semantic-seo

name	semantic-seo
description	Semantic SEO toolkit — pillar/cluster topic maps, SERP analysis, entity SEO, content-gap mapping, keyword research framework, internal-linking strategy.
user-invocable	true
allowed-tools	["Read","Glob","Grep","Write"]

Semantic SEO Skill

Owned by seo-analyst. Powers organic-search growth as a durable, non-paid acquisition channel. Modern SEO is semantic — search engines reward topical authority and entity coverage, not keyword stuffing.

1. Topic Cluster Model (Pillar + Cluster + Spoke)

flowchart TD
    Pillar["Pillar: Marketing Attribution Complete Guide"]
    C1["Cluster: MMM Fundamentals"]
    C2["Cluster: MTA Methods"]
    C3["Cluster: Incrementality Testing"]
    S1a["Spoke: Bayesian MMM in 30 days"]
    S1b["Spoke: Adstock and saturation curves"]
    S2a["Spoke: Shapley vs Markov MTA"]
    S2b["Spoke: Last-touch vs data-driven"]
    S3a["Spoke: Geo-lift sample size calculator"]
    S3b["Spoke: Ghost-ad test playbook"]
    Pillar --> C1 --> S1a
    C1 --> S1b
    Pillar --> C2 --> S2a
    C2 --> S2b
    Pillar --> C3 --> S3a
    C3 --> S3b

Rules:

• One pillar per topic per site (avoid keyword cannibalization)
• Pillar is 2,500-5,000 words, comprehensive, periodically refreshed
• Cluster pages link UP to pillar; spokes link UP to cluster and SIDEWAYS to siblings
• Pillar links DOWN to every cluster; never to every spoke (curate)

2. SERP Analysis Checklist

For each target keyword, before writing:

• Intent classification: navigational / informational / commercial / transactional
• SERP features present: featured snippet, People-Also-Ask, knowledge panel, image pack, video carousel, shopping, local pack, AI Overview
• Top 10 organic content type: listicle / how-to / definition / comparison / tool / case study / product page
• Top 10 word count distribution: median and IQR
• Domains ranking: are they DR-equal, weaker (opportunity), or stronger (barrier)
• Last-updated freshness: avg age of top-10 results
• AI Overview present: if yes, what is cited; can we earn citation?
• Brand vs publisher mix: helps decide content angle

3. Entity SEO Concepts

Search engines understand entities (people, places, organizations, products, concepts) and their relationships (knowledge graph), not just strings.

Practical Application

• Schema.org markup on every templated page:
  • Organization on homepage and about
  • Product or Service on offering pages
  • Article with author (linked to Person entity) on content
  • FAQPage on pages with explicit Q&A
  • BreadcrumbList on category/cluster pages
  • LocalBusiness for service-area businesses
• Entity establishment for the brand:
  • Wikidata entry (where notable)
  • Consistent NAP (Name/Address/Phone) across the web
  • Knowledge Panel claim (Google Business / Search Console)
  • Author profiles with traceable cross-links
• Co-occurrence: mention semantically related entities in content (e.g. an MMM article should reference adstock, saturation, Bayesian inference, MCMC, geo-lift)

4. Content-Gap Mapping Process

Step-by-step:

1. Inventory current content: crawl the site, classify by topic
2. Pull competitor content: identify 3-5 organic competitors (Ahrefs / Semrush / SimilarWeb)
3. Keyword overlap analysis: keywords competitors rank top-20 for that we don't
4. Filter by relevance: drop keywords outside our taxonomy
5. Cluster by topic: group keywords into themes
6. Score each cluster: search volume * intent * business value * winnability
7. Prioritize: top 5-10 clusters become content roadmap
8. Assign: pillar, cluster, or spoke status

Tools

• Ahrefs (content gap, Top Pages, Keywords Explorer)
• Semrush (keyword gap, Topic Research)
• SimilarWeb (traffic, top pages of competitors)
• Google Search Console (existing performance, query-page mismatch)
• Google Trends (seasonality, regional intent)

5. Keyword Research Framework

Score each candidate keyword across four dimensions:

Dimension	Source	Score 1-5
Search volume	Keyword tool	1 = <100/mo, 5 = >10k/mo
Keyword difficulty (KD)	Ahrefs/Semrush DR	Inverted: 1 = KD>70, 5 = KD<20
Intent fit	Manual SERP review	1 = mismatch, 5 = perfect fit
Business value	Internal	1 = brand awareness only, 5 = bottom-of-funnel

Composite = mean of four. Prioritize composite >= 3.5.

Keyword Type Mix Target

Type	Examples	% of content effort
Pillar (head, informational)	"marketing attribution"	10%
Cluster (mid-tail, informational/commercial)	"what is MMM"	30%
Spoke (long-tail, mixed)	"how to run a geo-lift test"	40%
Bottom-funnel (transactional)	"MMM consulting"	15%
Branded	"[brand] case study"	5%

6. Internal Linking Strategy

Principle	Implementation
Topic hubs	Every cluster has a hub page linking to all its spokes
Anchor diversity	Use varied descriptive anchors, not just exact-match keyword
Reciprocal cluster links	Sibling spokes link to each other when topically related
Pillar from sitewide	Pillar in main nav or footer (every page link-equity)
New content gets links	When publishing a new spoke, add 2-3 links from existing high-traffic pages
Audit orphans	Quarterly: no page should be >3 clicks from home

Anchor Text Distribution Target

• 40% exact / partial match
• 30% branded / URL
• 20% generic ("learn more", "see this")
• 10% long-tail descriptive

7. Industry-Aware Adjustments

Profile	SEO emphasis
B2B SaaS	Pillar+cluster on category, competitor comparison pages, integration pages
DTC E-com	Category + collection pages, product schema, review markup, gift guides
Pro Services	Service pages, case studies, local + service-area SEO
Regulated	Compliance-reviewed content, fair-balance disclaimers, no claim ambiguity
Creative-Production	Portfolio pages, case studies, local SEO, NAP consistency
Ad-Commercial	Thought-leadership, original research, agency directory listings

8. Output Deliverables

When seo-analyst runs, it produces:

• output/research/seo/pillar-cluster-map.md — mermaid diagram + table
• output/research/seo/serp-analysis-<keyword>.md — per priority keyword
• output/research/seo/content-gap-report.md — prioritized roadmap
• output/research/seo/keyword-research.csv — scored keyword inventory
• output/research/seo/internal-link-audit.md — orphans + opportunities

References

• marketing-expertise (E-E-A-T checklist supports authority signals)
• audience-segmentation (intent matching against personas)
• content-marketing topic taxonomy in marketing-expertise