Most agencies treat an AI visibility score the way they treat a website audit — run it once, hand it over, invoice, move on. That's a waste. A client's AI visibility score is a living number that changes every time ChatGPT updates its training data, every time Perplexity recrawls the web, every time your client's competitor publishes a better-structured page. Used correctly, that score becomes the anchor for a monthly reporting cadence that clients actually understand — and actually pay for month after month.
Why a One-Time Audit Doesn't Hold Up
A traditional SEO audit captures a moment in time. AI visibility is worse — it's a moving target tied to systems that change without announcement. Gemini's grounding behavior shifted noticeably in early 2025. Perplexity added more aggressive source citation weighting. Claude began pulling more heavily from structured, authoritative pages when answering commercial queries. If you scored a client's site in January and filed the report, that score is already stale. You're not auditing a static document. You're auditing a relationship between a business and a set of AI systems that are continuously re-evaluating it.
What the Score Actually Measures
Zygur's 0–100 AI visibility score evaluates a business's website across the factors AI systems use to decide whether to cite, recommend, or surface it in a response. That includes structured data completeness, content clarity, topical authority signals, entity recognition, and technical accessibility for AI crawlers. It's not a PageRank proxy. A site can rank on page one of Google and score a 31 on AI visibility — because the signals AI systems care about are meaningfully different from what search engines have rewarded for the past twenty years.
The Counterintuitive Part
Here's the thing most agencies miss: a low score is not a problem to solve once and close. It's a baseline to beat every month. Clients who start at a 28 and watch it climb to a 54 over three months are seeing the work. That number is the story. You don't need a twelve-slide deck explaining algorithm changes — you need one number that went up, and a clear explanation of why.
Building the Monthly Reporting Anchor
The structure is simple. At the start of the engagement, run the initial score. That's month zero. Every month after that, you run the score again, compare it to the prior month, and report on three things:
- —The score delta — did it go up, down, or hold steady, and by how much
- —What changed — which fixes were applied and what AI systems are now more likely to cite the client
- —What's next — the top two or three issues still dragging the score down, and what you're doing about them
That's a complete monthly report. It's not padded with keyword ranking tables the client doesn't read. It's not buried in a PDF they'll open once. It's a number, a story, and a plan. Clients who understand that ChatGPT now mentions their business when someone asks for a local accountant in their city — that's retention.
What to Fix Between Reports
Zygur generates the fixes automatically. Every score comes with AI-generated recommendations — structured data additions, content rewrites for clarity, schema markup, entity disambiguation, and technical fixes that improve how systems like Claude and Perplexity parse the page. You implement the fixes, rescore at the end of the month, and report the delta. The client doesn't need to know it took forty-five minutes. They need to know their score went from 41 to 58 and that three more AI platforms are now citing them in relevant queries.
Pricing It as a Retainer
This is where the model pays off for agencies. A one-time audit is a project. A monthly score anchor is a retainer. The value proposition is continuous — AI systems keep changing, competitors keep optimizing, and someone needs to keep your client visible. Monthly scoring plus fix implementation plus reporting is a clean recurring service that sits naturally alongside existing SEO or content retainers. Or it stands alone as an AEO-specific offering for clients who are already satisfied with their SEO and want to address AI search as a separate channel.
FAQ
How often should I rescore a client's AI visibility?
Monthly is the right cadence for most clients. It gives you enough time to implement fixes and see meaningful movement, without letting the score go stale for so long that you lose the narrative thread. For clients in competitive categories where multiple AI platforms are frequently recommending competitors, bi-weekly rescoring is reasonable.
What AI systems does the Zygur score account for?
The score reflects visibility factors relevant to ChatGPT, Claude, Perplexity, and Gemini — the four platforms where most commercial AI queries are happening right now. It also accounts for AI agents and automated systems that retrieve and synthesize web content on behalf of users.
Can a site have great SEO rankings and a low AI visibility score?
Yes, and it happens constantly. Google's ranking signals have been trained on two decades of web behavior. AI systems like Perplexity and Claude are optimizing for something different — structured, authoritative, entity-rich content they can cite with confidence. A site built entirely around keyword density and backlink volume can rank well in Google search and be nearly invisible to AI platforms.
How do I explain AI visibility scoring to a client who's never heard of AEO?
Ask them to open ChatGPT and type: 'Who are the best [service category] in [their city]?' If they're not in the answer, that's the explanation. The score tells you why they're not there. The monthly work moves the number until they are.
Is the Zygur score free to run?
The initial score is completely free — enter any URL at zygur.com, get a 0–100 score in about 60 seconds, no account required. Monthly monitoring and automated fix generation are part of the paid toolkit, which is what you'd use to run client retainers at scale.
Check your score free at zygur.com — takes 60 seconds, no account required.