Eamonn Conway, Managing Director at Fiducial Communications, provides some guidelines for ranking on LLMs.
Financial marketers have been here before. When websites arrived, then mobile, then voice, our playbooks had to change. The rise of large language models (LLMs) like ChatGPT is the next step-change. Already, firms are seeing their web traffic by 25% as AI chatbots and virtual agents intercept queries that once went to Google. And already, some financial firms are noticing that over 50% of their referrals are coming from LLMs as consumers embrace generative AI search.
As AI search rises, financial brands are losing visibility
Here’s the paradox: financial institutions have some of the most authoritative and trusted content anywhere online: market outlooks, white papers, product explainers, regulatory insights, yet much of it is invisible in AI-driven discovery.
Why? Because large language models can’t easily access or interpret it.
Many financial websites are still built on frameworks that hide their best material behind JavaScript-heavy interfaces or PDFs buried and well hidden. Others block crawlers for compliance reasons, meaning AI systems can’t “read” or learn from their pages. And too often, content that could demonstrate real expertise – and target audience consideration – such as investment insights or insurance comparisons is buried in long, unstructured paragraphs that LLMs can’t parse into usable snippets.
As a result, AI systems trained on open, structured data often turn to nimbler challengers, smaller firms or fintech disruptors whose content is lighter, cleaner, and easier to process. These challengers end up being the firms that AI recommends first when users ask things like:
“What’s the best small business insurance provider in the UK?”
“Which investment platforms are FCA-authorised and have low fees?”
“How do I choose between a SIPP and an ISA?”
Authority still wins, if it’s discoverable
The good news for financial marketers is that authority still matters.
LLMs are consensus-driven. They favour trustworthy, compliant, and credible sources to avoid legal or factual risks.
In practice, being an established or well-known firm gives you a natural head start. Your content already meets many of the right criteria: high editorial standards, regulatory accuracy, and a long history of credible publishing. Over time, this builds domain reputation, which AI systems recognise as a key signal of trustworthiness.
However, your competitors have the same advantage. Most major players in the industry also produce high-quality, compliant content that AI considers reliable.
Where few are investing enough attention is in accessibility and structure. Making sure your content is readable by AI systems and supported by proper schema markup can make all the difference. Schema markup is a bit like postal bar codes that make sorting easier and this is where early movers can gain a genuine competitive edge. By making your trusted content easier for AI to interpret, your firm can stand out even in a market full of credible voices. Over time, this competitive edge will compound further as AI identifies your content as meeting the key criteria.
A new kind of SEO: Making content AI-readable
Optimising for large language models requires a different mindset. One that focuses on making your trusted content viewable, readable, retrievable, and recommendable by AI systems.
That starts with structure.
A call to re-structure content, not replace it
LLMs work best when they can extract meaning from clear patterns, which are headings, sub-headings, FAQs, and logical formatting. Instead of relying on long PDFs or static brochures, financial marketers should create modular web content that directly answers conversational questions.
Each topic can become a structured section with schema markup, meta descriptions, and concise summaries that AI systems can easily interpret and recommend. Collaborate with your web developer to ensure your content is properly labeled with schema markup for optimal AI visibility.
This isn’t about tearing down everything you’ve built, but it’s about re-structuring how it’s presented.
Start by reviewing your content. Then ask:
• Is it published in a format AI can read (HTML, structured data, open access)?
• Does it contain natural-language questions and clear, factual answers?
• Is your website allowing AI crawlers to access these pages?
• Are your compliance and marketing teams aligned on what can be indexed safely?
These simple steps don’t require radical AI tools, just smarter content design and cross-department collaboration.
Optimise your video content for AI search
Many financial firms use explainer and educational videos as part of their marketing and distribute them across their websites. This presents an opportunity to make those existing assets readable by large language models (LLMs), but it requires an understanding of how AI actually processes video content.
LLMs don’t watch videos like humans do; they analyse the data attached to them. That includes titles, descriptions, captions, transcripts, and structured metadata such as video length and content type.
By optimising these elements and embedding videos with clear schema markup (VideoObject), marketers can help AI systems interpret the content of the video, who it’s for, and why it’s relevant.
Done well, this approach gives your content two layers of visibility: engagement from human audiences and discoverability within AI-driven search environments.
Looking to the future
This shift won’t happen overnight, but the window for early movers is now. LLMs learn from their own outputs, creating a “rich get richer” dynamic where frequently cited sources receive more citations in future queries.
“Just block AI” is not a viable strategy. Financial services need to be having the LLM conversations. Giving little to no context causes the LLMs to “hallucinate”. In the absence of relevant information to correctly answer questions, they return ‘best guess’ or potentially ‘made up’ content. ,
The longer financial services firms keep their heads in the sand, the greater the risk that more nimble competitors will claim the new territory. Just as the brands that built websites first became market leaders in the 2000s, those who optimise for AI search today will shape tomorrow’s discovery landscape.
