Get SEO Ranked by LLMs: How to Make Sure ChatGPT and Perplexity Have the Correct Information About Your Business
- Katina Ndlovu

- 21 hours ago
- 5 min read
Large language models like ChatGPT and Perplexity decide which businesses get surfaced based on clarity, structure, and trust.
If your website does not explicitly explain who you are, what you do, and how your services work in a format AI systems can reliably interpret, those systems will either misrepresent your business or ignore it entirely.
This guide documents a practical, repeatable method for making sure LLMs have correct, verifiable information about your business using real AI-generated questions, structured answers, and validated FAQ schema.
Written by Katina Ndlovu, a marketing strategist focused on SEO, AEO, and AI-ready content systems.

Get SEO Ranked by LLMs
Traditional SEO is dying.
That does not mean search is disappearing. It means how information is discovered, summarised, and trusted has changed.
Large language models like ChatGPT, Perplexity, and Google’s AI systems no longer rely solely on blue links and keyword density. They look for clear, structured, verifiable answers. If your website does not explicitly explain who you are, what you do, and how your services work in a machine-readable way, you are effectively invisible to AI-driven discovery.
This is where most businesses fall behind.
They publish blogs.
They optimise pages.But they fail to tell AI systems what their business actually means.
This guide documents a practical method I use to correct that gap and get SEO ranked by LLMs.
It shows how to:
Extract real, AI-visible questions about your brand
Turn those questions into controlled answers
Convert those answers into structured data LLMs can reliably interpret
Validate everything before it ever touches your site
This is not theory. It is a repeatable process designed to reduce ambiguity and increase trust signals across both search engines and language models.
Why LLM visibility now matters more than rankings
When someone asks ChatGPT, Perplexity, or Google AI:
“What does this company do?”
“Is this business trustworthy?”
“How does this service work?”
“Who should I choose?”
The model does not browse your site like a human. It reconstructs your business from signals.
Those signals come from:
Structured data
Consistent question-answer patterns
Clear entity definitions
Verifiable on-page explanations
If those signals are missing or inconsistent, the model fills in the gaps itself. That is where misinformation, poor summaries, or complete exclusion happens.
FAQ schema, when done correctly, is not about rich results.
It is about controlling how AI systems understand your business.
The method to Get SEO Ranked by LLMs
Search Perplexity for your business name, and take a screenshot of the 'Related Questions' that appear at the bottom of the results
Ask Perplexity for more 'Related Questions' and screenshot those too.
In ChatGPT, upload the 2 screenshots along with the first prompt below.
Check its answers, and make corrections if necessary. The, use the second prompt below to get your code.
Check your code in Google Rich Results Test (make sure to select code)
Once it passes the test, paste the code into your website (see steps for various website platforms below)
A) If your site is WordPress
Method 1: Add JSON-LD directly to the page (easiest)
Go to Pages → choose your FAQ page
Click Edit
Switch to Code Editor (top right options → “Code Editor”)
Scroll to the bottom of the page
Paste your JSON-LD <script> block
Save
Method 2: Inject into the <head> site-wide
Install plugin: Insert Headers and Footers (or WPCode)
Go to Settings → Insert Headers and Footers
Paste JSON-LD inside the Footer Scripts box
Save
This works if you want the FAQ data to appear everywhere or on custom templates.
B) If your site is Webflow
Open Webflow
Choose the page (e.g., FAQ, About, Services)
Go to Page Settings
Scroll to Inside <head> tag
Paste your JSON-LD <script> block
Publish the site
C) If your site is Squarespace
Open the page
Click Edit
Add a Code Block
IMPORTANT → Set block type to “HTML”
Paste the script
Hit Save
OR to add to <head>:
Go to Settings → Advanced → Code Injection → Page Header Code Injection
Paste the JSON-LD
D) If your site is Shopify
Option 1: Add JSON-LD to a page
Go to Online Store → Pages
Edit your FAQ page
Click Show HTML (<> icon)
Paste the JSON-LD
Save
Option 2: Add it to the theme
Go to Online Store → Themes → Edit Code
Open theme.liquid
Paste the JSON-LD before </head> or </body>
Save
E) If your site builder only allows “Custom HTML” blocks
(GoDaddy, Leadpages, ClickFunnels, etc.)
Almost all page builders have a:
“HTML block”
“Code snippet”
“Embed code”
“Custom script” module
Steps:
Add a HTML or Code block
Paste the entire <script type="application/ld+json"> block
Save
Make sure the text still renders on the page (the JSON-LD does not show visibly)
How to verify that it’s working
Test with Google’s Structured Data Tool:
Paste your URL and check:
FAQPage schema detected?
No errors?
Structured data is valid?
If it validates, you’re good.
The prompts
Initial Prompt
I will be creating a JSON-LD / schema.org markup for my website. I have attached some pictures containing related search questions.
Before we create the JSON-LD, I want you to provide draft answers to these questions in 1 or 2 sentences each, to the best of your ability. I will then refine your answers.
Second Prompt
Create the JSON-LD / schema.org markup
Why this method works for LLMs specifically
This workflow aligns with how language models ingest and prioritise information:
Perplexity surfaces AI-salient questionsThese are not random FAQs. They are questions already being generated and reinforced by AI systems.
Human-reviewed answers prevent hallucinationYou are correcting the model before it creates its own interpretation.
JSON-LD creates a stable knowledge structureSchema is not about appearance. It is about machine certainty.
Validation removes ambiguityIf Google’s parser can read it cleanly, LLM pipelines can ingest it reliably.
This is Answer Engine Optimisation in practice, not as a buzzword.
Common mistake I see businesses make
They assume AI “figures it out”.
It does not.
AI summarises what it can clearly understand. When your site lacks explicit definitions, the model relies on third-party mentions, outdated directories, or generic assumptions.
That is how:
Service descriptions become vague
Expertise gets flattened
Brand differentiation disappears
This method flips that dynamic. You define the narrative first.
How I use this in real client work
For service-based businesses, consultants, and founders, this process becomes part of a larger
visibility system:
Brand definition
On-site content clarity
Schema governance
AI-readable trust signals
It is especially effective for businesses operating in competitive markets like Sandton and Johannesburg, where being “another option” is not enough. AI needs to understand why you are the right option.
Closing: where this fits in your broader strategy
If you are still thinking about SEO as “ranking pages,” you are already behind.
The real shift is interpretation over indexing.
This method helps ensure that when someone asks an AI about your business, the model does not guess. It answers correctly, confidently, and consistently.
If you want help implementing this properly — refining the questions, tightening the answers, and making sure your schema aligns with your actual service offering — that is the work I do.
You do not need more content.You need clearer signals.
You can reach me through my site and start with a simple question about your visibility setup. From there, we can decide what actually needs fixing.
About the Author
Katina Ndlovu is a marketing strategist and founder of Katina Ndlovu Agency, based in Sandton, Johannesburg. She specialises in search visibility, AI-readiness, and structured content systems that help businesses remain understandable and credible in an LLM-driven search landscape.



Comments