The AI Compass framework measures how readable a financial institution is to an AI research-and-comparison agent. We score brands across product surfaces, then split every score into two layers: Technical Readiness and Content Architecture. The result is one number for the executive team and two diagnostic numbers for the teams that can move it.
When a consumer asks Claude, ChatGPT, or Perplexity to find a no-fee 2 percent cashback card, the agent reads the bank's own website, parses the structured data, and decides whether to recommend the product. The brand the agent can read becomes the brand the consumer hears about. The brand the agent cannot read becomes invisible at the moment of consideration. No lead form, no signal, no chance to recover.
Today, agent-driven traffic represents a small share of FI website visitors. Within the next two years, every major analyst forecast pegs it materially higher. The institutions that prepare now own the share. The institutions that wait lose ground.
AI Compass measures readiness against that future. We score what an agent actually sees when it visits the site. The methodology rewards brands that ship clean schema, semantic URLs, server-rendered content, and rich FAQ depth. It penalizes brands that hide rates behind JavaScript, block agents at the firewall, or leave canonical product URLs returning errors.
Every page earns a score of 100, split into two layers. Each layer carries 50 points. The dimensions are scored independently so different teams within an FI can see exactly where the work lives.
Each financial institution is scored across up to seven product surfaces. Surfaces a brand does not offer are excluded from the average rather than penalized.
The front door. Where the agent lands when asked about the brand. Scored on Organization schema, navigation clarity, product taxonomy visibility.
Flagship no-fee cashback card or closest equivalent. Scored on FinancialProduct and CreditCard schema, APR transparency, rewards structure clarity.
Flagship checking or spending account. Scored on BankAccount schema, fee transparency, monthly minimums, eligibility clarity.
Flagship high-yield savings or savings product. Scored on APY visibility in raw HTML, FDIC or NCUA disclosure, tier structure.
Brokerage, robo, or retirement product where offered. Scored on FinancialProduct schema, fee structure, account type clarity.
Mortgage, personal loan, or buy-now-pay-later flagship product. Scored on rate transparency, qualification criteria, loan type clarity.
The customer service surface agents will use most as adoption grows. Scored on FAQPage schema depth, contact path clarity, resolution-flow visibility.
Every surface is scored independently for mobile and desktop, then averaged. The surface score is the sum of Technical Readiness (50) and Content Architecture (50), for a total of 100 per surface.
The brand's overall Agent Readiness Score is the average of the surfaces it offers. A brand that publishes six surfaces is judged on six. A brand that publishes seven is judged on seven. No FI is penalized for not offering a product.
Within each category (Top 10 Banks, Super Regionals, Regional Banks, Credit Unions, Fintechs), the brand with the highest score is the category leader. Across all categories, the brand with the highest score is the overall leader. Per surface, we publish a separate "best in class" ranking.
