Why optimisation alone no longer works
For much of the last two decades, digital marketing has been built around tactics.
If something stopped working, you adjusted it. You changed keywords, rewrote headlines, refreshed metadata, tweaked layouts, or experimented with formats. Progress was made incrementally, and success came from staying one step ahead of the system.
That approach worked because the system was relatively simple. Search engines and social platforms rewarded specific actions, and those actions could be optimised in isolation.
That world has changed.
Today, visibility depends less on what you do in any single place and more on the signals you emit across all of them. Trust is no longer inferred from isolated optimisations, but from patterns that hold together over time.
When ranking stops being a proxy for relevance
Google rankings were always a proxy. They stood in for relevance, trust, and usefulness — imperfectly, but at scale.
AI systems don’t need that proxy.
They don’t rely on page order. They don’t privilege freshness in the same way. They don’t reward optimisation patterns just because they worked historically.
Instead, they decide which sources feel safe to reference when generating an answer.
That distinction matters.
Because a source can rank well and still be ignored by AI. And a source can be cited repeatedly without ever receiving meaningful search traffic.
Visibility is no longer about where you appear. It’s about whether you’re selected at all.
Why AI citation behaviour looks “wrong” through an SEO lens
From a traditional SEO perspective, that 12% figure feels counterintuitive.
If Google’s top results represent the “best” answers, why wouldn’t AI systems rely on them?
The answer is simple: AI systems are solving a different problem.
Search engines rank pages.
AI systems assemble explanations.
To do that, they look for:
- Consistent language across sources
- Stable definitions
- Clear ownership of a topic
- Explanations that hold up when paraphrased
- Signals of credibility beyond a single page
These aren’t ranking factors. They’re trust signals.
Output is abundant. Confidence is scarce.
The web is saturated with content designed to rank.
AI systems don’t need more of it.
What they need is confidence — confidence that a source can be reused without introducing risk. That confidence is inferred from patterns, not positions.
This is why many brands that invested heavily in SEO now find themselves “present but absent”: visible in rankings, invisible in AI answers.
If your content only performs inside Google’s ecosystem, you’re competing in a shrinking arena.
The same pattern shows up in brand consistency research
This isn’t unique to search.
Across branding and customer experience research, the same principle appears repeatedly: consistency predicts trust more reliably than isolated quality.
- Exploding Topics reports that companies with consistent branding see 10–20% higher revenue growth attributable to brand marketing, while fragmentation leaves that upside unrealised.
Source: Exploding Topics - In omnichannel contexts, Isitatech found that consistency of experience is 30% more predictive of customer satisfaction than the quality of individual interactions, with strong performers achieving around 89% retention versus 33% for weak ones.
Source: Isitatech
AI systems behave in a similar way. They privilege sources that feel coherent across contexts, not ones that simply perform well in isolation.
AI doesn’t reward optimisation – it rewards recognition
Optimisation still helps machines understand content.
But recognition is what makes them reuse it.
AI models learn which sources “sound right” for a topic by encountering them repeatedly, in similar forms, across different surfaces. Over time, certain names, phrases, and explanations become familiar.
That familiarity is what drives citation.
This is why being referenced in AI answers is closer to brand recognition than ranking. It’s not a single win. It’s an accumulation effect.
Why ranking-first strategies quietly fail in AI search
If your entire visibility strategy is built around ranking on Google, you’re optimising for one system’s incentives.
AI systems sit outside that logic.
They don’t care how much effort went into optimisation. They care whether your content reduces uncertainty.
A perfectly optimised page that contradicts your other materials, uses inconsistent terminology, or reframes the problem differently from one article to the next becomes harder to trust.
From an AI perspective, that’s risk.
A more honest question to ask
The old question was binary:
Do we rank?
The more useful question now is:
If an AI had to explain this topic clearly, would our perspective feel familiar enough to draw from?
Not quoted verbatim, not necessarily attributed but recognisable.
If the answer is no, ranking won’t save you.
What the 12% stat really tells us
That 12% overlap isn’t a technical quirk.
It’s a signal.
It tells us that AI discovery is operating on a different axis of trust than traditional search. One that prioritises coherence, consistency, and authority over tactical performance.
Google rankings still matter. But they’re no longer the ceiling — or even the centre — of visibility.
They’re just one signal among many.
Visibility now belongs to those who are chosen
AI search doesn’t surface the “best-ranked” result.
It surfaces the least risky explanation.
That’s why authority beats output.
Why signals beat tactics.
Why being recognised matters more than being present.
If your content only ranks on Google, you’re already invisible where visibility is heading.
