How to build a topic cluster for a B2B website: seven steps, common mistakes and a mini cluster ranked by commercial value, with DataForSEO data from September 2026.
Louis Mauclair · CEO
Updated October 2026 · 11 min read
TL;DR A topic cluster is a pillar page, a set of supporting pages, and internal links that follow that hierarchy. For a B2B site, build it in seven steps: map search intent, group queries by SERP overlap, set the hierarchy, rank by commercial value, plan the links, publish in order, then measure. Start small: a cluster your team cannot finish does more harm than good. This is the logic behind how we work as a B2B SEO agency.
Most guides explain topic clusters with dog toys or email marketing. The HubSpot and Semrush guides we read contain no table and no method for ranking pages by commercial value (checked October 26, 2026). That is the question that matters when your team is small and your sales cycle is long.
What is a topic cluster?
A topic cluster is a way to organize a site's pages around one subject: a pillar page that covers it broadly, supporting pages that answer each sub-question, and internal links that follow that hierarchy. For B2B, treat it as a map of your buyers' search intent, ranked by commercial value.
It is also called a content cluster (Semrush uses both terms) or, in French SEO, a cocon sémantique. Yoast calls clusters a good option for sites that still need to build authority, and warns that they take time and are never finished.
How to build a topic cluster for B2B in 7 steps
These steps follow our method: extraction and qualification of the market's keywords, clustering into a cocoon structure, content briefs, writing and on-page optimization, internal linking that follows the structure, then rank tracking and iteration.
1. Map search intent
Start from your buyers' questions, not a keyword export: sales calls, existing pages, Search Console queries, People Also Ask boxes. Tag each query three ways: an offer, a buying stage (problem, comparison, price, decision) and a target page.
Inventory what you already have first. An SEO audit keeps you from rebuilding a published page and shows you the ones nobody links to. To turn the map into a calendar, see our approach to content strategy.
2. Group queries by SERP, not by similar words
Two queries that look alike do not always deserve the same page. Search each one and compare the URLs of the top results. If at least three match, the intent is probably the same: one page. If none match, use two pages even when the wording is close. Three is a working convention, not a Google rule.
In the example below, "managed it services pricing" and "managed it services cost" share 3 of 8 URLs, so one page covers both. "Managed it services" and "what are managed it services" look interchangeable, yet their top results share no URL: a separate page for a separate SERP.
3. Set the hierarchy: pillar, clusters, supporting pages
Pillar page: one per offer. It covers the subject end to end and links to every cluster page.
Cluster page: one page per decision to inform (pricing, company size, comparison, contract).
Supporting page: one precise question, often upstream of the purchase, linking to its cluster page and then to the pillar.
Match each page to the dominant format of its SERP. For pillar page SEO, if the results for the pillar query are mostly educational guides, a pure sales page answers the search poorly: the pillar has to teach before it sells.
4. Rank by commercial value before completeness
Score every query on commercial value first, then difficulty, then volume. In B2B, volumes are small and the queries that open a quote are smaller still: a pricing query with a few dozen monthly searches can be worth more than a general query searched far more often.
The page count is never the goal. The HubSpot guide suggests starting with three to five pillar candidates and supporting each with 20 to 30 articles, which is 60 to 150 articles by our arithmetic: for a small B2B team, a calendar you will not finish. Size the plan to your team's capacity: a small cluster that is written, reviewed and linked beats a large one that is half published. Automation can speed up briefs; it never replaces human review before publication.
5. Plan directed internal links
Draw the links before you write: each supporting page links to its cluster page, each cluster page to the pillar, and the pillar to all of its cluster pages. A lateral link is justified only when a reader would want the next page.
Google recommends that every page you care about gets a link from at least one other page on your site, with descriptive, reasonably concise, relevant anchor text (Google Search Central). The mechanics are in our internal linking guide.
6. Publish in the right order
Publish the pillar, even a short version, then the high-value cluster pages, then the supporting pages. Each page ships with its links in place: with no inbound link, it is already an orphan.
List it in the sitemap the same day. Google says a sitemap is only a hint, with no guarantee of crawling, and that you should include the URLs you want to see in results (Google Search Central). Crawl and indexation checks fall under technical SEO.
7. Measure and iterate
Track the cluster page by page: rankings, indexed pages, leads per page. Look for cannibalization (two URLs alternating on the same query), then merge, differentiate or extend.
Measurement has to work first: in our external audits (seven action plans, September 14 to October 6, 2026, without account access), 6 of 6 audited sites had broken or incomplete tracking. Without reliable measurement, you cannot tell which cluster pays.
Where should you start on your site? Get an action plan with your priority levers and the order to run them. Get your free action plan
Topic cluster example: a mini cluster for a B2B service
Illustrative example. Generic B2B service: managed IT services, in the United States. This is not a client and not a recommendation. Source: DataForSEO, September 2026 for volumes and difficulty, Google US SERPs captured October 26, 2026.
Pillar: Target query: managed it services; Searches per month: 22,200; Difficulty: 29; Commercial value (our reading) and decision: High (CPC $65.28). Mixed SERP: provider homepage, Wikipedia, vendor guides, Reddit. Publish first.
Cluster: Target query: managed it services small business; Searches per month: 1,300; Difficulty: 6; Commercial value (our reading) and decision: High (CPC $82.41): precise buyer, clear offer. No URL shared with the pillar. Next.
Cluster: Target query: managed it services pricing; Searches per month: 390; Difficulty: 12; Commercial value (our reading) and decision: High: decision stage (CPC $27.38). "Cost" (140 searches, difficulty 6) shares 3 of 8 URLs: same page. Next.
Support: Target query: what are managed it services; Searches per month: 720; Difficulty: 24; Commercial value (our reading) and decision: Medium (CPC $27.59). Definition SERP, no URL shared with the pillar: a supporting page that links up.
Support: Target query: managed it services contract; Searches per month: 170; Difficulty: 3; Commercial value (our reading) and decision: Medium to high (CPC $20.04). Distinct SERP (agreements and templates): a short checklist page after the core.
Phase 2: Target query: it outsourcing; Searches per month: 1,000; Difficulty: 19; Commercial value (our reading) and decision: Medium (CPC $74.10). Broader topic, no URL shared with the pillar: later.
Excluded: Target query: managed it services near me; Searches per month: 9,900; Difficulty: 3; Commercial value (our reading) and decision: Conditional: local intent (CPC $61.40). Needs local pages; skip it for a national cluster.
Three takeaways:
Big numbers move. Monthly volume for "managed it services" ranged from 8,100 to 90,500 over twelve months. Never size a cluster on one figure.
The biggest number is not the best page. "Near me" has 9,900 searches, but its intent is local.
Close wording is not a shared SERP. "What are managed it services", "contract" and "it outsourcing" share no URL with the pillar.
The core to publish first is three pages: the pillar, the small-business page and the pricing page. The rest waits until they are linked and measured.
Topic clusters SEO and AI search: what the sources support
On the link between clusters and answer engines, little is proven. Here is what the sources say.
Google: AI Overviews and AI Mode may issue multiple related searches across subtopics and data sources, a technique Google calls "query fan-out." To appear as a supporting link, a page must be indexed and eligible to be shown with a snippet. Google writes that no special optimizations are necessary (documentation updated December 10, 2025).
OpenAI: OAI-SearchBot is used to surface websites in ChatGPT's search features, and sites that opt out of it are not shown in ChatGPT search answers (OpenAI documentation). Check your robots.txt.
Semrush claims that topic clusters answer the sub-queries AI systems generate and can increase citations, but cites no independent study (Semrush). Treat it as a hypothesis.
Our reading, which is not a measurement: a cluster plans for what these systems look for, one clear, indexable page per sub-question. No open source we consulted measures its effect on AI citations. The levers specific to GEO (crawlers allowed, quotable answers, external mentions) are in our guide on how to rank in ChatGPT.
Common topic cluster mistakes
A cluster too ambitious for the team
The plan covers the whole market, but the team publishes at its own pace. The result: one pillar live, cluster pages half written, links to pages that do not exist. Size the cluster to the team's capacity, then expand.
Orphan content
A page with no internal link is hard to discover and passes nothing to its neighbors. It can exist in the CMS without existing for Google.
Cannibalization
Two pages that target the same SERP fight for the same spot, and neither wins. The shared-URL test (step 2) catches it before you write.
Pages with no commercial intent
A supporting article that answers a question your buyers do not ask, or that leads to no cluster page, brings visits with no consequence. Every supporting page must point to its next page: the cluster page, then the offer.
Method: sources, criteria and dates
Volumes: DataForSEO, Google Ads search volume endpoint, United States, monthly average over 12 months to September 2026, queried October 26, 2026. Google Ads sometimes groups close variants; a missing volume means fewer than 10 searches or no data.
Difficulty and CPC: DataForSEO Labs keyword difficulty (0 to 100). CPC is the average in US dollars returned by DataForSEO: what advertisers pay, not what a lead is worth to you.
SERPs: top organic results from Google US captured October 26, 2026 (8 or 9 depending on the query), compared by exact URL.
Commercial value: an editorial reading (buying stage, SERP format, CPC), not a DataForSEO field. One capture on one date: SERPs and volumes move.
Conclusion
The B2B topic cluster that works is small, ordered and finished. Our view: start from commercial value, not volume; group by SERP, not by similar words; begin with a pillar and the cluster pages that open a quote; link everything the day you publish; measure, then extend the topic cluster at your team's pace.
Which cluster pages should you publish first? Get an action plan with your priority levers and the order to run them. Get your free action plan
Frequently asked questions
What is a topic cluster?
It is a group of pages organized around one subject: a pillar page, supporting pages and internal links that follow that hierarchy. In B2B, we use it as a map of buyers' search intent ranked by commercial value. It structures coverage; it does not guarantee a ranking.
What is a pillar page in SEO?
The pillar page is the main page of a cluster: it covers the subject broadly and links to every cluster page. If the top results are educational guides, the pillar has to teach before it sells.
Is a topic cluster the same as an SEO silo?
People use the terms loosely, and we did not open a source that settles the difference. In our view, a silo keeps links inside a strict category, while a cluster allows lateral links where a reader would want them. What matters more is whether the structure follows SERPs and commercial value.
Do topic clusters work for a new website?
Yoast describes clusters as a good option for sites still building authority, with the caveat that they take time and effort. For a new B2B site, build a small cluster around commercial queries and measure it before adding more.
Do topic clusters help you get cited in AI answers?
No open source we read measures it. Google says its AI features may issue multiple related searches across subtopics, and that a page must be indexed and snippet-eligible to appear as a link. A well-planned topic cluster puts you in those conditions, without any promise of a citation.
Louis leads growth and digital acquisition for B2B and e-commerce companies in France and the UAE. Before Klickbee, he co-built and sold an e-commerce business that served more than 13,000 customers. He moved to Dubai in 2026 to work with companies across the GCC.
GEO builds on SEO, it does not replace it: Google says optimizing for its AI features is still SEO. What is documented, what is hypothesis, and a 30-day plan for B2B teams.
Indexing, facets, product pages, redirects, reviews, tracking: 24 ecommerce SEO audit points, each with a priority level and the mistakes we see in the field.
A straight verdict on llms.txt for B2B sites: what is proven, what to fix first (robots.txt, rendering, entity pages), how to measure it in server logs and a minimal template.
Louis Mauclair·12 min read
Free action plan
Grow faster. Get your free action plan.
A tailored plan to generate more leads and sales from your existing traffic.