Build a Topic Cluster in Seven Steps
Topic clusters are not a content calendar shape. They are an internal linking system that teaches crawlers and readers which URL owns the parent intent. This playbook runs seven steps: pick the parent query, audit coverage, define the hub URL, map spoke intents, draft with Content Operations, relink sitewide, measure query movement in Search Console. Same hub, stronger spokes, no orphan publishing.
Clusters are architecture, not a folder name
Most teams discover they have a topic cluster when someone draws circles on a whiteboard. Crawlers discover clusters when one hub URL earns parent intent and every spoke links up with descriptive anchors. Without that architecture, related blog posts compete for the same queries and equity leaks sideways.
The failure mode is familiar. You publish twelve posts on a theme. Impressions rise. Clicks flatline. Cannibalization reports show three URLs on the same head term. None of them rank on page one because none of them behave like a hub. This playbook fixes that pattern before you ship spoke number thirteen.
Operator rule
One parent query, one hub URL, many spokes that link up before you publish the next spoke.
Topical Authority in 2026 explains why entity depth beats publish volume. Internal Linking Growth Playbook covers sitewide redistribution after the cluster ships. Create Pillar Page in Seven Steps goes deep on hub craft when you already know the cluster map. Read those when you need strategy context. This article is the build sequence.
Start with Search Console query groups, not brainstorming. Parent intent shows up as overlapping impressions on related terms landing on different URLs. That overlap is your cluster brief waiting to happen.
Step 1: Choose the parent query and hub URL
Export GSC queries for the theme over ninety days. Group by shared head terms and SERP overlap. Pick the parent query with enough impression volume to justify a hub and enough commercial proximity to matter to your business. A cluster on a theme you never monetize is an academic exercise.
- Confirm the parent query is not purely brand navigational.
- Check cannibalization before you commit the hub slug.
- Pick a URL pattern you will keep for years.
- Log the parent query cluster in your backlog with success queries attached.
- Screenshot SERP leaders so format decisions survive stakeholder churn.
Striking Distance Keywords often reveal parent and spoke candidates in the same export. Do not promote every modifier to a new hub. One parent, one keeper URL, many spokes.
When an existing URL already earns impressions on the parent cluster, refresh it into the hub. Consolidation beats net-new when equity exists. Keyword Cannibalization Workflow resolves keeper decisions when two URLs split the parent job.
User Intent SEO Guide classifies whether the parent query wants a guide, comparison, or product surface. Wrong format on the hub poisons every spoke you attach later.
Step 2: Audit existing coverage
Crawl your site and URL Library. List every URL that already ranks or could rank on cluster terms. Mark keepers, merge candidates, and gaps. Most teams skip this step and publish spoke seven while spoke two still cannibalizes the hub.
Coverage audit rows
- Keeper hub
- Spoke keepers
- Merge pairs
- True gaps
- Retire candidates
One URL owns parent intent. Refresh before expanding spoke count.
URLs that match narrow intent and need relink only, not net-new siblings.
Two URLs on same intent. Consolidate before new publish splits equity again.
Intent with no credible URL. Candidate spoke with distinct SERP proof.
Thin URLs that add noise. Redirect or noindex after hub ships.
AI Content Audit helps at portfolio scale. For one cluster, a spreadsheet with URL, primary cluster, impressions, and recommended action is enough. Attach the audit to the Growth Order so reviewers see scope.
Content Gap Analysis Growth Orders ranks cluster work against decay on money URLs. A cold cluster waits until keeper refreshes on revenue pages clear the queue.
Step 3: Map spoke intents
Each spoke answers one narrow question with a distinct SERP. Cap initial spokes at five to eight. Depth beats fifty thin pages that differ only by adjective. Programmatic SEO Complete Guide covers template scale later. Editorial clusters win on quality first.
Spoke quality test
Ship
- Distinct SERP with different dominant format or scope
- Clear link-up to hub in first third of body
- Original operator examples competitors lack
- Success queries logged before draft starts
Skip
- Same intent as hub with longer word count
- Duplicate modifier only, no new job to be done
- No internal link plan before publish
- Spoke exists because keyword tool exported a row
Mine People Also Ask, support tickets, sales call notes, and community questions. Spokes should map to questions your team already answers verbally. That is information gain without inventing demand.
Ten Blog Post Templates by Intent gives format starting points once intent labels are set. Templates are scaffolding, not substitutes for SERP review.
Step 4: Draft the hub first
The hub summarizes the entity, links to every spoke, and earns external references. Route to Content Operations with parent query, spoke list, Knowledge Base context, and link placeholders. Human review is mandatory. Learn Domains does not auto-publish.
Create Pillar Page in Seven Steps is the companion playbook for hub-only depth: H2 map, schema pass, refresh rhythm. Build the hub to that standard before you batch spokes.
- •Brief hub with parent query and spoke map attached.
- •Draft body with link placeholders for each planned spoke.
- •Add FAQ block for follow-up questions visible on page.
- •Approve and publish hub before spoke batch two when possible.
- •Submit hub URL in Search Console after meaningful body ship.
Hub first is not dogma when spokes already exist. If legacy URLs cover subtopics, refresh the hub same sprint you add link-down modules and spoke link-ups. Partial clusters confuse crawlers and readers equally.
SEO Content Workflow defines handoffs between brief, draft, review, and external publish record. Clusters should never skip review because they feel strategic.
Step 5: Ship spokes with link-up discipline
Every spoke includes a contextual link to the hub in the first third of the body. Update the hub to link down to each new spoke the same week. Link-up discipline is non-negotiable. Spokes without hub links are orphan pages wearing cluster branding.
Link-up discipline
If a spoke publishes without a hub link, fix it before the next spoke ships.
Sideways links between related spokes help readers and crawlers traverse the ring. Keep anchors descriptive. Generic learn more links waste the graph you are building.
GSC Data to Content Tasks turns query evidence into writer-ready spoke briefs. Attach primary cluster and keeper URL to every brief header.
Batch spoke publishing weekly, not daily, so relink passes stay finishable. Velocity without architecture is how clusters collapse into keyword spam.
Step 6: Relink sitewide
Audit nav, footer, related posts, and category pages. Point relevant traffic at the hub and priority spokes. Internal Link Audit checklist validates crawl paths. Orphan hubs with strong spokes still underperform.
- Add hub link from top ten traffic pages where context fits.
- Replace generic related-post widgets with cluster-aware modules when possible.
- Fix inbound anchors that say click here or read more.
- Submit updated sitemap after bulk relink pass.
- Log relink owner same as cluster owner so nothing stalls.
Internal Linking Growth Playbook covers redistribution patterns for multi-cluster sites. One relink sprint per cluster beats endless ad hoc edits.
Commercial pages should link to informational hub when the reader's next job is education, not hard pitch on paragraph two. Intent paths matter inside clusters too.
Step 7: Measure and refresh
After four weeks, compare hub and spoke query clicks to baseline. Refresh the hub when parent queries stall. Refresh spokes when narrow clusters decay. Content Decay Recovery Playbook runs the same loop for declining URLs inside the cluster.
Measurement discipline
One cluster, one parent query, one monthly review note. Scatter metrics prevent learning what worked.
Track cluster health as prefix segments in GSC when URL patterns allow. Impressions without clicks on the hub often mean CTR or format issues, not missing spokes.
GA4 GSC Combined Workflow validates engagement when clicks move. Asset Yield weights commercial clusters higher when revenue connects.
Quarterly hub refresh beats annual rewrites. SERP leaders update subtopics. Your hub should too.
Common cluster mistakes
Mistake one: hub as table of contents only. A hub that lists links without answering the parent job will not earn parent queries. Mistake two: spoke sprawl. Thirty thin URLs beat eight strong ones only in fantasy dashboards. Mistake three: publishing without relink. New spokes hidden from high-traffic entry points might as well not exist.
Mistake four: ignoring cannibalization during build. Every merge deferred becomes a redirect project later. Mistake five: measuring publish count instead of query movement. Clusters exist to move clicks on defined query groups, not to fill content calendars.
- Name parent query cluster before any draft starts.
- Cap initial spokes and finish relink before scaling.
- Refresh hub when parent clicks stall, not only when spokes decay.
- Document keeper URL decisions in the backlog for the next operator.
Weekly cluster standup rhythm
Week one: audit and hub draft. Week two: hub publish and first three spokes. Week three: remaining spokes and relink pass. Week four: GSC baseline and refresh orders. Solo operators run one cluster per month. Agencies cap one active cluster build per client to protect review capacity.
Mission Brief ICEE ranks cluster work against decay and striking-distance orders. High Impact parent queries on commercial themes beat low Impact spoke ideas most weeks.
Choosing cluster themes with commercial weight
Not every keyword theme deserves a cluster. Pick themes where parent intent connects to revenue path: product category you sell, problem your tool solves, audience you retain. Traffic-only clusters inflate hosting and review cost without asset yield.
Asset Yield Framework weights commercial proximity in prioritization. Cluster build on theme with no conversion path waits behind money URL decay.
Five Strategies Choose SEO Blog Topics helps reject themes that look interesting in keyword tools but fail commercial weight test.
Spoke fulfillment through Content Operations
Each spoke brief attaches to Content Operations with hub URL, intent label, information gain note, and link-up requirement. Batch spoke drafts only after brief batch approved.
QA rejects spokes that duplicate hub H2 bodies. Spoke should go deep on narrow job hub summarizes.
Writing SEO Friendly Content Checklist habits apply per spoke same as hub. Cluster quality is limited by weakest spoke.
Cluster measurement beyond clicks
Measure hub parent cluster clicks, spoke cluster clicks, assisted conversions when GA4 paths exist, and internal link click-through from hub to commercial pages.
Impression growth without hub click growth may mean title or format fix, not more spokes.
Recover Organic Traffic Without Publishing More reinforces refresh before expand when hub underperforms despite new spokes.
Cluster build checklist
- •Parent query cluster named with success queries.
- •Keeper hub URL confirmed after cannibal audit.
- •Spoke map with distinct SERP proof per spoke.
- •Hub draft approved with spoke link placeholders.
- •Link-up rule enforced on every spoke publish.
- •Sitewide relink pass scheduled same sprint as batch.
- •GSC baseline logged for hub and priority spokes.
- •Monthly review owner assigned in backlog.
Checklist lives in Growth Order or project ticket, not oral tradition. Next operator continues cluster without rediscovering architecture.
Topical Authority in 2026 treats completed cluster as entity subgraph milestone. Report cluster ship as graph improvement: fewer orphan URLs on theme, denser internal paths, measurable hub click trend.
Solo founders defer spoke expansion when hub parent queries have not moved after first relink pass. Fix hub depth and entry links before adding spokes that inherit weak architecture.
Agencies document cluster map in client wiki with hub URL, spoke list, and relink status so account transitions do not reset architecture memory.
When cluster spans programmatic and editorial spokes, hub link module in template must match editorial link-up rules so graph stays coherent for crawlers.
Document cluster ship date and baseline queries in shared backlog before starting next cluster. Unmeasured clusters cannot prove which architecture choices worked.
Internal Linking Tools compare software for relink execution. Tools assist crawl of gaps; this playbook defines which links to add after cluster ship.
Recover Organic Traffic Without Publishing More is the doctrine when cluster impressions rise but hub clicks stall: refresh and relink before publishing spoke fourteen.
Ship one cluster completely before starting the next. Parallel cluster builds split relink capacity and review focus.
Frequently asked questions
- What is a topic cluster?
- A hub URL plus linked spokes that together cover a parent intent with clear internal link architecture.
- How many spokes should I launch first?
- Five to eight strong spokes beat thirty thin pages. Expand after measurement proves hub strength.
- Should I pick an existing URL as the hub?
- Yes when equity and intent match. Refresh before creating a competing URL on the same cluster.
- How do I avoid cannibalization inside a cluster?
- One intent per URL, merge duplicates, relink legacy posts to the hub, and reject spokes that duplicate the parent job.
- When should the hub publish relative to spokes?
- Hub first when possible. If spokes exist, update hub and bidirectional links same sprint.
- Does Learn Domains build clusters automatically?
- It detects gaps, drafts, and suggests links. Humans approve and publish externally.
- How long until cluster results show?
- Often six to twelve weeks for query movement on parent and spoke clusters. Judge against baselines, not site-wide vanity.