AI Assistants Are Becoming the New Front Door for Banking Product Discovery

The Shift From Search Links to AI-Generated Answers

A fundamental change is underway in how consumers discover and evaluate banking products. According to new research from Bain & Company analyzing consumer behavior in Australia, AI assistants are increasingly becoming the primary entry point for customers seeking banking products—bypassing traditional channels like bank websites, apps, branches, and call centers entirely.

The research, conducted by Profound for Bain & Company using opt-in consumer panels, reveals that customers are forming preferences and creating shortlists before or without ever visiting a bank’s website. This represents a significant departure from the two-decade model where banks competed for visibility through branches, brokers, marketing scale, and online search rankings.

During the discovery stage, AI is now embedded in search engines like Google’s Gemini and operates through standalone assistants such as ChatGPT and Claude. These AI assistants provide synthesized responses to customer questions, which proves especially valuable when queries require explanation, guidance, or analysis of trade-offs between different banking products.

The research found that roughly one-third of banking-related AI queries involved home loans, with customers prompting assistants to find banks matching specific selection criteria such as fast preapproval. Credit card customers frequently pasted pages of fees into AI assistants, asking them to convert fine print into understandable text. Customers are also using AI assistants to interpret terms, receive advice, and even draft communications on their behalf.

How AI Assistants Source Their Recommendations

The Bain research provides insight into where AI assistants pull their information when responding to banking queries. Across all types of queries, social media and content sites dominated as sources. For banking-specific queries, AI assistants relied not only on banks’ own websites but also on other trusted sources, including comparison websites.

This sourcing pattern has significant implications. Recommendations are influenced by customer feedback found in reviews and forums in ways that traditional search engines do not capture. Banks can no longer gain awareness and relevance simply by driving customer traffic to their sites; they must position themselves for inclusion in AI-generated shortlists.

The research also revealed some unexpected patterns in share of voice. In transaction banking, major banks led in share of voice, but this metric did not reflect overall market share—highlighting an opportunity for smaller banks in AI brand discovery. For home loans, while major banks led, digital-first banks such as Unloan (owned by Commonwealth Bank) and Loans.com.au (owned by Firstmac) also ranked among the top-cited brands, demonstrating how challenger brands are shaping customer consideration.

Perhaps most notably, global banks with limited Australian presence ranked high in share of voice in large language model outputs. This was most pronounced in credit cards, where US competitors including Chase and Capital One—which do not offer consumer-focused cards in Australia—had a high share of mentions. Their visibility derived from the large amount of globally relevant content available through trusted websites that LLMs scrape, suggesting that geographic market presence matters less than content volume and authority in AI-driven discovery.

Brand Framing and Sentiment in AI Outputs

Being included in AI-generated recommendations is necessary but not sufficient. How the LLM behind an AI assistant frames a bank’s brand influences customer consideration. The Bain sentiment analysis found that positive framing of a brand correlated with repeated signals: clear anchor facts, structured and comparable product language, and credible third-party validation.

Negative framing tended to stem from regulatory scrutiny, articles and commentary about scams, fraud, and other customer harms, and persistent clusters of complaints about fees, disputes, or service delays. This dynamic elevates the importance of nurturing customer reviews from promoters while minimizing detractor feedback.

Some banks have already begun optimizing their brands for AI discoverability. Wells Fargo, for example, partnered with Schema App to layer structured data across its website to reduce AI hallucinations and improve the reliability of answers. This represents an early example of what may become standard practice as AI-driven discovery matures.

Trust remains a constraint in this new landscape. According to the research, customers are more likely to make a purchase with AI assistance if they feel a brand and transaction process are secure and credible. This suggests that while AI assistants can drive discovery and consideration, the final conversion still depends on traditional trust signals.

What This Means for SaaS Teams

While this research focuses on banking, the implications extend to any SaaS company competing for visibility in AI-generated recommendations. Several patterns are worth noting:

First, content strategy must evolve beyond SEO optimization for traditional search. Structured data, clear product language, and comparable feature descriptions appear to influence how LLMs represent brands. SaaS teams should audit how their products appear in AI assistant responses and identify gaps in how their value propositions are being communicated.

Second, third-party validation matters more in AI-driven discovery. Reviews, comparison site mentions, and credible external sources influence AI recommendations in ways that owned content alone cannot replicate. This elevates the importance of review management, analyst relations, and presence on comparison platforms.

Third, global content can create unexpected competitive dynamics. The finding that US credit card companies appeared prominently in Australian AI responses—despite not operating there—suggests that content volume and authority can override geographic relevance. SaaS companies with strong content footprints may find themselves appearing in markets they don’t actively serve, while regional players may struggle for visibility against global competitors.

It remains uncertain how quickly these patterns will evolve as AI assistants improve their ability to filter for geographic and contextual relevance. The research covers Australia specifically, and while Bain notes similar consumer willingness to rely on AI assistants in other countries, the specific dynamics may vary by market.

For SaaS operators, the strategic imperative is clear: brands must now shape how AI understands and recommends their products, not just how customers find them through traditional channels. The port of entry is shifting, and visibility strategies must shift with it.