The Context
By mid-2024, the SEO landscape for financial brands in India had shifted fundamentally. Google's AI Overviews had begun appearing for 30–40% of financial queries. ChatGPT, Gemini, and Perplexity were being used by millions of Indians for financial guidance — asking questions like "which bank gives the lowest personal loan interest rate in India?" and "what is the minimum CIBIL score for a credit card?"
These AI-powered answers were pulling from web sources — and those sources were the ones that ranked in featured snippets, had strong FAQ schema, and were structured to answer questions directly. BankBazaar had the best financial comparison data in India, but its content structure was designed for human readers browsing comparison tables, not for AI models extracting direct answers.
Understanding AEO vs. Traditional SEO
I spent six weeks studying how ChatGPT, Gemini, and Google AI Overviews select sources to cite. The pattern that emerged was clear: AI engines prefer content that is: (1) structured as direct Q&A rather than narrative prose, (2) factually specific with numbers and dates, (3) attributed to credible institutional sources, and (4) wrapped in appropriate structured data like FAQPage and SpeakableSpecification schema.
Traditional SEO optimises for ranking signals — backlinks, authority, keyword density. AEO optimises for extractability: can an AI model extract a specific, accurate answer from your content without reading the entire page?
The AEO Transformation
- Audited 400+ financial product pages and identified the 8–12 most common questions users and AI search engines were asking about each product
- Rewrote FAQ sections from generic to hyper-specific: "What is a personal loan interest rate?" → "What is the current SBI personal loan interest rate for salaried employees with a CIBIL score above 750?"
- Deployed FAQPage JSON-LD schema on all 400+ pages with answer length optimised for featured snippet (40–60 words)
- Added SpeakableSpecification markup on key definitive answers
- Built a monitoring system using n8n + Claude AI that tracks 200+ financial queries across ChatGPT, Gemini, and Perplexity weekly — logging which sources are cited
- For queries where BankBazaar is not cited, analysed the cited sources and reverse-engineered their content structure
- Implemented "source-worthy" content elements: direct quotes from RBI/SEBI publications, explicit data attribution, and "As of [Month Year]" freshness signals
- Created dedicated "Answer-First" landing pages for 50 high-value financial questions — pages where the direct answer appears in the first paragraph, before any comparison table
- Used GSC data to identify 200+ queries where BankBazaar ranked positions 2–5 but was not capturing the featured snippet
- Restructured page content to match the snippet format Google was already selecting from competitors: definition → key criteria → example → source
- Added "quick answer" boxes at the top of informational pages, formatted as a single 40-60 word paragraph with the key answer
Results
The featured snippet CTR improvement was the most immediately measurable result. Pages that captured snippets saw a 3× increase in click-through rate versus the same queries where BankBazaar ranked in position 1 without a snippet. For high-volume financial queries, this difference is significant — featured snippets command up to 8% CTR compared to 3% for a standard #1 result.
The AI citation monitoring system has become one of the most valuable SEO intelligence tools I've built. Knowing which of BankBazaar's pages are regularly cited by ChatGPT for financial queries — and which are being overlooked in favour of competitors — provides an ongoing roadmap for content improvement that is entirely invisible to traditional rank-tracking tools.
Key Takeaways
AEO requires a fundamental rethinking of how we structure financial content. The question is no longer "does this page rank well?" but "would an AI model choose to quote this page to answer a specific financial question?" Those are very different standards, and optimising for one doesn't automatically optimise for the other.
The most powerful insight from this project is that specificity is the currency of AI citation. Generic financial content that says "personal loan rates vary by bank" will never be cited by an AI. Specific content that says "as of June 2026, SBI's personal loan rate starts at 11.15% p.a. for salaried employees with a CIBIL score of 750+ (source: SBI official website)" has a far higher probability of being cited — because it gives the AI exactly what it needs to answer a user's specific question.