---
title: Keyword clustering: how to group keywords into pages
description: Keyword clustering step by step: results overlap versus meaning based grouping, a worked example from real keyword data, and how to fix cannibalisation.
canonical: https://www.postbing.com/blog/keyword-clustering-guide
published: 2026-11-16
updated: 2026-11-16
author: The Postbing Team
---

# Keyword clustering: how to group keywords into pages

Keyword clustering is grouping keywords so that each group becomes one page: keywords that share a search intent, and usually the same Google results, go on the same page, and keywords with different results get separate pages. The most reliable method compares the top ten results for each keyword. Clustering by meaning alone is faster but often puts the wrong keywords together.

Last updated: November 16, 2026

## What is keyword clustering for?

Clustering decides how many pages you need and which keyword each page targets. Without it, you either write one page that tries to serve five intents, or five pages that compete for one. Both waste writing time.

It is the second of the six decisions in the [SEO blog strategy guide](/blog/seo-blog-strategy-guide): after you have a list of searches worth answering, clustering turns that list into pages. A separate, larger grouping (pages into [topic clusters with a pillar](/blog/topic-clusters-and-pillar-pages)) comes after this step. People mix up the two because both are called clustering.

| | Keyword clustering | Topic clustering |
|---|---|---|
| Groups | Keywords into pages | Pages into a subject |
| Question it answers | Does this keyword need its own page? | Which pages link to which pillar? |
| Main test | Do the search results overlap? | Does a reader need these pages together? |
| Output | One primary keyword plus variants per page | A pillar and its spokes |

## Two ways to cluster: by results or by meaning

**Results based clustering** compares the top ten Google results for each pair of keywords. If several of the same URLs rank for both, Google treats them as the same need, so you target both with one page. A common practitioner threshold is three or more shared URLs in the top ten; it is a rule of thumb, not a Google rule, so treat borderline cases with judgment.

**Meaning based clustering** groups keywords by shared words or by semantic similarity. It is quick and works for an initial pass, but it misses cases where similar words have different intent ("CRM software" and "what is CRM") and cases where different words share one intent ("content calendar" and "editorial calendar").

Use meaning based grouping to sort a long list into rough buckets, then results based checks to make the final call on anything ambiguous.

## How to cluster keywords in seven steps

1. **Collect the list.** Export keyword ideas from your research tool, Search Console queries and your own list of buyer questions. Finding them is covered in [how to find blog topics](/blog/how-to-find-blog-topics).
2. **Clean it.** Remove duplicates, misspellings, brand terms you will not target, and keywords with no link to what you sell.
3. **Tag intent.** Mark each keyword as learn (what, why, how), compare (best, vs, alternatives), or act (pricing, template, tool). Keywords with different intent rarely belong together.
4. **Make rough buckets** by shared meaning within each intent group.
5. **Check results overlap** inside each bucket. Search each keyword in a private window from your target country and note the top ten URLs. Split any keyword whose results differ.
6. **Pick the primary keyword** for each final group: usually the one with the most searches that also states the intent clearly. The rest become close variants.
7. **Assign a page type**: how-to, explainer, comparison, checklist. The type then drives the brief and length.

For a list of 200 keywords, steps 1 to 4 take an hour or two in a spreadsheet. Step 5 is the slow one, which is why paid tools automate it by pulling results data in bulk.

## A worked example from this blog's own keyword list

Here is how some of the keywords behind this Blog Strategy cluster were grouped. Volumes are US monthly searches, November 2026 (Serpstat estimates, DataForSEO for the last row); third party tools often disagree on small numbers like these.

| Keyword | US monthly searches | Grouped onto |
|---|---|---|
| topic clusters seo | 90 | One page: topic clusters and pillar pages |
| pillar page seo | 90 | Same page |
| seo content calendar | 40 | One page: SEO content calendar |
| how to build a content calendar | 40 | Same page |
| blog content calendar template | 90 | Same page (the column template covers it) |
| seo content brief | 110 | One page: SEO content brief |
| how to write a content brief | 40 | Same page |
| how often should you blog | 90 | Own page |
| how many blog posts per week for seo | 10 | Same page as "how often should you blog" |

The reasoning: "pillar page seo" and "topic clusters seo" ask about two halves of one structure, so a reader looking for either needs both. "Blog content calendar template" has a more specific intent (a template), so the calendar page includes the column table to satisfy it. If a results check later showed that template searches are served only by downloadable files, that keyword would move to its own page.

## How to spot and fix keyword cannibalisation

Cannibalisation is when two of your own pages compete for one query and neither ranks as well as one combined page would. In Search Console, open the Performance report, filter by the query, and switch to the Pages tab. If two or more of your URLs collect impressions for the same query, and their positions swap from week to week, you have it.

The fix depends on whether both pages earn their place:

- **Same intent, both thin:** merge into the stronger URL and redirect the other to it with a 301.
- **Same intent, both substantial:** merge the best sections, redirect, then refresh.
- **Different intent, overlapping wording:** keep both, sharpen each title and opening answer, and link between them with anchor text that states the difference.
- **Near duplicates you must keep** (for example print versions): use `rel="canonical"` to point to the main version, as described in Google's documentation on consolidating duplicate URLs.

## Mistakes that make clusters wrong

- **Grouping by volume.** High volume keywords are not automatically the primary keyword if their intent is vaguer.
- **One page per variant.** "Best X", "top X" and "X list" are one page.
- **Never re-checking.** Results change. Recheck big clusters once or twice a year.
- **Mixing intents.** "What is a content brief" and "content brief template" may need different page types even though they share words.

Once each group is final, write a one page brief per group; [how to write an SEO content brief](/blog/how-to-write-an-seo-content-brief) shows the fields.

## FAQ

### What is the difference between keyword clustering and topic clustering?
Keyword clustering groups keywords into pages. Topic clustering groups pages into a subject with a pillar page and linked spokes. You do keyword clustering first.

### How many keywords should one page target?
One primary keyword plus however many close variants share its intent and results. That might be two keywords or twenty; the count matters less than the shared intent.

### How many shared results mean two keywords belong on one page?
Many practitioners use three or more shared URLs in the top ten as the threshold. It is a rule of thumb, not a Google rule, so check borderline pairs by reading the results yourself.

### Can I cluster keywords without a paid tool?
Yes. A spreadsheet and manual results checks work well for a few hundred keywords. Paid tools mainly save time on the results overlap step for large lists.

## Sources

- [Google: How to specify a canonical URL](https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls): rel="canonical" for duplicate or very similar pages (page last updated 10 July 2026), accessed 16 November 2026
- [Google Search Console Help: Performance report](https://support.google.com/webmasters/answer/7576553?hl=en): queries, pages and position metrics, accessed 16 November 2026
- Serpstat and DataForSEO keyword data, US, pulled 16 November 2026: monthly search volumes in the worked example

## Related reading

- [SEO blog strategy: the whole plan on one page](/blog/seo-blog-strategy-guide)
- [Topic clusters and pillar pages](/blog/topic-clusters-and-pillar-pages)
- [How to find blog topics worth writing](/blog/how-to-find-blog-topics)
- [How to write an SEO content brief](/blog/how-to-write-an-seo-content-brief)
- [Why Postbing avoids link schemes and risky tactics](/safe)
