In a recent BNI meeting I attended, the key topic of the presentation was AI Search, or AEO, or GEO, or as most people would ask:
“How do I get found on ChatGPT?”.
“Getting found by AI” sounds like a technical problem; a setting to change, a tag to add, a box to tick; or something to ask my web-developer about. It isn’t. The way an AI engine decides whether to name your business is actually quite close to the way a cautious human-buyer decides whether to call you.
After all, if we are training AI to be human-like, then it is not hard to conceive that it will be like us when it comes to searching for answers to a question.
It reads what you say about yourself. Then it goes and checks whether anyone else says the same thing.
You don’t get found by AI. You get corroborated.
That’s the short answer.
An AI engine will only name a business it can confidently identify and confidently vouch for. That takes two things: a clear, consistent story about who you are, and independent sources that back that story up. One without the other doesn’t work. A consistent story nobody else repeats is a brochure. A pile of mentions that contradict each other is noise.
Why consistency is the floor, not the ceiling
Before a machine can recommend you, it has to work out who you are. Your website says “growing businesses.” Your LinkedIn says “SMEs.” An old directory listing still has the trading name from 2019. A human prospect squints and moves on. AI models do something worse: they hedge, or name someone clearer instead. (Remember, AI seeks to give you the answer you want, the fastest way possible)
This isn’t new thinking, and the evidence for it is actually shifting in an interesting way. Whitespark’s Local Search Ranking Factors report (the industry’s benchmark survey, running since 2008) added a category for AI Search visibility for the first time in its 2026 edition. Darren Shaw’s own analysis of the results: “Citations as a local search strategy have been decreasing in importance for the past decade, but AI SEO is breathing new life into them. 3 of the top 5 AI Search visibility factors are citation factors.” His summary of what that means: “In AI SEO, mentions (citations) are the new link.”
Read that as: consistency’s old job, winning you the local pack, was fading. Its new job, making you legible to a machine, is exactly what’s replacing it. Or, just updating your website isn’t enough anymore… What about your yellow pages listing? Your Google Profile? The Online Trade magazine listing?
What actually gets you named
If you are a referral business, this is where it really gets interesting… Because AI has become your new referral partner. And how well it refers you is something you can help just like a human referral partner.
Ahrefs studied 75,000 brands and found the strongest correlations with AI Overview mentions were all off-site factors (i.e. information NOT on your website): branded web mentions, at 0.664, correlated roughly three times more strongly than backlinks (a figure other researchers have put at 0.218 on the same dataset). Ahrefs’ December 2025 follow-up, extending the same test to ChatGPT, found something I have always only known anecdotally: YouTube mentions correlated more strongly with AI brand visibility than any other metric measured. While a site’s own link volume and page count showed only weak correlation.
In other words, having mentions on YouTube, through video mentions like in reviews, or testimonials really helps with your visibility.
Semrush’s 2026 AI Visibility Index backs the same pattern from a different angle: a company’s AI narrative is no longer shaped solely by owned websites and brand-controlled content; AI platforms also rely on customer reviews, community discussions, independent publishers, retailers, and industry sources to understand and recommend brands.
The machine asks around. Which is exactly what your referred prospects have been doing for years. Bob tells them to call you. Then they check: on Google, on LinkedIn, in a group chat Bob will never see. 83% of B2B decision-makers self-research before ever speaking to sales (Reddit x SurveyMonkey Business Decision-Maker Survey, Dec 2025–Jan 2026), and when they research, peer recommendations are the most trusted source at 73%, ahead of vendor websites at 55% (same survey).
The AI engine has simply joined the queue of people checking up on you.
Specific beats polished
The only proper academic study of this (not an industry blog; an actual peer-reviewed paper out of Princeton, Georgia Tech and the Allen Institute for AI) tested nine content changes against 10,000 real queries to see what actually shifted whether AI engines cited a page. The winner: content-level characteristics associated with epistemic authority, namely statistics, source citations, and quotations from credible authorities, improved visibility by up to 40%, while classical SEO signals like keyword density showed minimal influence.
Translation: AI likes citable sources. If you claim something, make sure you have a credible source to back it up. For example, in this blog I have ensured that any finding I have provided is attributed to its source, be it Ahrefs, or Reddit, or otherwise mentioned as an opinion; and this is more important than stuffing keywords into your blogs.
Here’s the detail I actually love: a page ranked fifth in traditional search saw a 115% increase in AI visibility just from adding citations, while the top-ranked incumbent’s AI visibility share dropped 30% in the same test. Big backlink profiles don’t automatically win here. Specific, well-sourced content does. That’s the best news a founder-led business with no SEO budget is going to get all year.
Which is a fairly blunt instruction: vague content is unciteable. “We deliver exceptional service to our valued clients” gives a machine nothing to quote and a human nothing to believe. A named client, a specific before-and-after, a number, a sentence in someone else’s voice: that’s quotable. It’s also what makes a testimonial worth watching instead of skipping.
The honest caveats
Anyone selling you certainty here is selling you something.
AI referral traffic is still small. Google currently sends around 190 times more traffic to websites than ChatGPT does (Ahrefs, Dec 2025). For most Australian SMBs, Google and referrals still dominate. This is a positioning play for the next few years, not a replacement for what’s working now. Who knows what will happen in the future, but my opinion is that it will only grow.
There are no AI rankings. Answers shift by engine, by prompt, by month. Nobody can guarantee you a spot in ChatGPT. Even if you were to ask ChatGPT on 2 separate occasions the same question, the answer may be different.
Being cited and being named are two different things, and the gap is bigger than you’d think. A 2026 Semrush study tested this directly and found 62% of AI citations are “ghost citations”: the site gets a source link, but the AI never actually says the brand’s name in its answer. And even when a link is served up, hardly anyone clicks it. Separate Pew Research Center analysis of real browsing data found people clicked a traditional search result in just 8% of searches where an AI summary appeared, and clicked the link inside the summary itself in just 1% of visits. So the goal was never traffic. It’s being the name in the sentence, not the link underneath it.
(Worth noting: Google has publicly pushed back on the Pew figures, arguing the study period overlapped with unrelated testing. The direction of the finding still holds up across the wider research, but it’s disputed, and disputed evidence should be presented as disputed.)
Just reiterating, being cited just means your content from your website, post on LinkedIn or Reddit, fed the answer. Being named means the AI actually says your business by name; and that usually only happens if someone follows up and asks specifically who the source was, which isn’t guaranteed.
So what do you actually do?
Nothing exotic. Mostly things you already own.
Tell one story, everywhere. Same description of what you do, who you help and why it matters: website, LinkedIn, Google Business Profile, every directory you’ve forgotten you’re listed in.
Get the trust off the phone and onto the record. The reference that goes out verbally, the thing the referral partner says about you in a room; a machine can’t read either. Captured, it can.
Make your proof specific. Names, numbers, before and after, in the client’s own words. Not because it’s persuasive copywriting; it’s the only kind of content a generative engine finds worth quoting.
Then check what a stranger, or a machine, can actually see. Not what you know is true. What’s visible from outside.
That last step is what the Trust Diagnostic is all about. It looks at every channel a business owns through the eyes of someone encountering it for the first time. It turns out that gap now has a second audience; your friendly neighbourhood AI.
Trust you’ve earned but never made visible is invisible to your prospects. It’s also invisible to the machine they’re about to ask.
If you want to know what a stranger, or a machine, can currently see of your business, that’s what a Trust Diagnostic is for. It is the first step of the Trust Strategy Framework to make sure that the trust you already earned is clearly and consistently visible to anyone looking for you.
