AI visibility is fast becoming one of the most common problems agencies get asked to fix — and one of the most misunderstood. When you sit down with a client who has already taken a swing at it, you will almost always find the same thing: a new blog post, maybe a FAQ page, and a quiet hope that this was enough. It was not. That is not a knock on the client. It is a reflection of how new and poorly documented the rules of AI search actually are.
The Blog Post Instinct Is Reasonable — Just Wrong
Your client has spent years learning that content drives visibility. Write more, rank higher. It worked on Google. It made sense to apply the same logic to ChatGPT and Perplexity. The problem is that AI systems do not index and rank content the way a search engine does. They synthesize it. A blog post titled 'Why We Are the Best Plumber in Austin' does not teach an AI model anything useful about your client's business. It teaches it that your client writes marketing copy.
What AI Systems Actually Need to Recommend a Business
Platforms like Perplexity and ChatGPT browse the web in real time. Claude and Gemini draw on training data supplemented by live retrieval in some contexts. All of them are trying to answer a specific user question — 'Who is the best accountant for freelancers in Denver?' — and they need structured, unambiguous signals to form a confident answer.
- —Structured data markup (Schema.org) that explicitly declares business type, location, services, and credentials
- —A clear, machine-readable summary of what the business does and who it serves — not a brand story
- —Consistent entity information across the site, Google Business Profile, and third-party directories
- —An llms.txt file that gives AI crawlers a direct, prioritized map of the site's most important content
- —Citations from credible sources — reviews, press, industry directories — that AI models treat as corroboration
The Counterintuitive Part: More Content Can Actually Hurt
Here is something clients do not expect to hear: adding a dozen vague blog posts can reduce AI visibility. When a language model crawls or ingests a site with inconsistent, diffuse content, it forms a fuzzy picture of what the business actually does. Perplexity, for example, pulls from multiple page signals to construct its answer. If half those pages describe your client as a 'full-service digital partner' and the other half as a 'social media agency,' the model has no clear answer to give a user who asks for a social media agency in their city. Clarity beats volume every time.
How to Have This Conversation Without Making the Client Feel Foolish
The goal is not to tear down what they did. It is to reframe the problem. They were solving for SEO. You are solving for AI citation. These are related disciplines but they have different requirements. A useful framing: traditional SEO helps Google find a page. AI visibility optimization helps ChatGPT form a confident, accurate answer about a business. The blog post served the first goal adequately. It did almost nothing for the second.
Start With a Score, Not an Opinion
Running the client's URL through Zygur's free score tool is a more productive starting point than a verbal audit. They get a 0–100 AI visibility score in about 60 seconds. That number anchors the conversation in data rather than judgment. It is much easier to say 'your site scores a 34 on AI visibility' than to explain in the abstract why their FAQ page did not work.
What a Real Fix Actually Looks Like
A complete AI visibility fix addresses four layers. First, structured data — does the site have valid Schema markup that names the business category, service area, and key offerings in terms AI models recognize? Second, entity clarity — is the business described consistently and specifically across every page and every platform? Third, crawlability — does the site have an llms.txt file and clean technical signals that tell AI agents where to look and what matters? Fourth, authority — are there third-party mentions, reviews, or directory listings that corroborate what the site claims? One FAQ page addresses none of these. It is content dressing on top of a structural problem.
The Agency Opportunity Here Is Significant
Clients who have already attempted a fix and failed are ideal prospects for a structured AEO service. They understand the problem is real. They have already spent time on it. They are ready to pay for something that actually works. The conversation is not 'let me explain why AI visibility matters.' The conversation is 'let me show you what was missing from what you already tried.' That is a much shorter sales cycle.
FAQ: AI Visibility for Clients Who Already Tried
Why didn't a FAQ page improve my client's AI visibility?
FAQ pages add readable content but they do not add structured signals. ChatGPT and Perplexity rely on Schema markup, entity consistency, and corroborating citations to confidently recommend a business. A FAQ page in plain HTML does not supply any of those signals on its own.
Which AI platforms actually drive business recommendations?
ChatGPT, Perplexity, Claude, and Gemini are the four platforms most likely to recommend local and niche businesses in response to user queries. Perplexity is currently the most aggressive about pulling live web data. ChatGPT with browsing enabled and Gemini's live search mode are increasingly influential for commercial queries.
What is an llms.txt file and does my client's site need one?
An llms.txt file is a plain-text document placed at the root of a site that tells AI crawlers which pages are most important and how to interpret the site's content. It is the AI equivalent of a sitemap. Most business sites do not have one. Adding it is one of the faster wins in an AI visibility audit.
How long does it take to see results after fixing AI visibility?
Perplexity can reflect changes within days because it crawls in near real time. ChatGPT's browsing mode is similarly responsive. Results in Claude and Gemini may lag longer depending on their retrieval and training cycles. Most clients see measurable score improvement within two to four weeks of implementing structural fixes.
Can I use Zygur to show clients their current AI visibility score?
Yes. The free score tool at zygur.com accepts any URL and returns a 0–100 AI visibility score in about 60 seconds with no account required. It is a practical way to open a client conversation with data instead of abstract explanation.
Check your score free at zygur.com — takes 60 seconds, no account required.