A topic cluster can become a very organised waste of time.

A cluster earns its place when the supporting pages answer different questions. Ten articles repeating a broad topic create maintenance, not depth. Map the decisions a buyer needs to make before choosing which pages to commission.

Map the disagreement, not only the keyword

The person researching the solution may care about time saved, while the colleague approving it cares about cost and ownership. A useful cluster can address both without commissioning two versions of the same introduction. Write the buying question and intended reader against every page. Then identify the page that resolves their disagreement: an implementation guide, a cost explanation or a comparison of operating choices. Give that page a route from the relevant commercial page. A bigger cluster is not the objective; fewer unanswered questions in the evaluation is.

Build around a buying decision

A cluster becomes commercially interesting at the point where a question changes. The reader moves from recognising a problem to comparing approaches, preparing implementation or choosing an offer. That transition deserves a page or a link. Another article explaining the original problem may not. The plan should expose these transitions rather than merely group similar phrases. Its value comes from the route it creates, including the questions that persuade a suitable buyer to keep investigating.

Start with the commercial problem

Suppose you sell inventory software. “Business efficiency” is too broad to guide an editorial programme. “Prevent stock discrepancies across multiple warehouses” gives you a more useful starting point.

That problem can generate distinct questions: how discrepancies arise, how teams reconcile them, which workflows software should support, what implementation requires and when a spreadsheet remains enough. Each page earns a place by helping the reader make progress.

Google’s link guidance recommends descriptive anchors and crawlable links. The cluster strategy adds an editorial choice: connect pages because the reader needs the next answer, not merely because their keywords overlap.

Give each page a clear job

Write one sentence describing what a reader should know or decide after each article. If two pages do the same job, consider combining them. If a page has no obvious link to your expertise, challenge why it belongs in the plan.

The cluster should include enough depth to serve the problem, not an arbitrary article count. Sometimes five strong pages are enough. Sometimes a complex category needs more. Scope follows the questions, evidence and available resources.

The plan becomes executable when each page has a complete content brief. Once the pages exist, link them around the reader’s next question; a cluster diagram alone does not create that journey.

Make the plan useful before you publish

The exact path is not a rule for every visitor. It is a way to ensure the pages support a coherent journey instead of becoming an isolated blog archive.

Build in waves

Publish the central problem guide and the most commercially useful supporting page first. Add links from relevant existing pages. Review search queries, engagement and enquiry quality before extending the plan.

If the early content attracts the wrong audience, revise the topic choice rather than filling the remaining slots. If readers consistently ask a question you missed, that is a stronger reason for a new article than a keyword tool’s suggestion alone.

Know where clusters go wrong

Teams often produce near-duplicate articles for tiny keyword variations. They also write broad introductions because those are easier than useful implementation advice. The resulting cluster looks complete but sends a buyer elsewhere for the answer that matters.

My test is whether the collection helps a reader understand the problem, evaluate alternatives and choose a next step. If it only increases the number of indexed URLs, it has confused production with progress.

Choose the cluster that has a destination

Take a hypothetical inventory SaaS business choosing between a broad “business productivity” cluster and a narrower “multi-location stock control” cluster. The first has more apparent reach. The second connects to a problem the product actually solves.

These pages form a cluster because they help the same buyer progress, not because they repeat the same keyword. Publish the commercial destination and the most urgent supporting question before commissioning ten awareness articles.

Use a spreadsheet with page, primary question, intended reader, unique contribution and next action. Two briefs with the same question and destination need a reason to coexist. Different search wording alone is a weak reason.

Broaden coverage when evidence supports the audience and its route into the offer. Until then, give the cluster explicit boundaries.

Find the page you are accidentally commissioning twice

Write every proposed page as a sentence beginning “After reading this, the buyer can…” Put the intended audience and next destination beside it. Compare those three fields before approving production.

If two briefs serve the same person, enable the same decision and lead to the same destination, explain what makes the second necessary. A synonym in the keyword column is not enough. Merge the briefs when the underlying job is the same.

Now trace one route from an introductory question to a commercial decision. A missing link may require an implementation guide or comparison, rather than another awareness article. Publish that missing piece first. The completed map should show a useful sequence, not simply a large inventory of subjects.

Use sales and support to find the missing middle

Search tools can help reveal demand, but they often do not capture the implementation question that determines whether a buyer moves forward. Ask sales which concerns delay a decision. Ask customer success what new users misunderstood. Ask technical teams what prospects need to prepare before a project can begin.

Those conversations often identify the missing middle between an introductory guide and a service page. Content about prerequisites, ownership and risk may have less obvious search volume, yet help a good-fit buyer make progress. Publish it when it has a clear role, and link it from pages where the question naturally arises.

Document repeated questions in a shared sheet. Include the source of the question, the audience, the decision it affects and whether an existing page answers it. This prevents every request from becoming a new article while making genuine gaps visible.

Build a cluster plan you can resource

Estimate effort separately from potential value. A research-heavy comparison may require more review than a practical guide based on an established process. Do not assign the same production deadline to every page and then blame the writer when the difficult one lacks depth.

Decide which page should be the centre

A central resource does not have to contain everything. It should orient the reader, answer the overarching question and help them choose the relevant detail. If it becomes an enormous page that repeats every supporting article, it may be harder to use and maintain.

Choose a format appropriate to the topic. A problem guide can work when the cluster begins with diagnosis. A resource hub can work when visitors need several distinct workflows. A commercial category page can connect use cases and products. The structure should follow the reader’s task rather than a diagram used in a presentation.

Link at the point where the question changes

Link from “why this happens” to “how to diagnose it” when the reader is ready to investigate. Link from a comparison to implementation requirements when they are evaluating feasibility. Link from a practical guide to the relevant service when the need for help becomes clear. These transitions explain why the destination is useful.

Avoid placing every cluster link in every article. That can turn a focused explanation into a directory and make the important choice less obvious. A small group of relevant links, supported by clear navigation, is usually easier for a reader to use than a repeated wall of related terms.

The missing page may be an objection, not a topic

A cluster can cover the category thoroughly and still avoid the question that stops a buyer: migration effort, implementation responsibility or what happens when the chosen approach fails. Keyword similarity will not reliably reveal that gap.

Ask sales for a recurring objection with a concrete buying context. Identify the page where the concern should be answered. Sometimes a section added to an existing commercial page is enough; a separate article would make the answer harder to find.

Build a new page when the question requires substantial explanation and serves a distinct decision. Keep the boundary explicit. Coverage should grow because the buyer has another useful job to do, not because the cluster diagram has an empty branch.

Sources and further reading

Google Search Central: link best practices: Crawlable links and descriptive anchor text.

Google Search Central: SEO Starter Guide: Foundational discovery and content guidance.

References checked 9 October 2026. Sources support the documented facts and tool behaviour; the recommendations and worked examples explain how I would apply them.