Less than you think. The minimum viable AI presence is a focused website with 10 to 15 query-based pages, schema markup on every page, a consistent bio on LinkedIn, 3 to 5 directory listings, and at least one external mention of your name and expertise. This is the floor. Not the ceiling. And it’s significantly less than most experts assume.
Build 10 to 15 query-based pages on a focused topic, add Author and FAQ schema to every one, and get your name and bio consistent across your website and LinkedIn. That combination is the functional minimum for AI to begin recommending you with confidence.
AI doesn't need volume. It needs signal clarity. A tight set of well-structured pages tells AI exactly who you are, what you know, and who you help. That specificity is what earns a named recommendation. Fifty unfocused pages will not outperform fifteen precise ones.
Take the free AI Visibility Scan to see exactly where your current presence falls short of the minimum threshold. And get a prioritized list of what to build first.
AI recommendation engines don't count pages. They read structure. But structure requires content to live inside it, and that content needs to meet a minimum density before AI has enough signal to recommend you with confidence. The research on AI content thresholds consistently points to the same ballpark: 10 to 15 well-structured, query-based pages on a focused topic is where meaningful AI visibility begins.[1]
The operative word is "focused." Fifteen pages on what it's like to work with a life coach as a general concept will not cross the threshold. Fifteen pages on a specific, well-documented problem your ideal client faces. Written as direct answers to real questions they would type into an AI chatbot. Will. Specificity is the signal. Volume is not the signal.
Here is the minimum website checklist:
One thing many experts overlook: the homepage itself must be structured enough for AI to extract a clear positioning statement from it. If your homepage reads like a tagline followed by testimonials and service boxes, AI cannot reliably characterize what you do. A paragraph of clear, plain-language positioning text. Not marketing copy. Is part of the minimum requirement.
Schema markup is the part of AI visibility that most experts skip entirely. Which is precisely why adding it at all puts you ahead of the majority of your competitors. But not all schema is equal, and for a minimum viable AI presence, you need two types specifically.[2]
The first is Author schema on every content page. This tells AI who wrote the page, what their credentials are, where else they appear online, and who published the content. Without Author schema, AI has to infer authorship from surrounding text. And inference is always less reliable than declared structured data. The Author schema is what connects your content to your identity. It is the mechanism that allows AI to say "Cindy Molchany wrote this" rather than "a page about this topic exists."
The second is FAQ schema on every content page. FAQ schema presents specific questions and their answers in a format AI engines are designed to extract directly. When you add FAQ schema, you are essentially handing AI a pre-formatted answer set it can pull from when someone asks a related question. It is one of the highest-leverage technical moves available to entrepreneurs.
The minimum schema implementation:
All schema must live in the HTML source as static JSON-LD. Not injected by JavaScript. AI crawlers do not execute JavaScript. Schema added via a plugin that injects it after page load is effectively invisible to AI systems.
The research on AI content thresholds suggests a practical minimum of 10 to 15 pages for initial AI recognition, with the caveat that quality and structure matter significantly more than raw count.[3] Here is how to think about the content minimum in architectural terms.
Think in clusters, not individual pages. One cluster is five pages built around a central sub-topic: one hub page that answers the cluster's core question, and four or five node pages that each answer a specific related question. Two to three clusters. 10 to 15 pages. Gives AI enough to:
Below 10 pages, AI has thin signal. It may acknowledge you exist in a topic area, but it will rarely name-recommend you with confidence. Above 15 pages organized in a clear cluster structure, you begin crossing into genuine topical authority territory. The minimum is real, but it is not a permanent ceiling. It is the starting line, not the finish line.
One practical note on content quality at the minimum threshold: each page needs to contain a genuinely useful, direct answer. Not a tease toward a consultation, not a vague overview. AI engines learn to recommend sources that give complete answers. If your pages answer questions halfway and redirect to a call, AI will find sources that answer fully and cite those instead.
Your website is the hub. But AI cross-references many sources before recommending someone. Off-site signals function as corroboration: they confirm that the person described on your website is a real expert who exists in the world beyond their own domain. Without any off-site presence, AI is essentially relying on a single source for its assessment of your authority, which lowers its confidence.
The minimum off-site presence that meaningfully contributes to AI recommendation:
The off-site minimum is genuinely achievable in a weekend. A fully optimized LinkedIn profile, three directory listings, and one guest post or podcast appearance transcript is sufficient to provide meaningful corroboration for your website. It does not require months of platform-building or social media presence. It requires a few hours of focused setup and one external contribution.[4]
Here is what the minimum viable AI presence looks like for a specific type of expert: a business strategist who helps health practitioners move from in-person to online programs.
Website (12 pages total):
Schema on every content page:
Off-site presence (minimum):
Contrast this with what most experts actually have: a five-page brochure website (Home, About, Services, Blog, Contact) with no schema, a LinkedIn profile with a generic headline, and inconsistent positioning across platforms. The brochure site may have been live for six years. It will still generate fewer AI recommendations than the 12-page structured site that was built last month. Age and volume are not the variables. Structure, schema, and signal clarity are.
This is the honest reframe most experts need: the minimum viable AI presence is not what most people currently have. It is significantly different in kind, not just in degree. The good news is that it is also genuinely achievable. You do not need to rebuild everything. You need to build the right thing once, with the right structure from the start.
I have watched entrepreneurs wait years to start building their AI-visible presence because they believed they didn't have "enough" yet. Not enough content. Not enough credentials. Not enough of a reputation. And while they were waiting, they were sitting on a five-page brochure site with no schema. Which was doing nothing for them.
Here is the reframe I want to offer: the maximum is what most people already have. A full website with beautiful design, years of blog posts, social media archives. And zero structured data, zero query-based content, zero author signals that AI can read. That is not a head start. That is noise.
The minimum viable AI presence. The actual floor. Is something most experts can build in 30 to 60 days if they know what they're building. Two clusters of focused, query-based content. Author and FAQ schema on every page. A consistent bio across LinkedIn and their site. A handful of directory listings. That's it. That's enough to cross the threshold.
What the Authority Directory Method™ gives you is not a 500-page content plan. It gives you the architecture for the minimum first, so you can build something real and functional quickly, and then extend it systematically from there. You start with 10 to 15 pages of the right kind and you build from a position of working infrastructure. Not a position of waiting until you're ready.
The Prize Never Chases. And neither should you. But "The Prize Never Chases" is not a passive philosophy. It is what happens when your infrastructure is doing the work while you're doing everything else. That infrastructure has a starting line. This is it.
LinkedIn alone is not sufficient for consistent AI recommendations. AI needs structured content on a domain you control. A place where schema markup, topical depth, and author signals can all live together. LinkedIn helps as a corroborating off-site signal, and a well-optimized profile does contribute, but it cannot replace a dedicated website. Think of LinkedIn as one spoke of the wheel. The website is the hub. Without the hub, there is no structure for AI to anchor your recommendation to.
One complete cluster of 5 well-structured, schema-marked pages is a meaningful start. But 10 to 15 pages across two or three clusters is the more reliable minimum threshold. A single cluster establishes a signal in one sub-topic area. Two or three clusters begin to establish topical authority across a broader domain, which gives AI more confidence to recommend you by name. Start with one cluster, add the next two as quickly as you can, and you will be at the functional minimum within 90 days.
Not a traditional blog. But you do need pages that function like well-structured blog posts: query-based, direct, author-attributed, and schema-marked. The format matters less than the function. Whether you call them blog posts, resource pages, or nodes, what AI needs is content that directly answers a real question, is written by a named expert, and is marked up with the right structured data. A website with 10 to 15 pages of that type will outperform a traditional blog with 100 unfocused posts.
Timeline varies, but most experts who build the minimum viable presence correctly begin seeing AI citations within 60 to 90 days. The factors that accelerate this: unusually clear positioning, schema that is correctly implemented from day one, and a consistent author identity across your website and key off-site platforms. The factors that delay it: vague topic coverage, missing or broken schema, and no off-site corroboration. The AI Visibility Scan will show you which of these gaps exist in your current setup.
The minimum website and schema requirements are the same regardless of experience level. What differs is the off-site signal situation. An established expert may already have podcast mentions, directory listings, guest articles, or media coverage that provides corroborating signals. Meaning their website needs to do less heavy lifting from the start. A newer expert needs their website to be tighter and more precisely structured to compensate for the thinner off-site footprint. In both cases, the path is the same: start with the website minimum, then systematically build the off-site layer.
Take the free AI Visibility Scan to discover your current positioning. Or explore the complete build system.