Topic clusters in 2026: what still works, what doesn't

Topic clusters were the defining SEO play of the last decade: choose a pillar theme, publish a cluster of supporting articles, interlink them, and watch organic traffic compound. In 2026 the idea still matters — but the version that worked in 2020 now quietly loses money. Search behaviour has shifted toward specific, intent-led queries, AI-generated content has saturated the easy informational keywords, and Google's helpful-content signals have devalued shallow pillar pages assembled from competitor research alone.

What still works

• Buyer-intent clusters. Anchor every cluster to a real revenue motion. A page targeting "best CRM for healthcare clinics" out-earns a generic "what is a CRM" explainer every time, because it meets a buyer at the moment of decision.
• First-party experience. Pages that show original analysis, real screenshots, proprietary data, or a genuine point of view consistently beat interchangeable AI summaries. Demonstrable expertise is the new moat.
• Intentional internal linking. Pass authority from informational content toward commercial pages, not the other way around. Your money pages should receive links, not leak them.

What we've stopped recommending

We no longer advise shipping 1,500-word "ultimate guide" pillars built purely from what competitors already rank for. They are expensive to produce, slow to index, and increasingly invisible. In their place we build answer-first pages — around 600 focused words that resolve the query immediately — supported by a longer, genuinely useful asset such as a template or calculator for buyers who want depth.

How to audit your own clusters

Start with revenue, not keywords. List the queries that actually precede a sale, map which pages serve them, and prune or merge anything thin. A smaller set of pages that each earn their place will outperform a sprawling library of shallow content — in rankings, in conversions, and in the cost of maintaining it.