In early 2023, while most of the tech industry was still experimenting with ChatGPT, a fintech company was solving a problem that would define the future of digital discovery. The challenge was one traditional SEO couldn’t address: how to make financial data appear consistently and reliably in AI-generated answers across multiple platforms.
The framework that emerged (structured data pipelines optimized for algorithmic citation rather than search rankings) would become one of the earliest implementations of what the industry now calls “Search Everywhere Optimization.” Working within the compliance constraints of a regulated brokerage, the team led by Dinesh Modi, Director at fintech platform Public, built a system that ensured market data appeared reliably in responses from modern AI search platforms like ChatGPT, Perplexity, Gemini, Grok, and Claude.
Today, as click-through rates plummet and zero-click answers dominate, this early work offers a blueprint for how organizations must adapt to an AI-first discovery landscape.
The Zero-Click Challenge
By mid-2023, the challenge for fintechs was becoming clear. Users were asking AI chatbots questions like “What’s Apple’s P/E ratio?” or “Should I buy Tesla stock?” and getting answers that either cited outdated data, incorrect figures, or didn’t mention Public at all, despite the platform maintaining comprehensive, real-time financial information.
“Traditional SEO assumes users will click through to your site,” Dinesh explains. “But when ChatGPT gives someone a complete answer inline, that click never happens. The question became: how do you optimize for visibility when there’s no webpage to visit?”
At the time, no fintech firm had a systematic approach to this problem. While concepts like Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO) were being discussed academically, no one had adapted these theories to regulated financial services where accuracy isn’t just preferred, it’s mandatory under SEC guidelines.
Between May 2023 and January 2024, Modi developed and tested a solution.
The Three-Part Framework
His approach rested on three foundational elements:
1. Structured data pipelines. Rather than just publishing numbers on web pages, the team created systems that transformed financial metrics (market cap, P/E ratios, earnings forecasts, options chains, pre-market and after-hours quotes) into formats AI models could reliably extract and cite. Using schema markup, JSON-LD, and semantic ontologies, the data became machine-readable in a way that allowed algorithms to understand context. For example: this is a P/E ratio, calculated on this date, using this methodology.
2. Semantic content architecture. Content templates were designed specifically for AI consumption, with passage-level answer formats, contextual FAQ structures, and clear semantic markup aligned with how large language models parse information. “AI models don’t read like humans,” Modi notes. “They look for patterns, clear definitions, and unambiguous statements of fact. We restructured content to provide those elements.”
3. Cross-platform testing. Perhaps most crucially, systematic testing protocols were established across ChatGPT, Google’s Bard (now Gemini), Perplexity, and Anthropic’s Claude. “I would run the same financial queries across all these platforms, document which ones cited Public, analyze why others didn’t, and refine our structured data accordingly,” Modi says. “It was an iterative feedback loop: test, analyze, adjust, retest.”
By late 2023, the results were measurable. Public’s financial data began appearing as a cited source in AI-generated responses with increasing frequency. The company’s financial information for various investable asset classes was now discoverable not just through Google searches, but through the AI search interfaces where millions of investors and traders were increasingly turning for answers.
The First-Mover Implementation
While academics discussed “generative engine optimization” in papers, the team at Public was implementing it in production at a regulated brokerage where a single data error could hamper user trust and lead to compliance issues. And while others optimized for individual platforms, they built a unified framework that worked across AI search platforms like ChatGPT, Perplexity, Gemini, Grok, and Claude simultaneously.
According to SEO leaders at fintech platforms CommSec (Australia) and Dhan (India) who were grappling with similar discovery challenges, Public was notably early in addressing this challenge systematically. While competitors were still optimizing for Google rankings, Modi and his team were already testing how ChatGPT parsed financial data and what structural changes would make that data citeable across multiple AI systems.
The approach differed fundamentally from traditional SEO’s focus on keywords and backlinks. Instead of optimizing for human-readable page titles, the focus shifted to machine-parseable data structures that AI models could confidently reference. The unit of optimization moved from the webpage to the entity and its relationships.
The Numbers Prove the Shift
This early recognition of the paradigm shift is now being validated by industry-wide data showing the collapse of traditional search metrics.
When Google’s AI Overviews appear, click-through rates for the top-ranking page drop by 34.5% compared to similar queries without an overview, according to Ahrefs research. GrowthSRC Media found CTRs for Google’s top result fell from 28% to 19% (a 32% decline) after AI Overviews expanded.
A Bain & Company report estimates that roughly 80% of consumers now rely on zero-click answers in at least 40% of their searches, reducing organic web traffic by 15 to 25%. When AI summaries appear, Pew Research found that users click on traditional results only approximately 8% of the time, compared to 15% when no summary is shown.
“Users click on traditional search results only around 8% of the time when an AI summary appears,” Pew Research’s behavioral data confirms. “This surge in AI Overviews may be impacting clickthrough rates for organic listings,” Search Engine Journal’s Editor wrote in a recent analysis.
In this new paradigm, having a highly-ranked webpage matters less than being cited within the AI-generated answer itself. This shift became apparent to early implementers in 2023. By 2025, it had become an industry-wide concern.
From Early Implementation to Industry Standard
What was built early at Public is now being discussed across the digital marketing industry as an essential evolution of SEO. Terms like “Search Everywhere Optimization,” “Generative Engine Optimization,” and “AI Search Optimization” are entering the mainstream lexicon, describing approaches that were being implemented in production systems while others were still debating theory.
“Optimize the format, metadata, and semantics so AI sees your content as a source, not noise,” advises Lexi Mills, CEO at Shift6, a perspective that resonates with the work done by Modi during the same period.
Industry observers now recognize that competitive advantage in discovery has shifted. Organizations whose data architecture allows them to be cited by AI systems are gaining visibility, while those relying solely on traditional SEO tactics are losing ground. The framework developed at Public in mid-2023 anticipated this shift before it became conventional wisdom.
Beyond Financial Services
The implications extend well beyond fintech. Healthcare providers citing medical information in AI responses face similar accuracy requirements. News organizations watching traffic evaporate to AI summaries need comparable frameworks. Even companies building internal AI assistants face identical discoverability challenges.
The structured data approach developed by Dinesh Modi for financial compliance has become a reference point for organizations where accuracy isn’t optional. The core principles (structure over style, explicit context for AI systems, systematic cross-platform testing, compliance-first optimization) are now emerging as industry best practices across sectors facing similar challenges.
The fundamental insight remains: in an AI-first world, the unit of optimization is no longer the webpage. It’s the structured knowledge entity and its relationships.
Looking Forward
As of October 2025, the search landscape anticipated in 2023 has fully materialized. Google’s AI Overviews are ubiquitous. ChatGPT handles billions of queries monthly. Perplexity, Gemini, and Claude are mainstream tools. Voice assistants are more capable. And traditional website traffic continues its structural decline.
Organizations that ignored this shift are now scrambling to understand why their carefully optimized websites are invisible in AI responses. Meanwhile, those who adopted approaches like “Search Everywhere Optimization” (focusing on structured data, semantic clarity, and cross-platform testing) have maintained or even grown their digital relevance.
What seemed experimental in 2023 is now standard practice. The companies investing in structured data architecture, semantic markup, and cross-platform testing are the ones maintaining visibility as the discovery landscape continues to evolve. The shift from optimizing for search engines to optimizing for AI systems represents more than a technical adjustment – it’s a fundamental rethinking of how information becomes discoverable.
As the lines between search engines, AI assistants, and knowledge platforms continue to blur, the principles established by early implementers are becoming foundational. The future of discoverability won’t be about gaming algorithms. It will be about structuring knowledge so machines can understand and cite it accurately.
About the Research: This article draws on studies from Ahrefs, Pew Research Center, Bain & Company, Search Engine Journal, and GrowthSRC Media, as well as interviews with practitioners in the search optimization and technology sector.
