rampstackco/claude-skills / skills/pillar-content-architecture
pillar-content-architecture
VerifiedHow to design a content hub that earns topical authority. Pillar topic selection, cluster planning, internal linking architecture, URL structure, pillar and cluster page anatomy, topical authority signals for SEO and AEO/GEO, and the maintenance discipline that distinguishes intentional hubs from accidental orphans. Triggers on pillar content, content hub, topic cluster, topical authority, content architecture, hub and spoke, pillar page, cluster page, content silo, internal linking strategy. Also triggers when a content set is not ranking despite individual piece quality, when a pillar was launched without a cluster, or when content has accumulated without an architecture.
| File | Kind | Size |
|---|---|---|
| README.md | markdown | 368 B |
| SKILL.md | skill md | 21.6 KB |
| references/cluster-piece-anatomy.md | markdown | 6.3 KB |
| references/cluster-planning-patterns.md | markdown | 6.1 KB |
| references/common-pillar-failures.md | markdown | 9.2 KB |
| references/content-refresh-patterns.md | markdown | 7.8 KB |
| references/internal-linking-architecture.md | markdown | 7.1 KB |
| references/pillar-cluster-decision.md | markdown | 5.7 KB |
| references/pillar-page-anatomy.md | markdown | 7.3 KB |
| references/topic-selection-criteria.md | markdown | 6.2 KB |
| references/topical-authority-signals.md | markdown | 6.8 KB |
| references/url-structure-patterns.md | markdown | 6.2 KB |
references/cluster-planning-patterns.md
View on GitHub# Cluster planning patterns
Facet sourcing methods, sequencing patterns, heterogeneity check.
---
## Facet sourcing methods
A pillar's cluster is built from 8 to 15 facets. Source them deliberately rather than from the writer's imagination.
### Method 1: keyword research expansion
Start with the pillar keyword. Pull the long-tail variations from keyword tools (Ahrefs, Semrush, GSC, dataForSEO). Filter by intent: each cluster facet should have its own SERP intent and dominant format. Variations that share intent with the pillar are not separate facets; they belong inside the pillar itself.
### Method 2: "people also ask" extraction
Pull the PAA box for the pillar keyword and 3 to 5 supporting keywords. Each unique question is a candidate cluster facet. PAA is the cleanest signal of what readers actually ask in this topic space.
### Method 3: top-ranking competitor outline mining
Pull the top 5 SERP results for the pillar keyword. Extract H2 sections from each. Cluster the H2s across the 5 competitors; sections that appear in 3+ competitors are facets the topic objectively requires. Sections unique to 1 or 2 competitors may be differentiation opportunities.
### Method 4: first-party customer questions
Sales calls, support tickets, customer success notes, sales enablement Q&A. The questions customers actually ask are facets the pillar needs to address even if SERP signal is weak. First-party questions also produce differentiation: questions your customers ask but that competitor content does not address.
### Method 5: Frase or AirOps SERP analysis
Frase's entity gap analysis and AirOps' workflow builder both run versions of methods 1 to 3 in software. The output is a candidate facet list with SERP coverage scores. Use the tool output as the starting point; refine with first-party questions (method 4) for differentiation.
---
## Sequencing patterns
How to ship the hub over time.
### Pattern A: pillar first, cluster follows
Ship the pillar as a 3,000 to 5,000 word piece. Over the next 6 to 12 weeks, ship 1 to 2 cluster pieces per week. By month 3, the hub has the pillar plus 8 to 12 clusters and starts compounding.
**When this works.** Strong editorial team that can ship a pillar quickly. The pillar drives initial rankings; clusters fill in topical depth as they ship.
**The risk.** Pillar ranks for 6 weeks before the cluster is complete; AI engines see a thin hub and may not return to crawl when clusters land. Mitigated by including cluster placeholders in the pillar with "coming soon" or by pre-publishing cluster stubs.
### Pattern B: pillar plus first 5 clusters concurrent
Ship the pillar and the first 5 clusters within the same 4-week window. The hub launches with the link graph already partially formed; rankings start with depth signals already in place.
**When this works.** Larger team or agency support. The launch produces a more credible hub on day 1.
**The cost.** Higher up-front investment. The first 5 clusters lock the team's editorial capacity for the launch window.
### Pattern C: prove with 3 clusters, then commit to pillar
Ship 3 cluster pieces under a planned-but-not-yet-shipped pillar slot. Measure performance over 8 to 12 weeks. If the clusters draw traffic and the topic shows commercial signal, commit to the pillar. If not, pivot to a different pillar topic without the sunk cost.
**When this works.** Uncertain topic where the commercial signal needs validation before the pillar investment. Common in agencies running pillars for clients without strong upfront conviction.
**The discipline.** The 3 clusters must be writeable as standalones during the validation window; they cannot reference the pillar by name yet because it does not exist. Once the pillar ships, retrofit the cluster pieces with pillar callouts and bottom-up links.
---
## Heterogeneity check
A cluster passes heterogeneity if its facets cover the topic from multiple distinct angles. The check:
**Categorize each facet** by primary angle:
- Definition (what is X)
- How-to (how to do X)
- Comparison (X vs Y)
- Examples (5 examples of X, X case studies)
- Mistakes (common X mistakes)
- Costs (how much X costs, X pricing models)
- Alternatives (alternatives to X)
- Best-of (best X tools, top X providers)
- Use cases (when to use X)
- Niche variants (X for [specific audience])
- Foundational (related concept Y that supports X)
- Advanced (advanced X techniques)
**The rule.** A 12-cluster hub should have facets across 6+ distinct angles. If 10 of 12 are "how-to," the cluster is homogeneous and the pillar will look like a how-to series rather than a topical authority hub.
**The fix.** When facet brainstorming over-indexes on one angle, deliberately add 2 to 3 facets from different angles. The "5 examples of X" cluster piece is often the most under-served angle; teams forget to add concrete-example pieces and over-invest in abstract how-tos.
---
## No facet duplication
Two cluster pieces should not target overlapping queries. The check:
- For each cluster piece, list the 3 to 5 keywords it would target.
- For each pair of cluster pieces, check whether their keyword lists overlap by more than 30%.
- Overlapping clusters compete for the same SERP and self-cannibalize. Consolidate to one piece.
**The most common duplication.** "What is X" and "X explained" and "X definition" are the same facet. Pick one slug and one focus; do not split the foundational definition across three pieces.
---
## When the cluster is too big
A planned cluster with 25+ pieces is over-built. The hub becomes unmanageable: maintenance burden too high, link graph too dense to audit, reader navigation confusing.
**Fix 1.** Split into two pillars under a parent pillar. The original pillar covers the broadest topic; sub-pillars cover major sub-territories with their own clusters. Three-level hierarchy is rare but valid for very broad topics.
**Fix 2.** Trim to the 12 most-impactful facets. Rank facets by combined search volume + commercial relevance + heterogeneity contribution; keep the top 12; archive the rest as future-cluster candidates.
The 12-cluster sweet spot is empirical. Below 8, the hub looks thin; above 15, maintenance overhead kicks in; 10 to 12 is the band where most healthy hubs sit.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |