How to Get Found by AI: A Practical Guide for Business Owners

AI is changing how people search, recommend, and find businesses like yours.

But right now, there’s a lot of confusion - and more than a few myths - about how to show up in those results.

1. Why We’re Talking About This Now

I was at the Female Founder Summit in London this week - full of smart, thoughtful women doing incredible things. And, as tends to happen when I’m around, the conversation kept circling back to AI.

There was one theme I kept hearing:

“I’m training my AI so I’ll be easier to find.”

But the more I heard it, the more I realised:

most people were talking about training their own AI chat - not their public presence.

They were adding memories to ChatGPT or Claude, or feeding in brand documents and FAQs - which is useful, but not in the way they thought.

Here’s the truth: Training your own chat assistant might help you get better results in that one conversation, but it won’t make you more visible to anyone else. It doesn’t teach the internet about you. It doesn’t train the model globally. And it doesn’t improve your discoverability in search or AI summaries.

It’s the same confusion I saw when SEO first took off. Suddenly everyone was selling “optimisation,” and nobody could explain what that actually meant. It was gatekept, overpromised, and underexplained - and people were being charged for work they couldn’t even see.

Now we’re heading into that territory again with AI.

They’re calling it “AIEO” - AI Engine Optimisation - and I can already see the fog settling in. So before the mystery packages and buzzwordy sales pitches start spreading, I want to explain what this actually means - and what you can do, yourself or with help, without needing a translator.

2. What Really Happens When You “Train” an AI Chat (And What Doesn’t)

Let’s get one thing clear: If you’ve uploaded your brand guide to ChatGPT, or fed Claude a few customer FAQs - that’s great. It might help you get more relevant answers. It might even feel like the AI is “learning” you.

But here’s the reality:

That’s not training the global model. It’s not making your content more discoverable. And no one else - not Google, not other users, not other AI tools - can see that content.

It’s like whispering into your own ear and expecting the world to hear it.

So what is happening?

When you interact with ChatGPT or Claude, you’re working inside a single session - like a temporary container. Some tools (like ChatGPT Pro with memory enabled) can remember what you’ve said across conversations, but even that’s just for your own use.

Unless you’ve explicitly agreed to share your data for training - and even then, it’s anonymised and aggregated - that chat doesn’t go anywhere beyond your screen.

You’re not teaching “the AI.” You’re feeding a temporary assistant.

Why this matters

The misconception here is subtle, but dangerous. It leads people to believe they’re doing “AI marketing” simply by talking to their chatbot.

But if your goal is to be found - by AI tools, search eng

ines, or voice assistants - that information needs to be on the public web, in a format machines can recognise.

Schema. Structured data. Clean, consistent metadata. That’s what the models actually read.

3. Where AI Really Gets Its Information

If training your personal chatbot doesn’t make you discoverable - then what does?

The answer is surprisingly old-school:

AI gets its information the same way search engines do - by reading the internet.

Tools like ChatGPT, Claude, Perplexity, and others are trained on publicly available web content. That means:

• Websites

• Blogs

• FAQs

• Public datasets

• News articles

• Reviews

• Trusted directories

• Structured metadata (like schema)

  
  

Basically, anything that’s online, public, and structured enough to make sense.

AI doesn't “see” your website - it parses it

When an AI tool reads your site, it’s not admiring the layout or squinting at your font choice. It’s looking for patterns:

• Does this site mention a person or business name consistently?

• What kind of content is this - a product, a guide, a testimonial?

• Are there links pointing to it from other sites?

• Does the structure of the page help me understand what it’s about?

This is why schema markup, metadata, and clean headings matter - because they help AI understand what it's looking at.

If you’ve ever asked ChatGPT to “recommend a tool for...” and it gave you a real suggestion with a link - that didn’t come from magic. It came from crawlable, structured, high-signal content online.

Public = Discoverable

Here’s the golden rule:

If it’s not public, it’s not helping your discoverability.

That means:

• Private Google Docs → No

• Chat uploads → No

• Password-protected Notion pages → No

• Content buried 5 clicks deep with no backlinks → Also no

It doesn’t mean everything has to be public right now - but it does mean that if you want AI tools to reference you, they need to have access to something meaningful, structured, and visible.

4. What Schema Markup Actually Is (And Why It Matters for AI)

Let’s demystify this one too - because it sounds more technical than it is.

Schema markup is a way of adding extra, structured information to your website so that machines - not people - can understand what your content is about.

If SEO was about guessing what your audience might be searching for, schema is about saying:

“This is exactly what this page is. Here’s what’s on it. Here’s who wrote it. Here’s how it connects to the rest of my site.”

That clarity is what makes your content findable by AI.

Who’s behind schema?

Schema.org is a global collaboration - founded by Google, Microsoft, Yahoo, and Yandex - to create a shared vocabulary for the internet.

“Schema.org is a collaborative, community activity with a mission to create, maintain, and promote schemas for structured data on the Internet, on web pages, in email messages, and beyond.”

It’s open source, constantly updated, and used by over 45 million websites, powering more than 450 billion data objects. It’s not obscure. It’s essential.

Many applications - from Google to Pinterest to ChatGPT - use Schema.org’s vocabulary to interpret, rank, and display web content. So when you use it, you’re speaking the language of the machines.

 Visit Schema.org

What schema actually looks like

Schema markup isn’t visible on your webpage - it sits quietly in the background, usually in a format called JSON-LD. (You don’t need to write this by hand — I’ll show you how to get AI to do it for you in the next section.)

There are different types of schema for different types of content. That’s what we’ll look at next.

5. The Different Types of Schema (And When to Use Them)

Not all schema is the same. The type you use tells AI what kind of content it’s looking at - whether it’s a blog, a product, a person, or a tutorial.

The good news? You don’t need to remember them all. You just need to know the basics - and when each one is useful.

BlogPosting

Use this when: You’re publishing a blog, news piece, article, or opinion.

Why it matters: AI models are more likely to quote or reference articles that have clear attribution, publishing dates, and headlines.

Person

Use this when: You’re writing an About page or introducing yourself as a founder, therapist, coach, or specialist.

Why it matters: Helps AI understand you’re a real human, not just a brand - which builds credibility and can link other pages to you as an author or expert.

Organization

Use this when: Describing your business, charity, or company.

Why it matters: Makes your brand more “visible” to AI models - can also support branded search, help with Knowledge Panels, and reinforce legitimacy.

Product

Use this when: You’re offering something people can sign up for, buy, or try (like the Inkie platform or a course).

Why it matters: Lets AI understand your product, pricing, free trials, and links to it. Increases the chance of being included in comparisons or recommendations.

HowTo

Use this when: You’ve written a step-by-step tutorial, explainer, or process.

Why it matters: AI tools love instructional content. “How to...” queries are common in search, and using this schema makes it easier for your content to be summarised in answers.

FAQPage

Use this when: You have a page answering common customer questions in a Q&A format.

Why it matters: Can get picked up by Google and AI to directly answer questions - and show up as featured snippets or source links in AI results.

You can use one schema type per page - and choose the one that most accurately describes what’s there. Don’t try to tag everything just for the sake of it. AI likes clarity, not keyword salad.

6. How to Add Schema Without Writing a Line of Code

This is the part where most people think they’ll need a developer. But the truth is:

You don’t need to code - you just need to know what you’re asking for.

Schema markup is written in a format called JSON-LD. That sounds scarier than it is. It’s just a neat little block of text you copy and paste into your website.

Use AI to Generate Your Schema

You can use ChatGPT, Claude, or even your own Inkie assistant to create it for you. Here’s a prompt that works well:

“Write structured data in JSON-LD format for a [type of page], including the page title, description, author, date published, and image. Use Schema.org vocabulary.”

Example:

“Write BlogPosting schema in JSON-LD for a blog titled ‘How to Get Found by AI,’ written by Sophie Boulderstone, published by Inkie, with a featured image and link to www.inkie.ink/blog.”

AI will give you a neat block of schema markup. All you have to do is paste it into your page’s Structured Data field.

Where to Put It

If you’re using Wix:

• Go to your page settings

• Open SEO (Google) → Advanced SEO

• Look for Structured Data

• Paste in the code

• Give it a name like “Homepage Schema” or “HowTo Guide Schema”

If you’re using WordPress or another CMS, there are free plugins (like RankMath or Yoast) where you can do the same thing.

You don’t need to style it. You don’t need to explain it. You just need to make sure it’s clean and valid.

Test It Before You Publish

Use Google’s Rich Results Test to check your markup. Paste your code or page link in - it’ll tell you if everything’s correct. https://search.google.com/test/rich-results

7. Why Content Alone Isn’t Enough

Here’s a truth that doesn’t get said often enough:

It’s not just what’s on your website - it’s what’s around it.

You can write a beautiful blog, tag it with perfect schema, and still struggle to be found - by search engines or AI tools - if you haven’t shown that your site is part of something bigger.

AI models (like ChatGPT or Claude) don’t just read your site in isolation. They’re looking for credibility, repetition, and signal strength across the web.

What Are They Looking For?

Signal	What It Tells AI
Backlinks	Other sites linking to yours → “This is worth referencing.”
Social links	Your content is being shared publicly → “This is current and active.”
Press mentions	You’re being cited elsewhere → “This person or brand is recognised.”
Testimonials or reviews	Trustpilot, Google, etc → “This is a real business with real users.”
Consistent identity	Same name, logo, description → “This matches what we’ve seen elsewhere.”

All of these reinforce what your schema is telling AI. They confirm that your content is original, your business is active, and your message is consistent.

AI Links the Dots - So Make the Dots Easy to Find

It’s not enough to just write about your work. You need to:

• Link your website to your social media - and your socials back to your website

• Be consistent with your name, business name, and what you do

• Get a few reviews on platforms people trust - like Google or Trustpilot

• Share things publicly (a blog, a post, a mention - anything that shows you're out there)

• Use your About page well - say who you are, not just what you offer

You don’t need tech platforms or big PR. But you do need to show you’re real.

If a human can’t tell what you do from your online footprint, neither can AI.

8. Final Thought: Make It Clear, Not Complicated

You don’t need to become a tech expert. You don’t need to learn JSON or chase every algorithm update. But you do need to be legible.

That means:

• Telling the internet what your content is, and who it’s for

• Showing you're a real person or business, not a placeholder

• Giving search engines and AI tools the structure they need to understand you

If you’ve done your bit - added schema, kept things consistent, and stayed visible - then you’ve already done more than most.

This isn’t about chasing trends. It’s about making sure your work isn’t missed.

And the good news?

Every blog Inkie creates and auto-posts includes built-in schema markup - automatically. No plugins, no copy-pasting code, no guesswork.

You can try it today, with a 7-day free trial:  www.inkie.ink

Let it help you show up - clearly and consistently.