Keyword clustering and topic clusters
Keyword clustering means grouping search terms that the same page could answer, instead of writing one page per keyword. Topic clusters take that a step further, organising a whole subject into one hub page and several supporting pages. Here is how a small Nepali business does both without needing an enterprise SEO team.
Say you sell solar water heaters and, after doing keyword research, you have a list of forty terms: 'solar water heater price Nepal', 'solar geyser cost Kathmandu', 'solar water heater vs electric geyser', 'best solar heater for 4 person family', and so on. The instinct many small businesses have is to write forty separate pages, one per keyword. That is the wrong instinct, and it usually produces forty thin, competing pages instead of a handful of strong ones.
Keyword clustering is the fix: grouping keywords that describe the same underlying question into one cluster, so one well-built page can answer all of them. Topic clusters extend the idea across a whole subject area, organising many clusters into a hub-and-spoke structure. Neither concept requires special software or an agency, just a spreadsheet and some judgement about what a searcher actually wants.
Clustering, the basic idea
Two keywords belong in the same cluster if the same page, written once, would genuinely satisfy both searches. 'Solar water heater price Nepal' and 'solar geyser cost Kathmandu' belong together, they are the same question in different words, price. 'Solar water heater price' and 'solar water heater installation steps' do not belong together, one wants a number and one wants a process, and cramming both onto one page usually serves neither well.
The test is not whether the words look similar. It is whether the intent is the same. 'Solar water heater vs electric geyser' looks like it shares words with a plain price page, but it is a comparison query, the searcher wants a side-by-side, not a price list. That is its own cluster, and probably its own page.
How to actually build the clusters
Start from your full keyword list and work through it methodically rather than trying to sort it all in your head at once.
- Put every keyword into a spreadsheet, one row each.
- For each one, search it and note the intent it reveals: price, comparison, how-to, definition, location-specific, problem-solving. The results page itself tells you this, if the top results are all listicles, it is a listicle intent, if they are all calculators or price tables, it is a price intent.
- Group rows that share the same intent and could be answered by the same page into one cluster. Give each cluster a short label, 'price', 'vs electric geyser', 'installation', 'sizing by household', so the groups are easy to scan.
- For each cluster, pick the single keyword phrase that best represents what the page is really about, this becomes your primary target and usually informs the title and H1.
- Assign one page per cluster. If a cluster only has one or two very low-volume keywords in it and overlaps closely with a bigger cluster, consider folding it into that bigger cluster instead of building a separate thin page for it.
From clusters to a topic cluster structure
Once you have several clusters under one broad subject, solar water heaters, in this example, arrange them as a hub and several spokes. The hub is a broad overview page, 'Solar Water Heaters in Nepal: Complete Guide', that briefly covers every cluster and links out to the dedicated page for each. Each spoke page goes deep on its one cluster, price, sizing, installation, comparison, and links back to the hub.
This structure does two things a scattered set of standalone pages does not. It gives Google, and a reader, a clear map of how thoroughly you cover the subject, which matters for the kind of broad query, 'solar water heater Nepal', that no single narrow page could win on its own. And it spreads internal links deliberately rather than leaving pages to compete for scraps of attention from a generic related-posts box.
Build the hub after the spoke pages exist, not before. A hub with nothing solid underneath it is a thin overview competing against sites that have actually built out the subject. The spoke pages earn the hub its credibility, not the other way round.
Worked example: solar water heaters
Take the original list of forty keywords and see how it collapses. 'Solar water heater price Nepal', 'solar geyser cost Kathmandu', 'solar heater price per litre capacity' cluster into one price page. 'Solar water heater vs electric geyser', 'is solar water heater worth it in Nepal' cluster into one comparison page. 'Best solar heater for 4 person family', 'solar water heater sizing guide', 'how many litres solar heater for family of 5' cluster into one sizing page. 'Solar water heater installation steps', 'solar water heater maintenance', 'solar heater not heating water fix' cluster into one installation-and-troubleshooting page.
Forty keywords, four clusters, four strong pages plus one hub. Each of those four pages is now built to comprehensively answer everything in its cluster, rather than being one of forty thin pages each trying and mostly failing to rank for a single narrow phrase.
Mistakes to avoid
The two failure modes mirror each other, and both waste effort.
- Over-splitting: building a separate page for every keyword variant, which spreads your effort thin and creates keyword cannibalisation, several of your own pages competing against each other for the same intent.
- Under-splitting: cramming genuinely different intents, price, comparison, how-to, onto one page because they share a topic word. The page ends up long and unfocused, and ranks poorly for all three intents instead of well for one.
- Clustering by word similarity instead of by intent, the actual test is always what a single page would need to say to satisfy the search, not whether the words look alike.
- Building the hub page first and never finishing the spoke pages underneath it, leaving a broad, thin overview with nothing to back it up.
Key takeaways
- ✓Cluster keywords by intent, would one page genuinely satisfy both searches, not by how similar the words look.
- ✓One cluster, one page. Splitting a cluster across multiple pages usually causes keyword cannibalisation, not extra traffic.
- ✓A topic cluster is a hub page linking to several deep spoke pages, one per keyword cluster, with links running both ways.
- ✓Build the spoke pages first. A hub page with nothing solid underneath it is a thin overview, not a real topic cluster.
- ✓A large keyword list usually collapses into a handful of real clusters, that collapse is the point of the exercise, not a failure to find more pages to write.
Explore the data behind this guide
Keyword Clustering and Topic Clusters, Explained for a Small Business, FAQ
What is keyword clustering in simple terms?+
Grouping search terms that share the same underlying intent, so one page can answer all of them, instead of writing a separate thin page for every keyword variant. The test is whether one page would genuinely satisfy both searches, not whether the phrases share words.
What is the difference between keyword clustering and a topic cluster?+
Keyword clustering groups individual search terms into the intents a single page can answer. A topic cluster is the next layer up: organising several of those clusters into one hub page, a broad overview, that links to a dedicated spoke page for each cluster.
Do I need special software to do keyword clustering?+
No. A spreadsheet and a willingness to search each keyword and read the results page is enough for a small business's keyword list. Paid clustering tools mainly save time on very large lists, they are not required to do this correctly.
How many keywords should be in one cluster?+
As many as genuinely share the same intent and could be answered by one well-built page, there is no fixed number. A cluster with only one or two low-volume keywords is often better folded into a related bigger cluster than built as its own thin page.
Should I build the hub page or the spoke pages first?+
The spoke pages first. A hub page written before the detailed pages underneath it exist is just a thin overview with nothing to back it up. Build three or more solid spoke pages on a subject, then write the hub to tie them together and link out to each.
Related guides
Sources & data note
The pillar-and-cluster content model is documented, widely-adopted practice popularised by HubSpot, cited above, and the general principle of matching one page to one search intent reflects Google Search Central's published guidance on helpful content, also cited. The solar water heater worked example, including the specific keyword list and how it collapses into four clusters, is our own illustrative construction to demonstrate the method, not a real client's keyword research or measured search volumes. Guides are written from primary sources, Nepali government departments, operators, park authorities and standards bodies, and each guide lists the sources used for its own facts. Rules, fees and prices in Nepal change; treat figures as current at the review date shown on each guide and verify anything money- or visa-critical with the issuing authority before you rely on it.