Topic Clusters for Small Sites: How to Plan 30 Articles That Rank Together
Thirty articles built around one pillar page and four or five tightly linked subtopic groups will usually serve a small site better than thirty loosely related posts. The structure shows search engines and AI answer engines that you cover one subject in depth, not many subjects at a glance. No structure can promise a ranking, but it gives your content the best chance of being read as coherent.
Why 30 scattered posts rarely add up to authority
Most small business blogs grow the same way. Someone has an idea, a post goes up, and a few weeks later another one appears on a completely different subject. After a year there are thirty posts and traffic is still flat.
A topic cluster is a set of articles on one subject, organized around a central pillar page that gives the overview, with supporting articles that each go deep on one question. The mechanism worth caring about is topical focus, not post count. Ten articles that answer the questions around one subject reinforce each other. Ten unrelated articles each have to stand alone.
Here is what we have and have not verified. Google has said repeatedly that it wants to reward helpful, people-first content, and its guidance on creating helpful content says to ask whether your site has a primary purpose or focus. That is guidance, not a ranking formula. Nobody outside the search engines can promise an outcome from structure alone, and we won't.
Start with one core topic your business can credibly own
Your pillar subject sits where three things overlap: what you genuinely know, what your customers keep asking, and what people actually search for. Miss the first and the articles will read thin. Miss the second and they won't convert. Miss the third and nobody will find them.
A simple scope test helps. If you can't imagine finishing the topic in thirty articles, it's too broad ("home improvement"). If you run out of distinct questions at article twelve, it's too narrow ("replacing one model of boiler valve"). "Boiler servicing and repair for homeowners" is the right kind of size.
Pick one cluster before attempting two. A small team that spreads thirty articles across two subjects ends up with two half-built clusters, which is the scattered-posts problem in a tidier costume.
Map the cluster: one pillar, four or five subtopic groups, 30 articles
Break the pillar into four or five subtopic groups and assign roughly five to seven articles to each. Within each group, mix intent types: definitions, how-tos, comparisons, cost questions, troubleshooting, and buying decisions.
Here is an illustrative example, not a real case study. Imagine a local heating company with the pillar "Boiler servicing and repair." Its groups might be:
- Servicing and maintenance (what a service includes, how often, what it costs)
- Faults and troubleshooting (pressure loss, strange noises, no hot water)
- Repair versus replace (age thresholds, cost comparisons)
- Choosing a new boiler (types, sizing, brands compared)
- Safety and compliance (carbon monoxide, certification, landlord duties)
To find the actual questions, use real data before intuition. Search Console shows queries you already appear for, People Also Ask boxes show how searchers phrase follow-ups, and customer emails and sales calls show what people ask right before they buy. It is also worth asking ChatGPT, Perplexity and Google's AI Overviews the same questions and noting which sources they cite. That tells you what format and depth currently earns citations.
Link the cluster so it works as a system
Internal links are how the cluster becomes a structure instead of a folder of posts. The mechanics are simple:
- Every article links up to the pillar.
- Every article links sideways to two or three closely related siblings.
- The pillar links down to every article in the cluster.
- Anchor text describes the destination ("boiler pressure troubleshooting," not "click here").
The common mistakes are predictable. Orphan posts have no links pointing at them, so crawlers and readers rarely find them. Articles that link to everything dilute the signal, because a link to every sibling says nothing about which ones are actually related. Vague anchors waste the one piece of context a link can carry.
Clean structure also helps AI systems working out which page is the authoritative answer on your site. We haven't seen published detail on how any specific answer engine weighs internal links, so treat that as a reasonable inference, not a documented rule.
Sequence the publishing: what to write first and why
Random order wastes the structure. Start with the pillar outline, even though the full pillar page will come later. Then publish the highest-intent articles, the ones closest to a sale, such as cost, comparison and buying-decision pieces. Supporting explainers follow. Update the pillar each time a piece goes live so its links stay current.
A realistic cadence for a small team is one to two quality articles a week, which puts thirty articles at roughly four to six months. Faster is possible, but speed trades against checking facts and adding what only you know. Automation and AI drafting help with throughput: research, first drafts, formatting, translation. Human judgment still has to cover fact-checking, original examples, firsthand experience, and spotting near-duplicates. Two articles aimed at the same query will compete with each other, and no tool reliably catches that for you.
Measure, prune, and refresh instead of assuming it worked
Check Search Console and analytics at 60, 90 and 180 days. Look at impressions and queries first, since they move before clicks do. Then check which articles are indexed, which earn clicks, and whether the pillar page is gaining visibility as the cluster fills in.
Keyword cannibalization means two of your articles targeting the same query, so neither performs well. In Search Console, filter by a query and see whether multiple URLs appear for it. If they do, merge the weaker article into the stronger one, rewrite it toward a distinct angle, or redirect it.
Set honest expectations. Results vary by niche and competition, and a cluster is an investment that compounds slowly, not a quick fix.
What to do this week
- Pick your pillar topic using the overlap test: real expertise, repeated customer questions, realistic demand.
- List 30 candidate questions from Search Console, People Also Ask, customer emails and sales calls.
- Group them into four or five subtopic groups of five to seven articles each.
- Draft the pillar outline, with a placeholder for each planned article.
- Schedule the first five articles, highest-intent first, and write down their internal links before drafting.
Frequently Asked Questions
How many articles do you need in a topic cluster for it to work?
There's no verified minimum. Thirty is a practical target because it lets you cover a subject from several angles, but a tight cluster of ten to fifteen strong articles can still be coherent. Coverage matters more than the count.
Do topic clusters still matter now that people use ChatGPT and Google AI Overviews?
We think so, but with a caveat: the evidence on how answer engines select sources is thin. A site that covers one subject thoroughly and clearly is easier to cite than a scattered one, and the same structure serves human readers either way.
How long does it take for a topic cluster to start ranking on a small website?
Expect months, not weeks. Early signs, such as rising impressions, often show around the 60 to 90 day checks, but timing depends heavily on competition and your site's existing authority.
Should the pillar page be written first or last?
Both, in a sense. Write the outline first so the cluster has a plan, then finish and refine the full pillar as articles go live, since it should link to real, published pages.
Can a small business run more than one topic cluster at the same time?
It can, but we'd advise against it early on. Finish one cluster, check how it performs, then start a second with what you learned.
The point of a cluster isn't to publish thirty articles. It's to make your site the obvious place to find a complete answer on one subject. Plan the structure first and the writing gets easier, because every article already knows its job.
Sources
Photo: cottonbro studio / Pexels
This article was produced with the same Post Pilot workflow we sell: researched and drafted automatically, scored by an editorial pass, and reviewed before publication.