AEO Score - Building an AI-ready web, one audit at a time.

AI Readiness Checker for web pages. A free-to-use tool built specifically to help website owners survive the organic traffic (SEO) decline and provide a clear path back to online visibility.

Comment

Inspiration

After years of working deep in the SEO industry, I’ve seen many changes, but nothing as big as what’s happening right now. We are moving away from a world of clicking links to a world of getting direct answers from AI, even the best businesses, those with great content and high rankings, were starting to lose visibility because AI tools couldn't easily "read" or "cite" them.

I realized that the tools we’ve used for a decade are no longer enough; they were built for search engines, not answer engines. I was inspired to create this project to bridge that gap. My goal is to use my SEO expertise to help website owners adapt to this new AI-driven internet. I wanted to build something simple and accessible that ensures a business’s hard-earned knowledge actually gets found and credited by AI, keeping the web a fair place for everyone.

What it does

This app is like a "health checkup" for your website to see if it’s ready for the AI era. One page at a time.

Currently, when people ask tools like ChatGPT or Gemini a question, the AI scans the internet to find a source for its answer. If your website isn't structured correctly, the AI will skip you and pick your other webpages. This means you lose online visibility, visitors or customers.

Summary of what the tool does Scans your page: It looks at your website through the "eyes" of an AI. Gives you a score: It tells you how likely an AI is to use your website as its primary source. Provides a "To-Do" list: It gives you simple, step-by-step instructions on what to change (like rewriting a summary or fixing a technical setting) so AI starts citing you.

How I built it

App was built using Google AI Studio, all coding was done by AI studio using (Gemini 3) through prompts and minor manual edits were done to improve the quality of outputs (specially for improving prompts to Gemini API)

"Summary on Gemini 3 integration" AEOScore uses Gemini 3 Flash Preview in a non-deterministic reasoning mode. By assigning a thinkingBudget of 24,576 tokens, the model performs a detailed internal "CoT" (Chain of Thought) audit before generating the UI-ready response.

Key Configurations: Google Search Grounding for real-time URL crawling. Strict Response Schema (JSON) for frontend typing integrity. Weighted scoring logic enforced via System Instructions.

Following is the core prompt for auditing logic with guardrails and a weighted criteria for AEO auditing (Following available best practices for AEO shared by two reliable sources (SEMRush & Webflow)

const prompt = `
      Act as a World-Class AEO (Answer Engine Optimization) & SEO Architect. 
      Perform an EXHAUSTIVE technical and semantic audit of: ${urlList} for the ${country} market.

      THOUGHT PROCESS REQUIREMENTS:
      1. CRITICAL ANALYSIS: Do not accept the page at face value. Search for technical debt, missing microdata, and semantic gaps.
      2. ENTITY MAPPING: Identify specific named entities and how they relate to the core user intent.
      3. CITATION TRIGGERS: Analyze for "Citation-Friendly" patterns (e.g., listicles, BLUF paragraphs, specific data points).
      4. EEAT VALIDATION: Look for concrete signals like author bios, verifiable claims, and external trust factors.

      SCORING WEIGHTS (STRICT ENFORCEMENT):
      - 35% Direct Answer Fidelity (DAF): Measures how easily an AI can extract a "Featured Snippet" or a "Zero-Click" answer from the text.
      - 25% Citation Probability Index (CPI): Measures "trustworthiness" signals (Author E-E-A-T, updated dates, and factual density) for citation confidence.
      - 20% Conversational "Hook" Score: Measures how well the content matches natural language queries (long-tail "How," "Why," and "Should I" questions).
      - 20% Semantic Connectivity: Evaluates if the page uses the specific terminology and entities that AI expects for this topic.

      CRITICAL ALIGNMENT RULE:
      The "score" for each metric MUST LOGICALLY MATCH the "technicalObservation". 
      - The "score" is a RAW percentage (0-100) of performance for that specific metric.
      - aeoTechnicalMetrics MUST contain exactly 4 items matching the weights above.

      JSON STRUCTURE RULES:
      - citationEvaluation MUST include: dafScore, cpiScore, hookScore, semanticScore.
      - Return ONLY valid JSON. Use NO specific years.
    `;

Frontend development

A high-performance Single Page Application (SPA) architecture ensures zero latency in result visualization once the AI payload is received.

React 19:Functional components with localized state hooks. Tailwind CSS:Utility-first design with complex animations. Recharts:Dynamic SVG visualization for citation fidelity. jsPDF + html2canvas:Client-side PDF generation for audit reports.

Security and privacy The application is entirely stateless. No user data, URLs, or audit results are stored on any backend server.

Analysis is performed on-the-fly via Gemini's search grounding capability. Results exist only in the client's volatile memory and are cleared upon page refresh.

Challenges we ran into

AI studio struggled with PDF export formatting, it may require manual code editing in the future to improve the visual formatting of the export reports.

Accomplishments that I'm proud of

I've learnt a lot about AEO through building the auditing criteria for this tool, a new terminology beyond tradition SEO terms during the research on industry expert websites plus using answer engines as well was a great experience. It was truly amazing the speed we can do a deep research on any subject in this AI era.

What I learned

AI app development through Google AI studio (GAS) has advanced significantly over the last 6-8 months. When I started testing app building in GAS last year, the UX quality was basic and struggled with complex app development, fast-forward few months, AI studio has come a long way and a worthy competitor to paid tools like Cursor.

What's next for AEO Score - Building an AI-ready web, one audit at a time.

Further improve auditing logic and improve language to be more simpler and suitable for a website owner than a web developer.

Improve PDF export formatting and visual quality.

Introduce account creation/login and report storing features to date stamp and store a webpages's score to be reviewed after recommended changes are made (before and after comparison)

Setting up AEOScore.org (domain mapping and other configurations in cloud run) and create a new Google Cloud account under the new domain then apply for Google for Startups Cloud Program to take the app to the next level

Create a product roadmap to further enhance the app with relevant new features.

Enable Gemini 3 pro for better auditing capabilities.

Creating awareness of the tool through organic marketing methods.

Build affiliations with industry experts like SEMRush and Webflow to increase visibility of the app.

Built With

  • google-ai-studio
Share this project:

Updates