I’ve spent a lot of time recently looking into how AI search actually decides what to cite. A few findings stood out and should change how B2B marketing teams spend their time.
AI search engines don’t just copy Google’s rankings
Around two in three citations inside AI search engines come from pages that don’t rank on Google’s first page. That figure comes from Averi’s analysis of 680 million AI citations, published in March 2026.
That matters because many teams still treat AI search as SEO with a new lick of paint. If AI systems simply recycled the top ten blue links, ranking well on Google would be enough, but it isn’t. Large language models (LLMs) reward different signals, and a page can be invisible in classic search yet heavily cited in AI answers, or the other way round.
AI answers are becoming the default search interface
AI Overviews now appear on a large and growing share of Google searches. TechCrunch reported in July 2026 that AI Overviews were showing on 43% of searches, and the trend is upward. For many queries, the AI answer is no longer an experiment sitting above the results. It is the first thing a buyer reads.
The ground underneath keeps moving too. Google completed another spam update in August 2026, and the models behind AI Mode change every few months. Any tactic designed to game a specific model has a short shelf life.
Brand mentions correlate three times more strongly than backlinks
For twenty years, backlinks were the currency of Google rankings. The same 2026 research shows that brand mentions correlate roughly three times more strongly with AI visibility than link profiles do.
In plain terms, being talked about by credible third parties across the web, consistently and in context, is doing the heavy lifting. A trade publication describing what your company does, an analyst naming you in a category, a customer discussing you in a community thread: these are the signals AI systems use to decide who is credible enough to mention.
This is the part that resonated most with me given my communications background. Two decades of SEO playbooks, and the defining signal for AI visibility looks a lot like PR instinct: earn coverage, say the same clear thing everywhere, and let credible people talk about you.
What B2B marketing teams should do about it
The practical shift is from counting links to tracking mentions and making sure they say the right thing.
- Track unlinked brand mentions. Know where your company is named, in what context and alongside which competitors, whether or not there’s a link.
- Make your story consistent. AI systems build answers from what is corroborated across sources. If your website, press coverage and analyst descriptions describe you differently, the AI answer will be vague or wrong.
- Invest in earned media again. Credible third-party coverage is one of the strongest inputs to AI visibility, and it compounds over time.
- Stop chasing model-specific hacks. The models change too often for tricks to last. Authority does, which is why the attribution gap matters less than being the source.
If you manage organic search or content strategy, it’s worth asking a simple question this quarter: are you tracking brand mentions yet, or still anchored to link profiles? And if your team publishes on LinkedIn, it’s worth knowing how AI uses LinkedIn content.
SOURCES
Questions
Do backlinks still matter for AI search?
Backlinks still matter for traditional Google rankings, but 2026 research shows brand mentions correlate around three times more strongly with AI visibility than link profiles. Links alone are no longer the main signal for being cited in AI answers.
What is an unlinked brand mention?
An unlinked brand mention is any place your company is named online without a hyperlink back to your site, such as a news article, an analyst note, a forum thread or a review. AI systems read these mentions as evidence of who is credible in a category.
Is optimising for AI search the same as SEO?
No. Around two in three AI citations come from pages that don't rank on Google's first page, so AI systems are using different signals from classic search rankings. Good SEO helps, but it doesn't guarantee AI visibility.
Nick Clark is the founder of Datum Point, a B2B consultancy for AI visibility and presence. He has spent 25 years in B2B, technology and corporate comms and marketing, including as CEO of Nelson Bostock Group and Publitek. First published on LinkedIn on 28 August 2026. Follow Nick on LinkedIn.