The Bank That Doesn't Exist in AI Search. And the One That Does.

Australia's financial services sector is a competitive advertising environment. The big banks spend hundreds of millions annually on brand, performance, and digital, and they have the Google rankings to show for it. Strong domain authority, years of SEO investment, high organic visibility across every major product category. By traditional search metrics, they are doing everything right.

So when we ran Share of Model's Search Visibility module across CommBank and Westpac, the traditional search results looked roughly as expected. Both institutions appeared. Both ranked. Although there was a clear gap between them.

On generative engines ChatGPT, Gemini, Perplexity the same two brands told a more urgent story. Westpac has a generative engine visibility score of 59 out of 100. CommBank scores 4. Both are household names. Both have invested heavily in digital. The difference has nothing to do with brand awareness or marketing budget, and everything to do with the content signals that AI models draw on to form their answers, signals that most finance brands aren't yet measuring, let alone acting on.

The metric: brand mention presence

Share of Model tracks two distinct signals. Domain presence: whether a bank's website is cited as a source, and brand mention presence: whether the brand name appears in answer at all, regardless of a link.

In a zero-click world, brand mention presence is where the real competition is happening. The consumer hears a name, forms an association, builds a consideration set, before they've clicked anything.

Across Westpac's tracked categories, brand mention rates on ChatGPT range from 59% to 93%. On Gemini, 73% to 89%. Westpac is being named in the vast majority of relevant AI conversations about Australian financial products without needing to be clicked to. That's a form of brand reach that doesn't show up in any traditional analytics dashboard.

For CommBank, brand mention presence across ten tracked categories rounds to zero on almost every platform. This isn't a reflection of CommBank's brand strength or product quality, it's a measurement gap. Without visibility into how AI models are forming their answers, it's impossible to know the gap exists, let alone close it.

Third parties are filling the space

When a bank isn't present in an AI answer, something else takes its place. Across the queries run in CommBank's product territory, the most-cited external source is a government financial guidance website, outranking CommBank's own domain in visibility score. Comparison platforms sit just behind it, serving product rankings without any of the brand's own framing.

For Westpac, Finder and Canstar appear as the second and third most-cited sources. These platforms co-author the AI narrative around Westpac's products - manageable when your own site leads strongly, but a risk if that lead narrows.

The Sources and Links feature reveals not just whether your content is being cited, but who else has a say in how AI understands your brand. For most financial services marketers, that picture is a surprise.

Topic gaps are not evenly distributed

What the Thematics and Topics view consistently reveals - across financial services, is that AI visibility doesn't mirror a brand's overall digital footprint. A bank can be well-represented in AI answers for home loans and transaction accounts while being virtually absent for insurance or superannuation queries. The categories where consumers most actively seek guidance are often the ones with the biggest gaps.

This matters because those gaps shift over time and vary by platform. A brand appearing strongly on Google AI Overviews for a given topic may tell a completely different story on ChatGPT or Perplexity. Tracking visibility at the category level, across every platform simultaneously, is what turns a broad awareness problem into a specific, addressable content brief - and what shows you where a competitor is gaining ground before it shows up anywhere else.

What the AI is actually trying to answer

The Query Fan Out view reveals the sub-queries AI models generate internally before forming an answer. When someone asks "what's the best savings account in Australia," the model isn't pulling from one source - it's running pricing breakdowns, persona-fit evaluations, feature comparisons, and how-it-works queries simultaneously.

For CommBank, content appears in buyer guide sub-queries just 13% of the time. These are the query types that determine whether a brand ends up in the final recommendation. Fan-out analysis turns that into a direct content brief: the formats AI draws on most - comparison content, persona-specific guidance, plain-language explainers - are exactly what financial services brands have historically underproduced.

The window is still open

For any financial institution still measuring success primarily through traditional search rankings, the opportunity is clear. Understand where your search visibility gaps are, by platform, by topic, by query type, by source - and you can close them. Share of Model's Search Visibility module makes that gap visible, and turns it into an actionable brief.