A keyword clustering tool answers the planning question that comes right after keyword research: which of these keywords belong on the same page? Paste your list and this tool sorts it into clusters, names a primary keyword for each, adds up the search volume and labels the likely intent. Each cluster is a candidate page, which turns a spreadsheet of hundreds of terms into a page plan you can review and act on.
How to Use the Keyword Clustering Tool
- Keywords: paste one keyword per line, up to 2,000 per run. To weight clusters by demand, add the monthly search volume after a comma or a tab, as in "shopify seo agency, 1700". Enter volumes as plain numbers without thousands separators (4700, not 4,700), because the tool reads the comma inside 4,700 as a separator. Lines without a volume are still clustered.
- Grouping strictness: a slider from 30% to 90%, set to 50% by default. Lower values make fewer, broader clusters. Higher values split the list into more, tighter clusters.
- Words to ignore (optional): comma-separated words or phrases to remove before comparing, such as your brand name, a year or a city that appears in every keyword. Without this, one shared word can pull unrelated keywords together.
Click Cluster keywords to run it. A note shows how many keywords went into how many clusters, and the results table lists each cluster with:
- Primary keyword (page): the keyword that started the cluster, which is its highest-volume keyword. Treat it as the main target for that page's title and H1.
- Keywords in the cluster: every keyword grouped there, with its volume if you gave one.
- Total volume: the sum of the cluster's volumes. The table is sorted by this total, so the biggest opportunities come first.
- Likely intent: informational, commercial, transactional, navigational, local or mixed, based on the words in the primary keyword.
Download CSV gives one row per keyword with the cluster number, primary keyword, keyword, volume and that keyword's own intent label, ready for a spreadsheet or a content brief. Everything runs in your browser, and your list is not uploaded or stored.
How Keyword Clustering Works in This Tool
This is an automatic keyword clustering tool based on word overlap. It does not query Google, so it is instant and free, and you can see exactly why two keywords ended up together. The steps:
- Clean each keyword. The tool lowercases it, removes your ignore words and drops filler words that do not change the topic, such as how, what, best, top, for, vs, near me and the years 2024 to 2026. It then trims simple endings, so services becomes service, agencies becomes agency and marketing becomes market.
- Sort by volume. Keywords are processed from highest volume to lowest, so the biggest term in each group becomes its primary keyword. Without volumes, the keyword with the fewest meaningful words goes first.
- Score against each cluster. Each keyword is compared with the primary keyword of every existing cluster. The score blends two measures: 60% weight on the share of the shorter keyword's words that are shared, and 40% on the shared words as a share of all distinct words in both.
- Join or start a cluster. If the best score reaches your strictness setting, the keyword joins that cluster. If not, it starts a new cluster and becomes its primary keyword.
- Label and sort. Each cluster gets an intent label from its primary keyword, and clusters are sorted by total volume.
Because filler words are removed, "best shopify seo agency" and "shopify seo agency" count as a full match. That is deliberate: a modifier such as best rarely justifies a separate page. It also means the score cannot tell a how-to guide from a service page when both share the same core words, which is what the intent column and a quick look at the search results are for.
Worked Example: One Cluster or Three?
These five keywords are the tool's own sample list. The volumes are example numbers, not real data or benchmarks.
| Keyword | Example volume | Words compared | Score against "shopify seo services" |
|---|---|---|---|
| shopify seo services | 4,700 | shopify, seo, service | Primary keyword |
| seo for shopify | 2,400 | seo, shopify | 87% |
| shopify seo agency | 1,700 | shopify, seo, agency | 60% |
| shopify seo checklist | 450 | shopify, seo, checklist | 60% |
| how to do seo on shopify | 300 | seo, shopify | 87% |
Take "shopify seo agency". It shares 2 of its 3 words with the primary keyword (67%, weighted at 60%, gives 40 points), and 2 of the 4 distinct words across both keywords (50%, weighted at 40%, gives 20 points), for a score of 60%.
At the default 50%, all five keywords form one cluster with a total volume of 9,550, labeled commercial because the primary keyword contains "services". At 70%, the tool returns three clusters: "shopify seo services" with "seo for shopify" and "how to do seo on shopify" (7,400), "shopify seo agency" (1,700) and "shopify seo checklist" (450).
Neither run is the final answer. "Shopify seo services" and "shopify seo agency" tend to return the same kind of results, agency and service pages, so they normally belong on one page even though the 70% run split them. "How to do seo on shopify" and "shopify seo checklist" return guides, so they belong on a blog post even though the 50% run put them with the service page. The tool does the bulk sorting in seconds; the close calls need a human look at the results page.
Choosing the Right Grouping Strictness
| Setting | What you get | Use it for |
|---|---|---|
| 30% to 40% | Few, broad groups that share one or two core words | Topic planning: finding pillar topics and the subtopics under them |
| 50% to 60% | Page-level clusters | Deciding which keywords one page should target |
| 70% to 90% | Many small, tight groups | PPC ad groups, or spotting near-duplicate keywords |
Start at 50%, read the largest clusters and move the slider if they look too mixed or too fragmented. If the same topic is spread across several small clusters, lower the setting. If one cluster mixes guides with service terms, raise it or split that cluster by hand.
Word-Overlap, Semantic and SERP-Based Keyword Clustering
There are three main ways a keyword clustering tool decides that two keywords belong together, and each fits a different job.
| Method | How it groups | Strengths | Weak spots |
|---|---|---|---|
| Word overlap (this tool) | Shared meaningful words after cleanup | Free, instant, private and explainable; handles thousands of keywords | Misses synonyms such as cheap and affordable; cannot see intent differences between keywords with the same words |
| Semantic (embeddings or AI) | Similar meaning, judged by a language model | Catches synonyms and rephrased questions | Can merge keywords that mean the same but rank different pages; the reason for a grouping is harder to inspect |
| SERP overlap | Keywords whose top 10 results share several URLs | Closest to how Google already treats the keywords | Needs live results data for every keyword, which costs money and time at scale; results change over time |
A practical workflow uses them in order: word overlap to sort a large export into rough groups for free, then a SERP check on the clusters that matter most or look borderline. A semantic keyword clustering tool earns its place when your list is full of synonyms and rephrasings, which is common with question keywords.
From Keyword Clusters to a Page Plan
- One cluster, one page. Each cluster that survives review becomes one URL. The primary keyword goes in the title tag, the H1 and the first paragraph.
- Use the rest of the cluster as the outline. Secondary keywords and questions become H2s, FAQ questions and phrases in the copy, so one page covers the whole group.
- Map clusters to existing URLs. Before writing anything new, check whether a page already targets the cluster. Improving an existing page usually beats publishing a competing one. Our keyword mapping service does this across a whole site.
- Check intent per cluster. Run the primary keywords through our search intent tool and confirm the page type: guide, category, product, service or comparison.
- Group clusters into topics. Related clusters become a topic cluster: one pillar page linked to supporting pages. Our topic cluster service plans and builds these structures.
- Prioritize. Rank clusters by total volume, how close the intent is to a sale and how realistic it is to compete, then publish in that order.
How Many Keywords Should One Cluster Have?
There is no correct number: a cluster should hold every keyword one page can satisfy, whether that is two or fifty. A niche product might have a primary keyword and a handful of variants. A broad service term can gather dozens of long-tail versions that all want the same page.
What matters is that the keywords share a page type and an answer. If a cluster has grown to hundreds of keywords, it is probably a topic rather than a page, and running just that cluster again at a higher strictness will show the pages inside it.
Keyword Clustering Mistakes That Create Thin or Competing Pages
- One page per keyword. Separate pages for "shopify seo agency" and "shopify seo company" split links and relevance across near-identical pages that compete with each other.
- Trusting any algorithm on close calls. Every automatic method gets borderline keywords wrong. Review the biggest clusters by hand before you brief writers.
- Leaving brand and location words in. A brand or city in every keyword pulls everything into one cluster. Add them to the ignore box.
- Clustering without volume. Without volumes, the shortest keyword becomes the primary keyword, which may not be the one people search most.
- Mixing intents on one URL. A how-to guide and a hire-us page on the same URL usually serve neither searcher well.
- Never revisiting the plan. Search results shift, so recheck important clusters when rankings stall.
Keyword Clustering for PPC Ad Groups
The same tool works for paid search: a cluster at a high strictness setting is a tight ad group. Set the slider to 70% or more, and each cluster's keywords share enough words that one set of ads and one landing page can match them closely. Use the per-keyword intent column in the CSV to keep transactional and informational keywords apart, since informational terms often work better as negative keywords than as ad groups. Our negative keywords guide covers how to build that list.
Turn Clusters Into Pages That Rank
Clustering is the sorting step. The work that moves traffic is choosing which clusters to win, briefing and writing the pages, and linking them into a structure search engines understand. Our SEO services cover keyword research and mapping, and our content marketing team writes the pages from the cluster outlines. If you have a clustered list and want a second opinion on the page plan, send it to us and we will point out the overlaps and gaps we see.
For more free tools, including the keyword density checker for checking drafts, see our marketing tools page.