cosmicstack-labs/seo-strategy
Comprehensive guide to search engine optimization strategy covering technical SEO, content optimization, on-page and off-page tactics, analytics, and performance measurement. Designed for practitioners building organic search presence from the ground up.
npx skills add https://github.com/cosmicstack-labs/mercury-agent-skills --skill seo-strategyGoogle's mission is to organize the world's information and make it universally accessible. Every algorithm update rewards content and technical setups that serve user intent. SEO is not about "gaming the system" — it is about making your site the best answer to a searcher's question.
No amount of technical optimization can save thin, irrelevant, or unhelpful content. SEO begins with understanding what your audience searches for and creating resources that genuinely satisfy that need. Technical SEO opens the door; content quality earns the ranking.
Search engines evaluate authority through signals like backlinks, brand mentions, E-E-A-T signals, and user engagement metrics. You cannot buy authority or shortcut it with spammy tactics. Consistent value delivery over time is the only durable path.
If search engines cannot crawl, render, and index your pages, your content quality is irrelevant. Technical SEO is not optional — it is the prerequisite for every other effort. Core Web Vitals, mobile usability, and structured data are minimum requirements in 2024.
Meaningful organic growth takes 3–6 months minimum for new content and 6–18 months for new domains. Algorithm updates, competitor moves, and shifting user behavior mean SEO requires continuous investment. Quick wins exist, but sustainable rankings are earned through consistency.
Every SEO decision should be grounded in data: search volume, click-through rates, conversion rates, ranking positions, crawl errors, and user behavior. If you cannot measure it, you cannot optimize it.
| Level | Name | Characteristics | Technical Foundation | Content Approach | Measurement |
|-------|------|----------------|---------------------|------------------|-------------|
| L1 | Unoptimized | No tracking, no structured approach, content published without SEO consideration | No sitemap, default robots.txt, no structured data | Ad-hoc, no keyword research | None or basic Google Analytics |
| L2 | Foundational | Basic keyword research, manual tracking, meta tags optimized | Static sitemap, basic robots.txt, manual canonical tags | Keyword-stuffed, single-page focus | Google Search Console set up |
| L3 | Systematic | Structured keyword strategy, regular content publishing, basic link building | Dynamic sitemap, automated canonicalization, Core Web Vitals monitored | Topic clusters, content calendar, keyword mapping | Monthly reporting, rank tracking |
| L4 | Data-Driven | Content gap analysis, competitive intelligence, programmatic SEO | Structured data across site, CDN/edge optimization, mobile-first | Authority building, E-E-A-T signals, pillar pages | Conversion tracking, attribution modeling |
| L5 | Optimized & Scaling | Automated SEO workflows, predictive analytics, multi-market | Full SEO automation, real-time monitoring, AI-assisted optimization | Content at scale with quality control | Full attribution, predictive ranking, ROI modeling |
Search engines discover pages through crawling. If a page cannot be crawled, it cannot rank.
Crawl Budget Management: Google allocates a crawl budget to each site. Large sites (10,000+ pages) must optimize crawl efficiency.
# Pseudocode for crawl budget optimization
prioritize_pages = {
"high": ["/products/*", "/guides/*"], # Crawl daily
"medium": ["/blog/*", "/resources/*"], # Crawl weekly
"low": ["/tags/*", "/archive/*"], # Crawl monthly
"noindex": ["/admin/*", "/cart/*", "/search/*"] # Blocked
}
Common crawl blockers:
Being crawled does not guarantee being indexed. Indexability means the search engine has stored the page in its database and can show it in results.
Indexability checklist:
index not noindex)Sitemaps tell search engines which pages exist and when they were last updated.
Best practices:
<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
<url>
<loc>https://example.com/ultimate-guide-seo</loc>
<lastmod>2024-12-01</lastmod>
<changefreq>monthly</changefreq>
<priority>0.9</priority>
</url>
<url>
<loc>https://example.com/blog</loc>
<lastmod>2024-12-15</lastmod>
<changefreq>weekly</changefreq>
<priority>0.8</priority>
</url>
</urlset>
Sitemap: https://example.com/sitemap.xmlControls which parts of your site search engines can crawl.
Example:
User-agent: *
Disallow: /admin/
Disallow: /cart/
Disallow: /checkout/
Disallow: /search/
Disallow: /*?sort=
Disallow: /*filter=
Sitemap: https://example.com/sitemap.xml
Common mistake: Using Disallow: / to block all crawlers during development, then forgetting to remove it in production. Always verify robots.txt via Search Console.
Google's performance metrics measuring real-world user experience. They are ranking signals.
| Metric | What It Measures | Good | Needs Improvement | Poor |
|--------|-----------------|------|-------------------|------|
| LCP (Largest Contentful Paint) | Loading performance | ≤ 2.5s | 2.5s – 4.0s | > 4.0s |
| FID/INP (Interaction to Next Paint) | Interactivity | ≤ 100ms (FID) / ≤ 200ms (INP) | 100–300ms / 200–500ms | > 300ms / > 500ms |
| CLS (Cumulative Layout Shift) | Visual stability | ≤ 0.1 | 0.1 – 0.25 | > 0.25 |
Optimization tactics:
requestIdleCallbackwidth and height attributes), reserve space for ads/embeds, avoid inserting content above existing contentHelps search engines understand page content and enables rich results (featured snippets, knowledge panels, FAQ rich results).
Common schema types:
{
"@context": "https://schema.org",
"@type": "Article",
"headline": "Complete Guide to SEO Strategy",
"description": "A comprehensive guide covering technical SEO, content optimization, and measurement.",
"author": {
"@type": "Person",
"name": "SEO Specialist"
},
"datePublished": "2024-01-15",
"dateModified": "2024-12-01",
"image": "https://example.com/images/seo-guide.jpg"
}
Priority schemas for most sites:
Testing tools: Google Rich Results Test, Schema.org validator
Keyword research is the practice of identifying the terms your target audience uses in search engines. It informs content creation, optimization, and strategy.
Research process:
Keyword difficulty assessment:
def priority_score(search_volume, difficulty, relevance):
"""
Calculate keyword priority on a 0-100 scale.
Higher is better.
"""
volume_score = min(search_volume / 1000, 100) # Normalize volume
diff_score = 100 - difficulty # Lower difficulty = better
return (volume_score * 0.3) + (diff_score * 0.4) + (relevance * 0.3)
The topic cluster model organizes content around pillar pages supported by cluster content.
Structure:
Pillar Page: "Complete SEO Guide" (comprehensive, covers broad topic)
├── Cluster: "Keyword Research" → detailed guide
├── Cluster: "Technical SEO" → detailed guide
├── Cluster: "Link Building" → detailed guide
├── Cluster: "SEO Analytics" → detailed guide
└── Cluster: "Local SEO" → detailed guide
How it works:
Benefits: Higher rankings for both broad and specific queries, improved crawl efficiency, stronger topical authority signals.
Topic authority is the degree to which a website is recognized as an expert on a particular subject. Google evaluates this through:
Building topic authority:
E-E-A-T is Google's framework for evaluating content quality, especially for Your Money or Your Life (YMYL) topics.
| Component | What It Means | How to Demonstrate |
|-----------|---------------|-------------------|
| Experience | First-hand knowledge | Personal stories, case studies, original research |
| Expertise | Formal or demonstrated knowledge | Author credentials, cited sources, technical accuracy |
| Authoritativeness | Recognized as a go-to source | Backlinks from industry leaders, brand mentions |
| Trustworthiness | Honest, accurate, transparent | Clear sourcing, author bios, contact info, secure site |
Actionable steps:
Title tags are the clickable headline in search results. They are the single most important on-page SEO element.
Formula: Primary Keyword | Brand Name or Primary Keyword: Secondary Keyword
Best practices:
Examples:
<!-- Good -->
<title>SEO Strategy Guide: Technical, Content & On-Page Optimization</title>
<!-- Bad -->
<title>SEO | Home | Blog | Products | Contact Us</title>
Meta descriptions are the summary snippet under the title in search results. They do not directly affect rankings but significantly impact click-through rates.
Best practices:
Examples:
<!-- Good -->
<meta name="description" content="Learn proven SEO strategies for 2024. This comprehensive guide covers technical optimization, content creation, link building, and measurement. Start ranking higher today." />
<!-- Bad -->
<meta name="description" content="SEO tips and tricks and strategies and best practices for SEO optimization and ranking." />
Headers structure content for both users and search engines. They signal the hierarchy and topics of your page.
Rules:
Example structure:
<h1>Complete Guide to SEO Strategy</h1>
<h2>What is SEO?</h2>
<h2>Technical SEO</h2>
<h3>Crawlability</h3>
<h3>Indexability</h3>
<h2>Content SEO</h2>
<h3>Keyword Research</h3>
<h3>Content Clusters</h3>
Internal links distribute page authority throughout your site and help search engines understand site structure.
Best practices:
Anchor text best practices:
<!-- Good -->
<p>Learn more about <a href="/technical-seo">technical SEO best practices</a> to improve crawlability.</p>
<!-- OK -->
<p>Check out our <a href="/technical-seo">Technical SEO Guide</a>.</p>
<!-- Bad -->
<p>Click <a href="/technical-seo">here</a> to read about technical SEO.</p>
Backlinks (inbound links from other sites to yours) remain one of the strongest ranking signals. Quality matters far more than quantity.
Link quality factors:
Link acquisition strategies:
Outreach email template:
Subject: Quick suggestion for your [Article Title] page
Hi [Name],
I was reading your excellent piece on [Topic] and noticed the
section about [Subtopics]. I thought you might find our
[Content Title] helpful — it covers [Specific Value] with
original data from [Source].
Here's the link: [URL]
No pressure at all, but if it adds value for your readers,
a mention would be appreciated. Either way, great work on
the article!
Best,
[Your Name]
Domain Authority is a third-party metric (developed by Moz) that predicts how well a site will rank. While not a Google metric, it is useful for competitive benchmarking.
Factors that influence DA:
Important: Do not obsess over DA. It is a correlational metric, not a causal one. Focus on the underlying signals (quality content, good backlinks, technical health) and DA will follow.
Free tool from Google for monitoring and maintaining your site's presence in search results.
Key features to use:
Weekly monitoring checklist:
□ Check for new crawl errors
□ Review performance trends (last 28 days vs previous period)
□ Inspect any pages with sudden traffic drops
□ Verify new sitemap submissions
□ Check Core Web Vitals for regressions
Essential for understanding what happens after users arrive from search.
Key reports:
Key events to track:
// GA4 event tracking examples
gtag('event', 'page_view');
gtag('event', 'view_item', { items: [{id: 'product_123'}] });
gtag('event', 'add_to_cart', { items: [{id: 'product_123'}] });
gtag('event', 'purchase', {
transaction_id: 'T12345',
value: 49.99,
currency: 'USD'
});
gtag('event', 'form_submit', { form_name: 'newsletter' });
Tools that simulate how search engines crawl your site.
| Tool | Best For | Key Feature |
|------|----------|-------------|
| Screaming Frog | Site audits, finding broken links, duplicate content | Crawls up to 500 URLs free |
| Sitebulb | Visual site audits with actionable reports | Prioritizes issues by impact |
| Ahrefs Site Audit | Continuous site health monitoring | Integrates with rank tracking |
| DeepCrawl (Lumar) | Enterprise-scale crawling | API access for automation |
| Google Search Console | Official Google crawl data | Real-world crawl data, not simulated |
What to audit monthly:
Monitor keyword positions over time.
| Tool | Strengths | Limitations |
|------|-----------|-------------|
| Ahrefs | Largest index, accurate | Expensive |
| SEMrush | Good for competitive analysis | Keyword data can vary |
| Moz Pro | Beginner-friendly | Smaller index |
| Serpstat | Affordable | Less accurate for low-volume keywords |
| AccuRanker | High accuracy, fast | Limited features beyond tracking |
| KPI | What It Measures | Target Movement | How to Track |
|-----|-----------------|-----------------|--------------|
| Organic Traffic | Total visits from search engines | ↑ Month-over-month | GA4 (Traffic Acquisition) |
| Keyword Rankings | Position in search results for target terms | ↑ Average position | Rank tracking tool |
| Click-Through Rate (CTR) | % of impressions that result in clicks | ↑ Match search intent | Google Search Console |
| Impression Share | % of eligible impressions you appear for | ↑ Target 80%+ for key terms | Google Search Console |
| Conversion Rate | % of organic visitors who complete a goal | ↑ Align content with intent | GA4 (Events) |
| Bounce Rate | % of users who leave after one page | ↓ (for content pages) | GA4 (Engagement) |
| Pages Per Session | Average pages viewed per visit | ↑ Indicates content engagement | GA4 (Engagement) |
| Crawl Budget Usage | How many pages Google crawls per day | ↑ for new content, ↓ for waste | Search Console (Settings) |
| Index Coverage | % of submitted pages that are indexed | ↑ Target 95%+ | Search Console (Coverage) |
| Core Web Vitals | Real-user performance metrics | All "Good" threshold | Search Console + PageSpeed Insights |
Weekly (15 minutes):
Monthly (1 hour):
Quarterly (2–3 hours):
Understanding which organic touchpoints drive conversions:
| Model | Description | Best For |
|-------|-------------|----------|
| Last Click | Credit goes to the last interaction before conversion | Simple reporting |
| First Click | Credit goes to the first interaction | Top-of-funnel awareness |
| Linear | Equal credit to all interactions | Balanced view |
| Time Decay | More credit to interactions closer to conversion | Short sales cycles |
| Position Based | 40% first, 40% last, 20% middle | Complex buying journeys |
| Data-Driven | Algorithmically assigned based on actual impact | Large data sets (ML-based) |
The mistake: Publishing multiple pages targeting the same keyword, causing pages to compete against each other.
Fix: Map one primary keyword per page. If pages overlap, consolidate content or add canonical tags. Use a keyword-to-page mapping spreadsheet.
The mistake: Creating content that targets a keyword but does not match what the searcher actually wants.
Fix: Before writing, search the keyword and study the top 10 results. If they are mostly listicles, write a listicle. If they are guides, write a guide. Match the format, depth, and angle.
The mistake: Buying backlinks, using PBNs, keyword stuffing, or other black-hat tactics that work temporarily but trigger penalties.
Fix: Commit to white-hat tactics. Organic growth is slower but durable. A Google penalty can wipe out months of progress overnight.
The mistake: Focusing only on content and keywords while ignoring site speed, mobile usability, crawl errors, or structured data.
Fix: Run a technical audit quarterly. Fix critical issues (crawl errors, slow pages, broken links) before creating new content.
The mistake: Publishing hundreds of low-value pages (auto-generated content, short blog posts, shallow category pages) hoping volume will drive traffic.
Fix: Quality over quantity. One excellent 3,000-word guide outperforms ten 300-word posts. Google rewards depth and utility.
The mistake: Using exact-match anchor text for every internal or external link (e.g., always linking "best SEO tools" to the same page).
Fix: Vary anchor text naturally. Use partial matches, branded terms, and generic phrases. Over-optimization triggers spam filters.
The mistake: A desktop-only experience that fails on mobile devices.
Fix: Google uses mobile-first indexing. Test every page on mobile. Ensure touch targets are large enough, text is readable without zooming, and content renders without horizontal scrolling.
The mistake: Doing SEO work without measuring results, so you cannot tell what is working.
Fix: Set up tracking before starting. Define KPIs. Report monthly. If you cannot measure it, do not optimize it.
The mistake: Publishing content once and never updating it. Rankings decay as content becomes outdated.
Fix: Review and refresh content annually. Update statistics, add new information, improve readability, and re-optimize for current best practices.
The mistake: Optimizing for ranking and clicks without caring whether visitors convert.
Fix: SEO does not end at the click. Ensure landing pages have clear CTAs, compelling value propositions, and a path to conversion. Track organic conversion rate as a primary KPI.
The mistake: Each page exists in isolation with no strategic internal link structure.
Fix: Build content clusters with pillar pages. Every new piece of content should link to at least 2–3 existing pages. Use descriptive anchor text.
The mistake: HTTP and HTTPS versions both indexable, www and non-www, trailing slash variations, or URL parameters creating identical pages.
Fix: Choose a canonical URL structure (e.g., https://www.example.com). Set 301 redirects for all variations. Use canonical tags consistently. Fix parameter-based duplicates.
The mistake: Adding incorrect or spammy structured data thinking it guarantees rich results.
Fix: Structured data helps search engines understand content — it does not guarantee rich results. Use valid schemas. Test with Google's Rich Results Test. Never markup content that is not visible to users.
The mistake: Doing an SEO push for 3 months and then stopping, expecting permanent results.
Fix: SEO is ongoing. Algorithm updates, competitor activity, and content decay require continuous investment. Build SEO into your regular content and development processes.
Take cosmicstack-labs/seo-strategy from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.