How to Get Cited by AI Search Engines When Backlinks No Longer Decide It.

AI search engines cite the brands that are already relevant, trusted, and talked about across the web, not the ones sitting on the biggest pile of backlinks. When Ahrefs studied 75,000 brands, the strongest predictor of showing up in Google's AI Overviews was how often a brand gets mentioned around the web, which correlated at 0.664, while the raw number of backlinks came in far behind at 0.218. That gap changes what you actually work on if you want ChatGPT, Perplexity, Gemini, or Google's AI answers to name your business. This piece covers what a citation is, why your current SEO only takes you part of the way, and the specific things that make an AI engine comfortable putting your brand in an answer.

What it means to get cited by an AI search engine

Getting cited means an AI assistant chose your brand as a source it trusts enough to name or link inside the answer it writes. When someone asks ChatGPT or Perplexity a question, the model pulls a shortlist of sources, drafts a reply from them, and shows the brands and links it leaned on. Those named sources are the citations, and they are what send you traffic and credit in an AI-first search world.

Picture someone asking an assistant which agency in Jakarta handles AI search. The answer might describe the field, name three firms, and link two of them. The named firms got cited, and every other agency in the market is simply absent from that moment, however well they rank in a normal Google search. That is the new shelf you are competing for.

Being cited is not the same as ranking. Ranking is your position in the old list of blue links, while a citation is being named inside the generated answer itself. A brand can sit on page one and never get cited, and a smaller brand can get cited without being the top organic result. The two overlap less than they used to. Ahrefs found that only 38% of AI Overview citations now come from pages ranking in the traditional top 10, down from 76% a year earlier, across 863,000 search results and four million citation links. A strong ranking still helps, but it stopped being the thing that decides whether an AI names you.

Why your current SEO only takes you halfway

Classic SEO still does real work here. Getting crawled, indexed, and ranking well puts your pages in the pool that AI engines draw from, and a site that search engines cannot read properly will not get cited by the systems built on top of them. None of that goes away. The problem is that it is no longer enough on its own.

That 38% figure is the tell. As AI answers pull more from sources beyond the top 10 and from the many related questions a model asks itself behind the scenes, you can do everything right on your own website and still be missing from the answer because nobody else is talking about you. The audience decides more of this than your homepage does. AI reflects what real people already signal about a brand, so the work shifts from polishing your own pages to becoming the brand your market keeps referring to.

The stakes are climbing fast enough to matter this year. Semrush, looking at billions of visits across more than 50,000 websites, found AI referral traffic grew 66% in 2025, from 462 million to 767 million monthly visits. It projects that AI search visitors could pass traditional search visitors by 2028. The base is still small, but the direction is clear, and the brands that get cited early are the ones building the habit of being named while it is still cheap to do.

Brand mentions now matter more than backlinks

The Ahrefs study of 75,000 brands is the clearest evidence of what actually moves AI visibility. Branded web mentions, meaning the number of times your brand name appears in relevant context across the web, was the top correlate at 0.664. The number of backlinks landed at 0.218, roughly three times weaker, and branded anchors and branded search volume both outranked link count too. The size of the gap between brands is what should get your attention. Brands in the top quarter for web mentions averaged 169 AI Overview mentions, more than ten times the group below them, while the bottom half were close to invisible in AI answers. The brands that already get mentioned keep pulling ahead, while the ones nobody talks about stay out of AI answers almost entirely.

For a business owner, the takeaway is practical. The number of places your brand name shows up in context, in articles, industry roundups, directories, forums, and videos, predicts your AI visibility better than any link-building campaign. This is not about buying links or gaming anchor text. It is earned presence, and you build it by being worth mentioning in the places your customers and your industry already read.

Answer the questions your customers are actually asking

AI answers are built around questions, so the brands that answer questions well become the sources those answers are built from. If your site thoroughly covers the real questions in your category, an AI engine has a reason to reach for you, even when the question is worded a little differently from anything you targeted.

This is the thinking behind our Topic Ownership Strategy at Search Agency. Instead of chasing one keyword at a time, you cover a whole topic your audience cares about, with one strong overview and a cluster of pages around it that answer the narrower questions people ask, what something is, why it matters, how it compares, what it costs, what other buyers say, and how to get it. Together those pages let you own the query space rather than winning a single lucky phrase.

Say you sell payroll software. The topic you want to own is not one keyword but the whole space around running payroll in Indonesia, so you would cover how to calculate BPJS contributions, when THR has to be paid, what the penalties are for late filing, how the main payroll tools compare, and what other finance teams say about them. Answer all of it well and an AI has a confident source to pull from no matter how the payroll question is phrased. The business version is simpler than it sounds. Pick the handful of topics where you know your subject better than your competitors, and answer everything a customer could ask about them better than anyone else does.

Earn mentions on the sources AI already trusts

AI engines lean on a small set of sources far more than most owners realize. Profound, tracking answer engines from August 2024 through October 2025, found Reddit accounted for about 3.11% of all citations, YouTube 2.13%, and Wikipedia 1.35%. Reddit is the single most-cited source inside Perplexity and the second most-cited in both ChatGPT and Google AI Overviews, while Wikipedia ranks first inside ChatGPT. When a model needs to check what people really think about a product or a company, those are the wells it goes back to.

The sources that matter most for you, though, are specific to your industry. A B2B software brand needs a strong presence on G2 and Clutch and mentions in analyst coverage. A hotel needs Tripadvisor reviews and healthy scores on the booking platforms. An ecommerce brand needs ratings on the marketplaces its buyers use. A finance or media brand needs press and named analyst mentions. The pattern holds across all of them, which is that AI trusts you more when the sources your audience already believes are saying good things about you.

Getting there is ordinary relationship and reputation work, not a technical trick. Earn reviews on the platforms that count in your category, get included in the roundups and comparisons your buyers read, answer real questions in the communities where your customers gather, and make sure the story those sources tell about you is current and accurate. You cannot force your way onto Reddit or into a Wikipedia entry, and attempts to plant your brand there tend to get removed or held against you, so the durable route is to be useful and notable enough that other people put you there. Every honest mention in a place AI already trusts is a deposit toward being named in the answer.

Make your brand easy to recognize as one thing

AI systems try to understand your brand as an entity, a single known thing with a name, a description, a location, and a set of offerings. When your business is described one way on your website, another way on LinkedIn, something slightly different on your Google Business Profile, and inconsistently across directories and press, the model ends up with a blurry picture and hesitates to name you with confidence. Consistency is what removes that hesitation. The mismatches are usually small and easy to miss, an old company name in one directory, a different founder listed on an outdated profile, a service you dropped two years ago still advertised somewhere, and each one gives the model a reason to pause before it names you.

The fix is unglamorous but it pays off. Use the same brand name and the same short description everywhere, keep the core facts about your business identical across your major profiles, make sure your Google Business Profile and important directory listings are accurate, and link your official profiles to one another so the connections are easy to follow. Structured data, also called schema markup, is how you state those facts in a format search systems read directly, and Google's own documentation explains how it helps their systems understand what a page and an organization are actually about. We put this into a full brand entity consistency check for clients, because a brand a machine cannot pin down is a brand it will not risk citing.

Be part of the conversation your customers are already in

Reviews, ratings, testimonials, community threads, and creator or video mentions are all signals an AI reads as real people vouching for you. Recency and sentiment carry more weight here than raw volume, so a steady flow of recent, positive mentions says more about you than a wall of reviews from five years ago. Video deserves more attention than most business owners give it, since YouTube is one of the most-cited domains in AI answers, at 2.13% of citations in Profound's data, and it is where a lot of "best" and "how to" discovery happens before anyone opens a search box. A short customer story, a product walkthrough, or an honest explainer on your own channel gives an AI something concrete to point to. Make it easy and worthwhile for happy customers to talk about you, show up in the channels and communities your audience already uses, and keep the presence current rather than treating it as a one-time push.

Total Graph Authority, the model that ties it together

Relevance, trust, and presence are not three separate projects, and the brands that win AI answers are strong on all three at once. This is what we call Total Graph Authority at Search Agency. You become relevant by answering your audience's real questions, trusted by earning validation from the sources they already believe, and present by showing up in the conversations they are already having. AI answers mirror those real audience signals rather than rewarding any trick played on the machine, which is why the approach keeps working as the engines change.

When one of the three is weak, the gap shows up plainly in the AI answer. A brand that comes up for its own name but disappears from category questions usually has thin topic coverage. A well-known site that never appears in comparisons is usually missing third-party validation. A business named under slightly wrong variations has an identity the machine cannot pin down. Each symptom points straight at which pillar to fix first.

A quick way to see where you stand is to check what the major AI assistants currently say about your business and where the picture goes thin. If you would rather have it mapped properly, our AI Search Visibility Audit shows where you are already being cited, where you are missing, and which of the three pillars is holding you back, so you can spend your effort on the part that will actually move your visibility.