Flagship Skill ยท Pillar content architecture

The pillar content architecture skill.

Pillar, cluster, and orphan content are three different things.

A senior content architect's playbook for designing content hubs that earn 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.

Audience: SEO content strategists, content architects, in-house teams designing content hubs, agencies running topical authority programs.

View the skill on GitHubBrowse the full catalog

What this skill is for

The hub-scope skill in a six-skill content workflow.

Six skills compose into the content discipline. Each one has its own scope; the distinction is what keeps each skill sharp. This skill plugs in between strategy plus research and per-piece briefing plus writing.

  1. 01

    content-strategyPROGRAM scope

    Decides what topic areas to invest in across a quarter or year.

  2. 02

    seo-keywordRESEARCH input

    Surfaces specific keyword opportunities with volume, difficulty, intent.

  3. 03

    pillar-content-architectureHUB scope (this skill)

    Designs the hub structure for one chosen topic: pillar plus cluster plus link graph plus URL pattern plus refresh cadence.

  4. 04

    content-brief-authoringPER-PIECE scope

    Briefs each piece within the hub.

  5. 05

    content-and-copyEXECUTION scope

    Writes each piece.

  6. 06

    seo-content-gap-auditAUDIT-SIDE

    Catches pieces still missing from the hub later.

The keystone distinction

Three categories of content. Most programs are mostly the broken one.

Search engines and AI engines both read internal-link graphs to infer topical authority. Same content volume, different architectural outcome. The discipline is moving toward intentional clusters and pillars and away from accidental orphan accumulation.

The anchor

Pillar

Comprehensive piece covering a major topic area. 2,500 to 5,000 plus words. Anchors the topical authority claim. Receives links from cluster pieces; links out to clusters selectively, never as a related-reading footer dump.

The orbit

Cluster

Narrower piece covering one facet of the pillar topic. 800 to 2,000 words. Links up to pillar in first 200 words and closing; may link sideways to sibling clusters when natural. Each cluster answers one specific question.

The drift

Orphan

Piece that does not connect to a pillar or cluster. May rank or convert standalone but does not compound into topical authority. Some orphans are fine by design (release notes, customer stories); most are accidents.

The pathology. Most content programs are 80% orphans, 15% accidental clusters, 5% accidental pillars. The discipline is moving toward 70% intentional clusters, 20% intentional pillars, 10% standalone-by-design.

Hub anatomy

What a hub actually looks like.

The pillar at the center; 8 to 12 cluster pieces around it; internal links flowing top-down (pillar to cluster callouts), bottom-up (cluster pieces linking up to pillar), and lateral (cluster to cluster, selective). The link graph is the architecture; without it, the same pieces are just a pile of orphans.

Hub anatomy: pillar at the center linked to eight cluster piecesA pillar at the center of the hub is connected by lines to eight cluster pieces orbiting around it, each cluster representing a distinct facet of the pillar topic.ClusterDefinitionClusterHow-toClusterComparisonClusterExamplesClusterMistakesClusterCostsClusterAlternativesClusterUse casesPillar3,000 to 5,000 wordsAnchors the topic
The pillar (center) is comprehensive at 3,000 to 5,000 words. Each cluster orbits at 800 to 2,000 words covering one specific facet (definition, how-to, comparison, examples, mistakes, costs, alternatives, use cases). Real hubs typically run 10 to 12 clusters; eight shown here for legibility.

The framework

Twelve considerations for hub architecture.

When designing or auditing a content hub, walk these 12 considerations. Each one names a discipline; skipping any of them produces one of the failure modes.

  1. 01Topic earns the pillar (5 criteria)
  2. 02Cluster size targets 10 to 12
  3. 03Facet heterogeneity across angles
  4. 04No facet duplication across clusters
  5. 05Pillar to cluster to pillar link graph
  6. 06Anchor text discipline; no stuffing
  7. 07URL structure signals the hub
  8. 08Pillar anatomy includes TL;DR and FAQ
  9. 09Cluster anatomy links UP to pillar early
  10. 10AEO/GEO and SEO together, not competing
  11. 11Annual pillar refresh + triggered cluster refresh
  12. 12Hub ownership is durable across 3 to 5 years

What is in the skill

Thirteen sections covered in the body.

The SKILL.md spans the full hub-architecture discipline from the keystone framing through topic selection, cluster planning, link graph design, URL structure, page anatomies, topical authority signals, and the refresh cadence that keeps hubs alive.

  1. 01

    What this skill covers

    Composition with five sister and adjacent skills (content-strategy, seo-keyword, content-brief-authoring, content-and-copy, seo-content-gap-audit). The clean reading order: program decides what, research surfaces opportunities, this skill designs the hub, briefs go per-piece, writing executes, audit catches gaps.

  2. 02

    Pillar vs cluster vs orphan content

    The keystone distinction. Pillars anchor; clusters orbit; orphans drift. Most content programs are 80% orphans; the discipline is intentional architecture that compounds. Same content volume, different architectural outcome.

  3. 03

    Topic selection for pillar pages

    Five-criterion framework: search volume, facet count, commercial relevance, competitive feasibility, editorial commitment. The 'everything is a pillar' anti-pattern: length does not make a pillar; the architecture does.

  4. 04

    Cluster planning

    Cluster size sweet spot 10 to 12. Facet sourcing from keyword research, PAA, competitor analysis, first-party customer questions. Heterogeneity check across angles (definition, how-to, comparison, examples, mistakes, costs, alternatives). No facet duplication.

  5. 05

    Internal linking architecture

    Top-down (pillar to cluster), bottom-up (cluster to pillar twice), lateral (cluster to cluster, selective). Anchor text discipline. The linking inventory pattern. Quarterly link audit checklist.

  6. 06

    URL structure and information architecture

    Subfolder over subdomain. /topic/cluster-piece/ pattern over /blog/cluster-piece/. Slug conventions. Breadcrumb structure. The /blog/ trap and when migration is worth it.

  7. 07

    Pillar page anatomy

    Hero, TL;DR (150 to 250 words, citation-ready), comprehensive body with cluster callouts woven in, use cases, FAQ with FAQPage schema, closing with cluster next-step. Schema choices (Article, HowTo). Length 3,000 to 5,000 calibrated to SERP.

  8. 08

    Cluster piece anatomy

    Focused hero, pillar callout in first 200 words, focused body, selective lateral callouts, closing that links back to pillar or to next-cluster journey step. Length 800 to 2,000 calibrated to cluster keyword SERP.

  9. 09

    Topical authority signals

    SEO signals (link graph density, entity coverage, URL clarity, breadcrumbs, schema, E-E-A-T). AEO/GEO signals (answer paragraphs, TL;DR, FAQPage schema, named statistics, named experts, distinctive POV). Two-engine optimization, complementary not competing.

  10. 10

    Maintenance and refresh patterns

    Annual pillar refresh checklist. Triggered cluster refresh on five signals (keyword decay, sub-topic emergence, link decay, AI citation drop, customer feedback). Cluster expansion and pruning. Hub lifecycle (5 to 7 year horizon).

  11. 11

    Common failure modes

    Eleven-plus patterns: 80 blog posts no rankings, 8000-word pillar that does not rank, half-built hub, no bottom-up links, self-cannibalization, /blog/ URL trap, facet homogeneity, no maintenance ownership, missing AEO/GEO signals, lifecycle ownership gap, over-built 25-cluster chaos.

  12. 12

    The framework: 12 considerations

    Topic earns the pillar, cluster size targets 10 to 12, facet heterogeneity, no duplication, three-direction link graph, anchor discipline, hub URL structure, TL;DR and FAQ for AI, cluster anatomy linking up, two-engine optimization, refresh cadence, durable ownership.

  13. 13

    Architecture is the moat

    Comprehensive content can be replicated; authoritative sources can be cited; individual pieces can be matched. What is harder to replicate is the compounding of an intentional hub built and maintained over years. Architecture is the only part of content marketing that compounds reliably.

Reference files

Ten references that go alongside the SKILL.md.

The references hold the decision tree, the topic-selection framework, the cluster-planning patterns, the linking architecture, the URL conventions, the page anatomies, the topical authority signals, the refresh patterns, and the failure-mode catalog. Hub architecture has many sub-disciplines; the references reflect the breadth.

  • references/pillar-cluster-decision.md

    Five-question decision tree for whether a piece is pillar, cluster, or standalone. Worked examples for B2B SaaS, ecommerce, services. The 'everything is a pillar' anti-pattern with diagnostic question.

  • references/topic-selection-criteria.md

    Five-criterion framework expanded with worked examples across B2B SaaS, ecommerce, services, and media. The long-tail variant pivot when competitive feasibility fails. When to walk away from a pillar idea.

  • references/cluster-planning-patterns.md

    Five facet-sourcing methods (keyword expansion, PAA, competitor outline mining, first-party customer questions, Frase/AirOps SERP analysis). Three sequencing patterns. Heterogeneity check across 12 angle categories. No-duplication discipline.

  • references/internal-linking-architecture.md

    Three link directions in detail: top-down, bottom-up, lateral. Anchor text discipline with examples. Linking inventory pattern (spreadsheet, Notion, dbt, CMS). 10-step quarterly audit checklist. When the link graph is too dense.

  • references/url-structure-patterns.md

    Subfolder vs subdomain decision. Hub URL pattern. Slug conventions (short, descriptive, keyword-aware, date-free, hyphens, lowercase). Breadcrumb schema example. The /blog/ trap and when migration pays back. Multilingual patterns.

  • references/pillar-page-anatomy.md

    Section-by-section anatomy. TL;DR placement and length (150 to 250 words). FAQPage schema with example. Internal-link callout placement within body. Article vs HowTo schema. Length calibration via SERP analysis. Updates and freshness signals.

  • references/cluster-piece-anatomy.md

    Focused-question framing. Pillar callout placement (first 200 words and closing). Lateral linking judgment. Featured snippet bait at H2 level. Schema markup. The 2,500-word and 600-word traps. When clusters work as standalones.

  • references/topical-authority-signals.md

    Why hubs produce stronger signals than orphans. SEO signals (PageRank flow, entity coverage, schema, E-E-A-T). AEO and GEO signals (answer paragraphs, TL;DR, FAQPage schema, named stats, named experts, POV). Two-engine optimization framing. Hub-level metrics worth tracking.

  • references/content-refresh-patterns.md

    8-step annual pillar refresh checklist. Five triggered cluster refresh signals. Cluster expansion criteria with 20-cluster cap. Cluster pruning when to and how to. Hub lifecycle (5 to 7 year horizon). Maintenance ownership patterns and the 'set and forget' failure.

  • references/common-pillar-failures.md

    11+ failure patterns with diagnoses and fixes. Cross-references to other reference files. Pattern across most failures: architecture is half-built, owned by nobody durable, or treats AEO/GEO and SEO as competing optimizations.

Browse all reference files on GitHub

Pairs with these platforms

Five content and SEO platforms for hub work.

The skill is the architecture discipline; the integration pages are the platform-specific tactics. Frase runs entity gap analysis at hub level. AirOps wraps hub builds in a managed workflow surface. Webflow and Contentful are publishing destinations where pillar and cluster pieces land with structured CMS support. Profound tracks AI search visibility for the hub: brand mention rate, citation share by engine, top-cited pieces.

Notion is intentionally not cross-linked here. Its existing microsite under the workflow category covers creative-direction briefs (the project-brief sibling) specifically; it does not address hub architecture. For teams running hub briefs as Notion database rows, the pattern still works: one row per piece, schema fields for pillar-or-cluster, parent pillar reference, facet category. The Notion MCP CRUDs that database; this skill designs what goes in the rows.

Bridges to sister and adjacent skills

Five sister skills compose into the content discipline.

This skill does not stand alone. Five skills connect to it: program companion above, research input upstream, per-piece companion below, execution companion further below, audit-side companion that catches what slipped. The architecture is dense because the value is heavily compositional.

  • Program companion (above)

    content-strategy

    Decides what topic areas to invest in across a quarter or year. Pillars, calendar, governance, formats. This skill takes one chosen topic from content-strategy and designs the hub.

  • Upstream research input

    seo-keyword

    Keyword research that surfaces candidate pillar topics with volume, difficulty, and intent. The cluster-facet sourcing in this skill draws heavily on keyword research output for long-tail variations of the pillar keyword.

  • Per-piece companion (below)

    content-brief-authoring

    Briefs each piece within the hub: pillar gets briefed first, then each cluster piece. The 12 fields apply to both pillar and cluster briefs; the weight of each field shifts by content type.

  • Execution scope

    content-and-copy

    Writes each piece. The hub provides the architecture; the briefs provide the contracts; content-and-copy provides the writing. Execution against the brief, calibrated to brand voice.

  • Audit-side companion

    seo-content-gap-audit

    Catches missing pieces in an existing content set after the fact, using Ahrefs MCP data. This skill prevents gaps in newly-built hubs; that skill catches the ones that emerge in legacy content sets.

Where this skill fits in the content suite

The second of five in the content suite.

Pillar content architecture is the second skill in a planned five-skill content suite. Skill 1 (content-brief-authoring) covers per-piece briefs. This skill covers hub architecture across the brief and the writing. Subsequent skills in the suite cover programmatic SEO (templated content at scale), editorial QA (review and publish discipline), and AI content collaboration (running an editorial team where some writers are agents).

Each suite skill assumes the brief discipline from skill 1 and the architectural discipline from this skill. Programmatic SEO briefs templates that generate hundreds of pieces from one architectural pattern; editorial QA reviews drafts against the brief and the architecture; AI content collaboration runs the brief-to-writer handoff at the team level when some writers are agents working inside the architecture.

The suite ships sequentially. Skills 3 to 5 do not exist yet; cross-linking lands when the skill pages ship. For now, the architecture discipline anchors the program and the five content and SEO platforms above cover the platform-specific tactics underneath.

Open source under MIT

Read the SKILL.md on GitHub.

The skill source lives in the rampstackco/claude-skills repository alongside dozens of other skills covering the full lifecycle of brand and product work. MIT licensed.

View SKILL.mdBrowse the full catalog

Frequently asked questions.

What is the difference between a pillar, a cluster, and an orphan?
A pillar is a comprehensive 2,500 to 5,000 word piece anchoring a topic; it receives links from cluster pieces and links out to clusters selectively. A cluster is a narrower 800 to 2,000 word piece covering one facet of the pillar topic; it links up to the pillar twice (early and in closing) and may link sideways to sibling clusters. An orphan is a piece that does not connect to a pillar or cluster; some orphans are fine by design (release notes, customer stories), but most are accidents that fail to compound topical authority.
How is this different from content-strategy and content-brief-authoring?
content-strategy is program-scope: multiple topic areas, editorial calendar, governance across a quarter or year. content-brief-authoring is per-piece scope: brief for one content artifact. This skill is hub-scope: one pillar plus its cluster, the architecture that ties them together. The clean reading order: content-strategy decides which topics to invest in, this skill designs each chosen topic's hub structure, content-brief-authoring briefs each piece within the hub, content-and-copy writes each piece.
How big should a cluster be?
Sweet spot is 10 to 12 cluster pieces. Below 5 reads thin; above 25 becomes unmanageable. A hub at year 1 might launch with 8 to 10 clusters and expand to 14 to 18 by year 3 as new facets emerge. A planned cluster of 25 plus pieces is usually over-built; either trim to the 12 highest-impact facets or split into two pillars under a parent topic. The 12-cluster band is empirical across most healthy hubs.
What does the link graph actually look like?
Three directions, all required. Top-down: the pillar links out to each cluster contextually within the body, not from a related-reading footer. Bottom-up: every cluster links up to the pillar at least twice, in the first 200 words and in the closing. Lateral: selective cluster-to-cluster links when one cluster naturally references another, typically 0 to 3 lateral links per piece. The bottom-up direction is the discipline that makes the pillar compound; without it the pillar is just a long article.
How does AEO/GEO factor into hub architecture?
AI engines (ChatGPT, Perplexity, Claude, Gemini, Google AI Mode) reward the same hub signals as search engines plus a few additional ones: TL;DR sections that get cited verbatim, 40 to 60 word answer paragraphs immediately following H2 headings, FAQPage schema, specific statistics with named sources, named experts and methods, and distinctive POV. Optimizing for AEO/GEO does not trade off against SEO; the hub architecture that earns search rankings is the same one that earns AI citations. Both reward depth, structure, and entity coverage.
How often should we refresh a pillar?
Annual pillar refresh on a fixed cadence: re-validate the SERP, update statistics, add sections for emerged facets, prune sections no longer relevant, refresh internal-link callouts to clusters since the cluster has likely grown, update the TL;DR, bump the last-updated date, re-submit to search engines. Cluster refresh is triggered, not scheduled: refresh when keyword performance drops, a new sub-topic emerges, links go stale, AI citations decline, or customer feedback flags a gap. Hubs themselves have a 5 to 7 year lifecycle; after that, wholesale restructuring is often appropriate.