) that summarizes what the company does in language optimized for AI systems. It's the equivalent of a robots.txt for AI — a direct communication channel that tells crawlers and agents exactly what the site is and who it serves. Many AI systems are beginning to read and weight these files. A business without one is missing a straightforward signal.

How long does it take to improve an AI visibility score?

Technical fixes — schema markup, llms.txt, structured FAQ content — can be implemented in days and start sending signals immediately. Content-based improvements take longer because AI systems need time to encounter, index, and weight new information. Realistically, meaningful score movement happens over 4–12 weeks, depending on how much the site is being crawled by retrieval-augmented systems like Perplexity.