Topic cluster

A topic cluster is a content structure that groups one broad pillar page with several pages on related subtopics, all connected by internal links. It helps readers, search engines such as Google and Bing, and the AI assistants that rely on them see that a site covers a subject in depth, sub-question by sub-question.

Why topic clusters matter for AI search

AI engines do not answer a prompt from one search. With query fan-out, they run several related searches and cite pages that rank for any of them. In Ahrefs' March 2026 analysis of 863,000 searches, only 37.9% of the URLs cited in Google's AI Overviews also appeared in the top 10 for the query itself. Ahrefs suggests many of the others were found through fan-out sub-queries.

A topic cluster is the content answer to that shift. Instead of one page fighting for the main query, a site has a focused page for each sub-question, and each page can be retrieved on its own search.

How a topic cluster works

A cluster has three parts:

A pillar page that covers the broad topic and links to every subtopic page.
Cluster pages that each answer one subtopic in depth and link back to the pillar.
Internal links that tie them together, with anchor text that Google wants "descriptive, reasonably concise, and relevant." Its rule of thumb: "Every page you care about should have a link from at least one other page on your site."

HubSpot built the model on 2015 research by two members of its team. Before the switch, its own blog had thousands of pages on overlapping topics, competing for the same searches.

Pillar page vs cluster page

The pillar page and the cluster pages do different jobs. Mixing them up weakens both: a pillar that tries to answer everything in depth, or cluster pages that repeat the pillar.

	Pillar page	Cluster page
Scope	The whole topic, at overview level	One subtopic, in depth
Typical search	Broad: "email marketing"	Specific: "email subject line length"
Links	Out to every cluster page	Back to the pillar, and to related pages

Example: Google's lawn example as a cluster

Google's own guide gives a fan-out example. For the query "how to fix a lawn that's full of weeds", sub-queries might include "best herbicides for lawns", "remove weeds without chemicals" and "how to prevent weeds in lawn".

A garden retailer could map that to a cluster: a pillar page on fixing a weedy lawn, plus three cluster pages comparing herbicides, explaining chemical-free removal and setting out a prevention calendar. Each page can rank for one sub-query, and the links show search engines they belong together.

How to build a topic cluster

Start from the prompts your buyers ask, list the sub-questions behind each one, then group your existing posts under pillars and merge the ones that overlap. Write only the cluster pages that answer a real question; depth on one subject is what builds topical authority. Google warns that creating separate content for every variation of a search, "or fan-out queries", primarily to manipulate rankings or AI responses violates its scaled content abuse policy.

To choose which cluster to build first, look at the answers. Hawkscan's action plan turns what it reads in AI answers into actions ranked by impact and updates them continuously, so a new cluster competes with your other fixes on evidence.