I presented at SEOCharity Summer Edition, live in front of an audience and shared exactly how I got AI models to start naming me. Not a client. Me. My own name, my own business, my own experiment.
If you have ever typed a question into ChatGPT, Perplexity or Google’s AI Overview and wondered why your competitor showed up and you did not, this post is the answer. Getting cited by ChatGPT is not luck, and it is not a secret prompt. It is a four-phase workflow, and I have now run it twice: once for a healthcare client whose domain authority went from 15 to 28, and once on myself, in public, over six months.
Key Takeaways
- Getting cited by ChatGPT follows a four-phase workflow: fix your own website, build content for AI, publish everywhere, then test and measure.
- GEO and AEO sit on top of SEO, not instead of it. AI models still need a technically sound, well-structured website to trust before they will cite it.
- I ran this exact workflow on myself, and it took six months before I started showing up by name in Perplexity, Google AI Mode and ChatGPT-style answers.
- Consistency across platforms (your site, Reddit, Quora, Medium, LinkedIn) matters more than any single “viral” piece of content.
- You test citation the way I did: ask a clean-slate LLM with no memory of you, and see what it actually says.
What does it mean to “get cited” by ChatGPT?
Getting cited means an AI model names you, your business or your content directly in its answer, sometimes with a link, sometimes just your name and what you do. It happens when the model has crawled or retrieved content about you from multiple credible sources and decided you are a relevant answer to someone’s question, not when you have simply written a lot about a topic.
This is different from ranking on Google. On Google, you are competing for a blue link. On an AI model, you are competing to be the answer itself, or one of a handful of names mentioned inside it. I found out mine were being named when I asked a “clean slate” AI, one with no memory of me, who works at the intersection of SEO, content and AI in the UK. My name came back. More on that below, because honestly, that moment surprised me more than I expected it to.
How is GEO different from traditional SEO?
Generative Engine Optimisation (GEO) is the practice of structuring your content and your presence so that AI models can find, trust and cite you. Answer Engine Optimisation (AEO) is the closely related discipline of writing content that directly answers the exact question someone is asking. Both sit on top of SEO. They do not replace it.
In short: you cannot skip SEO and jump to GEO. AI models still favour technically sound, well-structured, genuinely useful websites when deciding who to cite, so the foundation work matters just as much as it always did.
Phase 1: How do you get your own website ready before chasing AI citations?
I call this “your own house” for a reason. Before I touch content for AI, I fix what is already there. This is the phase most people skip because it is not glamorous, and it is exactly why most people never get cited.
On-page, I go through every page and check for:
- A keyword-led title and H1
- A proper meta title and meta description
- Structured H2s that actually reflect how someone searches
- Alt text on every image
- Anchor text that means something (not “click here”)
- Excerpts and tags that are filled in, not left blank
- Internal linking that connects related content
- Credits and external links to sources, because AI models reward pages that show their working
Technical optimisation, I check whether the site is actually built for bots and AI agents, not just human visitors:
- Page speed
- Indexing status
- Canonicals set correctly
- Redirects cleaned up
- Core content served in raw HTML, not hidden behind JavaScript rendering, because some AI crawlers still cannot see JS-rendered content
- An llm.txt file, the newer equivalent of robots.txt, telling AI crawlers what they can access
I use Sitebulb to audit all of this. It is not the only tool that does the job, but it is the one that lives on my laptop, and it catches the things I would otherwise miss on a tired Tuesday.
Phase 2: What does a content workflow built for AI citation actually look like?
Content built for AI citation starts with real questions, not assumptions. You pull queries from Google Search Console, do proper keyword research, use query fan-out prompts to see the related questions AI models generate, and then go and listen to where people are actually talking: Reddit, Quora, Medium, and the “People Also Ask” box.
Once I know what people are really asking, I feed AI a structured prompt with everything it needs: the goal of the piece, the audience, the intent behind the search, the keyword clusters, and the content gap I am trying to fill. I align all of it with brand guidelines and my own expertise, because AI can draft fast, but it cannot draft as me. That part is still mine.
Then it is proofread, publish, optimise, measure. In that order, every time. I do not publish anything I have not read twice, and I do not consider a piece finished the day it goes live. This is where SEO, AEO and GEO strategies get blended together, all wrapped in the brand’s actual voice and tone, because a technically perfect page that sounds like nobody will not get cited either.
Phase 3: Why do you need to publish beyond your own website?
AI models pull from more than your own site. If the only place you exist online is your website, you are giving the model one source to trust instead of ten. This is why Phase 3 is about showing up on Medium, Reddit and Quora, on the social platforms your audience actually uses (Instagram, Facebook, LinkedIn, YouTube, TikTok), and on relevant sites within your niche.
The rule I hold myself to: same content, same positioning, uniform everywhere. Not copy-pasted, but recognisably you, wherever someone finds you. Show up consistently. It takes time, and there is no way around that. I want to be honest about that part, because most content advice skips it.
Phase 4: How do you actually test whether you are being cited?
You check with a clean-slate LLM, one with no memory of previous conversations and no bias from you having talked to it before, and you ask it a question your ideal customer would ask. Then you test, experiment, and adjust based on what comes back.
The honest answer to “how long does this take” is: mine took six months. Not six weeks. Six months of Phase 1 through 3, repeated, before Phase 4 gave me a result worth screenshotting. If someone tells you it happens overnight, they have not actually run the experiment.
What happened when I ran this workflow on myself?
I have done this before for a client. A healthcare brand I worked with went from a domain authority of 15 to 28 in six months, built on exactly this kind of forum research and content workflow, and it eventually started appearing in LLM-generated answers. That result was exciting, but it was also, in a way, easy to be sceptical about. It was one client, one niche, one set of circumstances.
So this time, I ran the whole thing on myself. My own website, my own content, my own name, in public, for anyone to check.
Six months in, I asked a clean-slate AI to share a list of people based in the UK working at the intersection of SEO, content and AI. I came back at number one, described as working in “SEO, content strategy, AI-assisted marketing workflows, GEO, n8n SEO workflows,” flagged as an early, growing personal brand. Fair enough. That is exactly where I am.
Then I asked Perplexity a similar question. I came back at number twenty on a list of over a hundred sources, described as the founder of GrowwDigitaly, an AI-driven SEO and content strategist, with my AirOps interview and Sitebulb article cited as a source.
Then Google’s AI Mode. This one genuinely caught me off guard. Under a section titled “Real Operators: The Verified Professionals,” it opened with: “Leading this selection is Shweta Gupta, founder of GrowwDigitaly, who actively bridges traditional search optimisation with custom AI frameworks, LLM readiness, and content strategy.” Leading the selection. I read that line three times.
A separate Google AI Mode answer, this one about B2B demand generation and content hubs, placed me further down the list but still there, still described accurately: “Founder of GrowwDigitaly, a UK-based agency specialising in content strategy, SEO execution, and AI-assisted marketing workflows.”
Here is the part I will not pretend was glamorous: my Reddit profile, one of the places I was consistently showing up and answering real questions, has two followers. And it still got picked up as a source. That is the bit nobody tells you about GEO. You are not chasing followers or virality. You are chasing consistency, in the right places, said in a way that is unmistakably you, so that when an AI model goes looking for evidence, it finds the same story everywhere it checks.
What is Search Everywhere Optimisation, and why does it matter now?
Search Everywhere Optimisation is the idea that visibility no longer lives on one search engine results page. Your audience is asking questions on Google, ChatGPT, Perplexity, Reddit, YouTube and inside AI Overviews, and you need a presence built for all of it, not just the one you are used to optimising for.
This is not a rebrand of SEO. It is an honest acknowledgement that “ranking on Google” was never really the goal. Being found, trusted and recommended was always the goal. Google just used to be the only place that happened. It is not anymore.
Frequently Asked Questions
How long does it take to get cited by ChatGPT?
There is no fixed number, but in my own experiment it took six months of consistent work across all four phases before I started appearing by name in AI-generated answers. Anyone promising results in weeks is not describing a real workflow.
Do I need to be active on Reddit and Quora to get cited by AI models?
Not exclusively, but it helps enormously. AI models weigh up multiple sources before citing someone, and forums are where real, unfiltered conversation happens. My Quora activity, despite a tiny following, was cited as a source in at least one of my results.
What is the difference between AEO and GEO?
AEO (Answer Engine Optimisation) is about writing content that directly and clearly answers a specific question. GEO (Generative Engine Optimisation) is the broader practice of structuring your entire content and platform presence so AI models can find, trust and cite you. AEO is largely about the sentence. GEO is about the ecosystem around it.
Can a small business realistically get cited by AI models, or is this only for big brands?
It is realistic for small businesses and solo founders. My own experiment was a business of one, with a two-follower Reddit profile, and it worked. What matters is consistency and genuine expertise, not budget or team size.
Do citations from ChatGPT and Perplexity actually bring in customers?
They build trust and visibility at the exact moment someone is asking the question your business answers, which is a stronger position than a generic ad. I cannot promise a direct line from citation to sale for every business, and I would not want to overclaim that, but visibility at the point of genuine intent is valuable by any honest measure.
If your brand is not showing up in AI answers yet
I built this workflow because I was tired of guessing, and I tested it on myself because I did not want to sell anyone a strategy I had not lived through first. It worked, slowly, honestly, and with a Quora profile that still only has two followers.
If you have a website, you have content, and you are still not showing up when your customers ask AI models the questions your business should be answering, that is exactly the gap I work in. No templates, no secret sauce, just the same four phases, built around your business instead of mine.
Get in touch through growwdigitaly.com and let’s find out what your version of this looks like.
Shweta Gupta is a UK-based marketer specialising in SEO, content strategy, and AI-assisted marketing workflows. She has experience across SEO, paid media, content optimisation, website management, and digital marketing across various industries.