Seo Optimizer

SEO Optimizer

Overview

Optimizes digital visibility through entity-based SEO, structured data, and content architecture. Focuses on AI Overviews optimization, E-E-A-T authority signals, and semantic topic clustering rather than keyword-centric approaches.

When to use: Search ranking optimization, structured data implementation, content audits, topical authority planning, featured snippet targeting, Core Web Vitals improvement.

When NOT to use: Paid advertising campaigns, social media strategy, email marketing, brand identity design.

Quick Reference

Common Mistakes

Delegation

• Audit a site for technical SEO issues, missing schema, and Core Web Vitals problems: Use Explore agent to crawl pages and identify structured data gaps
• Implement JSON-LD schema markup and meta tags across page templates: Use Task agent to add Organization, Article, FAQ, and Breadcrumb schemas
• Plan a topical authority strategy with pillar pages and content clusters: Use Plan agent to map keyword intent, cluster hierarchy, and internal linking architecture

References

• AI Overviews optimization and search visibility
• E-E-A-T standards and authority building
• Semantic entity SEO and topic clusters
• Schema markup patterns with JSON-LD
• Content audit methodology and scoring

AI Overviews Optimization

Google AI Overviews use LLMs to synthesize answers directly in search results. Visibility in these AI-generated summaries is a primary goal of content optimization.

Scannable Answer Boxes

AI models extract information from structured, scannable text. Format content so models can identify and synthesize key points.

Heading format: Use H2/H3 headings that are direct questions:

## How Does Lazy Loading Work?

**Lazy loading** defers the initialization of resources until they are needed.
It reduces initial page load time by loading images, scripts, and components
only when they enter the viewport or are triggered by user interaction.

Answer pattern: Follow the question heading with a 1-2 sentence direct answer before expanding into detail. Bold the core concept in the answer.

Entity Extraction Optimization

AI searches for entities (people, places, things, concepts) and their relationships. Make entities explicit rather than implied.

First paragraph rule: Define the core entity in the first paragraph of the page. State what it is, what category it belongs to, and its primary purpose.

Schema.org markup: Provide a machine-readable entity map using JSON-LD:

<script type="application/ld+json">
  {
    "@context": "https://schema.org",
    "@type": "Article",
    "headline": "How Lazy Loading Improves Performance",
    "author": {
      "@type": "Person",
      "name": "Jane Smith",
      "jobTitle": "Senior Performance Engineer"
    },
    "about": {
      "@type": "Thing",
      "name": "Lazy Loading",
      "description": "A design pattern that defers resource initialization"
    }
  }
</script>

Topical Depth (The 360-Degree Rule)

AI Overviews favor content that covers a topic from multiple angles. A single page should address:

Question-Based and Semantically Related Keywords

Question keywords target the query patterns that trigger AI Overview summaries:

• "What are the benefits of..."
• "Step-by-step guide to..."
• "How does X compare to Y..."

Semantically related terms are concepts that naturally co-occur in high-quality discussions of a topic. Include them without forcing:

Topic: "React Server Components"
Related terms: streaming, hydration, bundle size, server-side rendering,
               client components, use server directive, RSC payload

Visual Data Insights

AI Overviews can extract insights from tables, charts, and structured data.

Data table requirement: Include at least one data table or comparison chart per long-form article. Tables are highly extractable by AI models.

Image alt text: Describe the _insight_ of the image, not just its visual content:

<!-- Weak: describes appearance -->
<img alt="Bar chart showing three colored bars" />

<!-- Strong: describes the insight -->
<img
  alt="Performance comparison showing lazy loading reduces LCP by 40% compared to eager loading"
/>

Content Structure for AI Overview Visibility

Effective structure for AI extraction:

1. Lead with the answer -- first 50 words should directly answer the page's core question 2. Use numbered lists for processes -- AI models extract step-by-step content as featured snippets 3. Use comparison tables -- side-by-side format is preferred for "X vs Y" queries 4. Keep paragraphs to 2-4 sentences -- shorter blocks are easier for models to parse 5. Bold key terms on first mention -- signals term importance to extraction algorithms

Content Audit Methodology

A systematic approach to evaluating and optimizing content pages for search visibility. Uses a 30-point scoring system across seven categories.

Scoring Overview

Score interpretation:

Title Tag (4 Points)

Formula: [Concept]: [What You'll Understand] in [Topic]

Strong examples:

• "Closures: How Functions Remember Their Scope in JavaScript" (58 chars)
• "Event Loop: How Async Code Actually Runs in JavaScript" (54 chars)

Weak patterns: Too short ("Closures"), too long (>60 chars), no hook ("JavaScript Closures"), missing qualifier.

Meta Description (4 Points)

Formula: [Action word] [what it is] in [Topic]. [Specific things they'll learn]: [topic 1], [topic 2], and [topic 3].

Strong example: "Learn JavaScript closures and how functions remember their scope. Covers lexical scoping, practical use cases, memory considerations, and common closure patterns." (159 chars)

Keyword Placement (5 Points)

Density guideline: Do not exceed 3-4 mentions of the exact phrase per 1,000 words. Use variations naturally.

Content Structure (6 Points)

Content length guidelines:

Featured Snippets (4 Points)

Snippet format mapping:

Internal Linking (4 Points)

Anchor text quality:

Technical SEO (3 Points)

URL slug rules: lowercase, hyphens not underscores, include primary keyword, under 50 characters, no IDs or random strings.

Audit Report Template

After auditing, document findings with:

1. Score summary -- points per category with status 2. Target keywords -- primary and secondary keywords with search intent 3. Category-by-category analysis -- specific issues found per section 4. Priority fixes -- high/medium/low ranked action items 5. Implementation checklist -- verification steps after making fixes

Core Web Vitals Performance Check

While not scored in the 30-point content audit, performance directly impacts rankings. Verify these thresholds on every audited page:

Common performance fixes for content pages:

E-E-A-T Standards

Experience, Expertise, Authoritativeness, and Trust (E-E-A-T) are the signals search engines use to separate human-quality expertise from generic content.

Experience (The "I" Factor)

Content demonstrating first-hand experience ranks higher than abstract summaries.

Signals to include:

• Personal pronouns indicating direct experience: "In our testing," "I found that..."
• Unique screenshots, case study data, or proprietary research
• Specific metrics from real implementations rather than theoretical claims
• Before/after comparisons from actual projects

Example framing:

## Our Migration Results

After migrating 12 production services from REST to GraphQL over 6 months,
we measured a **34% reduction in API response time** and a **22% decrease
in frontend bundle size** due to eliminated over-fetching.

Expertise

Demonstrate domain knowledge through credentials and depth.

Author bios: Every major article should have a detailed author bio with:

• Professional credentials and current role
• Links to other published work or profiles
• Relevant experience summary (years, specializations)

SME review labels: High-stakes content (health, finance, legal) must be reviewed by a subject matter expert:

_Reviewed by Dr. Sarah Chen, Board-Certified Financial Analyst, CFA.
Last reviewed: January 2026._

Authoritativeness

Pillar page strategy: Build thorough pages (2,000+ words) that establish the site as a definitive source for a topic cluster. Each pillar page should:

• Cover the topic exhaustively from multiple angles
• Link to 5-10 supporting cluster articles
• Receive internal links back from those cluster articles
• Target the highest-volume keyword in the cluster

Backlink quality: Focus on citations from reputable domains in the specific industry rather than volume of links from unrelated sites.

Trust

Transparency signals:

• Clearly state publishing and last-updated dates
• Cite sources of data with links
• Disclose affiliations and sponsorships
• Display contact information and organizational details

Technical trust:

• HTTPS with valid certificates
• Core Web Vitals within acceptable thresholds (LCP < 2.5s, CLS < 0.1, INP < 200ms). After the December 2025 core update, sites with poor vitals face steeper traffic losses in competitive queries
• Accessible design (WCAG compliance)

AI Attribution

Transparency about AI usage is a trust signal.

Labeling pattern:

_Written by [Author Name], with AI assistance for data synthesis.
All facts verified by the author._

Validation requirement: Human verification of all AI-generated facts is mandatory. AI-generated content without expert review is flagged as low-quality by E-E-A-T filters.

E-E-A-T Checklist

Schema Markup Patterns

JSON-LD is the preferred format for structured data. It is embedded in a <script> tag and does not affect page rendering.

Core Patterns

Organization Schema

Define the publishing entity on every page:

<script type="application/ld+json">
  {
    "@context": "https://schema.org",
    "@type": "Organization",
    "name": "Your Company",
    "url": "https://yourcompany.com",
    "logo": "https://yourcompany.com/logo.png",
    "description": "Brief description of what the organization does",
    "sameAs": [
      "https://twitter.com/yourcompany",
      "https://linkedin.com/company/yourcompany",
      "https://github.com/yourcompany"
    ],
    "contactPoint": {
      "@type": "ContactPoint",
      "contactType": "customer support",
      "url": "https://yourcompany.com/support"
    }
  }
</script>

Article Schema

For blog posts, guides, and documentation pages:

<script type="application/ld+json">
  {
    "@context": "https://schema.org",
    "@type": "Article",
    "headline": "How to Optimize React Performance",
    "description": "A guide to React performance optimization techniques",
    "datePublished": "2026-01-15",
    "dateModified": "2026-01-28",
    "author": {
      "@type": "Person",
      "name": "Jane Smith",
      "url": "https://yourcompany.com/team/jane-smith",
      "jobTitle": "Senior Frontend Engineer"
    },
    "publisher": {
      "@type": "Organization",
      "name": "Your Company",
      "logo": {
        "@type": "ImageObject",
        "url": "https://yourcompany.com/logo.png"
      }
    },
    "mainEntityOfPage": {
      "@type": "WebPage",
      "@id": "https://yourcompany.com/guides/react-performance"
    }
  }
</script>

FAQ Schema

For pages with question-and-answer content:

<script type="application/ld+json">
  {
    "@context": "https://schema.org",
    "@type": "FAQPage",
    "mainEntity": [
      {
        "@type": "Question",
        "name": "What is React Server Components?",
        "acceptedAnswer": {
          "@type": "Answer",
          "text": "React Server Components render on the server and send only the HTML result to the client, reducing JavaScript bundle size and improving initial load performance."
        }
      },
      {
        "@type": "Question",
        "name": "How do Server Components differ from SSR?",
        "acceptedAnswer": {
          "@type": "Answer",
          "text": "SSR renders the full component tree on the server and hydrates on the client. Server Components never hydrate — they stay on the server permanently, eliminating their client-side JavaScript entirely."
        }
      }
    ]
  }
</script>

Breadcrumb Schema

For establishing page hierarchy in search results:

<script type="application/ld+json">
  {
    "@context": "https://schema.org",
    "@type": "BreadcrumbList",
    "itemListElement": [
      {
        "@type": "ListItem",
        "position": 1,
        "name": "Home",
        "item": "https://yourcompany.com"
      },
      {
        "@type": "ListItem",
        "position": 2,
        "name": "Guides",
        "item": "https://yourcompany.com/guides"
      },
      {
        "@type": "ListItem",
        "position": 3,
        "name": "React Performance"
      }
    ]
  }
</script>

HowTo Schema

For step-by-step tutorials:

<script type="application/ld+json">
  {
    "@context": "https://schema.org",
    "@type": "HowTo",
    "name": "How to Set Up Lazy Loading in React",
    "description": "Step-by-step guide to implementing lazy loading with React.lazy and Suspense",
    "step": [
      {
        "@type": "HowToStep",
        "name": "Import React.lazy",
        "text": "Replace static imports with React.lazy() for route-level components."
      },
      {
        "@type": "HowToStep",
        "name": "Wrap in Suspense",
        "text": "Add a Suspense boundary with a fallback component around the lazy-loaded component."
      },
      {
        "@type": "HowToStep",
        "name": "Test loading states",
        "text": "Verify the fallback renders during chunk loading using network throttling."
      }
    ]
  }
</script>

Implementation Guidelines

Validation

Test structured data using:

• Google's Rich Results Test: validates eligibility for rich results
• Schema.org Validator: validates syntax correctness
• Lighthouse audit: checks for structured data presence and errors

Common validation errors:

Semantic Entity SEO

Semantic SEO focuses on user intent and relationships between concepts rather than keyword density. Search engines map content to entities in a knowledge graph.

The Intent Map

Map every page to a specific user intent category:

Implementation: Add intent as metadata or a comment in the page frontmatter. Every page should serve exactly one primary intent.

Topic Clusters

Group related content around a central pillar page.

Architecture:

                    ┌──────────────┐
                    │  Pillar Page │
                    │  (3000+ words) │
                    └──────┬───────┘
                           │
          ┌────────────────┼────────────────┐
          │                │                │
    ┌─────┴─────┐   ┌─────┴─────┐   ┌─────┴─────┐
    │ Cluster A │   │ Cluster B │   │ Cluster C │
    │ (1500 words)│  │ (1500 words)│  │ (1500 words)│
    └───────────┘   └───────────┘   └───────────┘

Internal linking rules:

• Cluster articles link UP to the pillar page
• The pillar page links DOWN to each cluster article
• Cluster articles link ACROSS to related clusters where natural
• Use descriptive anchor text (not "click here")

Example cluster for "React Performance":

Schema.org for Entities

Use structured data to explicitly define entities for the knowledge graph.

Organization schema -- define who you are:

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Acme Corp",
  "url": "https://acme.com",
  "logo": "https://acme.com/logo.png",
  "sameAs": ["https://twitter.com/acmecorp", "https://github.com/acmecorp"]
}

FAQ schema -- define the questions you answer:

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What is lazy loading?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Lazy loading defers resource initialization until needed."
      }
    }
  ]
}

Breadcrumb schema -- define the page hierarchy:

{
  "@context": "https://schema.org",
  "@type": "BreadcrumbList",
  "itemListElement": [
    {
      "@type": "ListItem",
      "position": 1,
      "name": "Home",
      "item": "https://example.com"
    },
    {
      "@type": "ListItem",
      "position": 2,
      "name": "Guides",
      "item": "https://example.com/guides"
    },
    { "@type": "ListItem", "position": 3, "name": "Performance" }
  ]
}

NLP-Friendly Writing

Write in patterns that NLP models parse reliably.

Active sentences: Subject, verb, object order. Avoid passive voice for key definitions.

Strong: "React Server Components render on the server."
Weak:   "Rendering on the server is done by React Server Components."

Semantically related terms: Include terms that naturally co-occur in quality discussions of the topic. Do not force them; weave them in where they fit.

Consistent terminology: Pick one term and use it throughout. Do not alternate between "endpoint," "route," and "URL" for the same concept.

Content Freshness

AI-powered search engines favor current data.

Refresh protocol: Update critical articles every 6 months with:

• Current statistics and benchmarks
• Updated tool versions and recommendations
• New patterns that have emerged since publication
• Removal of deprecated approaches