Keyword clustering

关键词输入

每行一个关键词，支持中英文词库做首轮轻量聚类。

聚类结果

按共享词片段快速分桶，适合先做词库清洗，再继续人工复核。

## 关键词聚类 (3)
- 关键词聚类工具
- 关键词聚类教程
- 关键词聚类怎么做

## SEO (2)
- SEO关键词分组
- SEO内容规划

## UTM (2)
- utm builder
- utm link generator

## Keyword (1)
- keyword clustering tool

The first round of lightweight clustering of Chinese and English keywords is done online, and the same topics, similar modifiers and potential word grabbing items are quickly divided into buckets. It is suitable for SEO vocabulary cleaning, topic page planning, column splitting, FAQ topic selection and internal link map sorting. The results are more suitable for preliminary screening first, and then continue manual review and page allocation.

After clustering, continue to complete the page Meta tags

Really focus on the title, description, canonical and share card

Generate more standardized SEO-friendly URL Slug for keyword phrases

Suitable for use when setting fixed links on topic pages, column pages and article pages

After writing the content, check the length, paragraphs and reading time

It is convenient to verify whether the content density of the page corresponding to each keyword group is sufficient.

Multilingual keywords further output hreflang tags

Suitable for international sites to synchronize content on the same topic to multi-language versions

Batch organize the newly planned pages into Sitemap

It is suitable to complete the site crawling entrance after the topic page and new column are launched.

Configure more robust crawling rules for test pages and staging pages

Prevent keyword test pages or intermediate pages from accidentally being included in the index

What is keyword clustering, and what problems does it solve in SEO?

The core of keyword clustering is not to "arrange" a bunch of words neatly, but to answer a more critical question: which words should fall on the same page, and which words should continue to be split into different pages. True SEO planning rarely starts with writing a title, but with a thesaurus structure. If the lexicon is not grouped first, whether it is creating columns, topic pages, FAQs, internal links, or content calendars, problems such as repeated writing, page competition for words, and site structure confusion will easily occur.

From a workflow perspective, keyword clustering is usually between “keyword collection” and “page planning”. You may first get a batch of query terms from Search Console, Baidu Statistics, Google Ads Keyword Planner, on-site search terms, competing product pages, or customer service Q&A, and then clean, remove duplications, and denoise, and then put the same topics, similar modifiers, and obviously related long-tail words into a bucket through clustering. After completing this step, the keyword list will look more like an executable map rather than a messy export table.

It should be noted that there is not only one way to do keyword clustering. This page is more suitable for the "first round of lightweight clustering", which is to quickly bucket based on shared word fragments to help you first identify topic clusters, page clusters and potential word clusters. It is suitable for vocabulary cleaning, topic page drafting, FAQ planning and pre-processing before content collaboration, but it is not equivalent to strict SERP overlap clustering. If you want to decide whether two words can be used in the same article, you still need to make manual judgment based on the search results page, ranking intention and business conversion path.

For the content team, the value of clustering is to reduce the impulse to "open a page when a word comes to mind"; for the SEO team, the value of clustering is to detect page word grabbing and structural faults earlier; for the advertising team, the value of clustering is to align keyword groups with ad groups, landing pages, and information architecture. Because of this, a useful clustering result should not only stay at the word list level, but should continue to extend to the main page, auxiliary pages, target intent, anchor text and internal link direction.

Use Cases

Group the query words exported from Search Console for the first round to determine which words should be incorporated into existing pages and which words are worthy of opening a new topic page
Organize the long-tail words exported by Baidu or Google keyword tools, and first separate tool words, tutorial words, price words, comparison words and question words
When making a keyword map for the SaaS official website, first divide product function words, solution words, industry words and alternative words into buckets, and then arrange the landing page
When building a help center or FAQ column, break down question words like "what is it, how to do it, why, error reporting, and repair" into content clusters
When planning the category page of an independent website, separate product words, brand words, attribute words, scene words and regional words to avoid category pages and article pages competing for words.
Before revamping the website, regroup the vocabulary of the old site to identify which columns should be merged and which columns should be split from one large page into multiple sub-pages.
When selecting topics for the content team, divide the introductory words, tutorial words, template words, case words and checklist words under the same topic into different writing tasks
When organizing ad groups for advertising, separate words with high commercial intent and pure information words to reduce the problem of one landing page taking on multiple intents.
When optimizing internal links, put subject words that can support each other into the same group, and reverse the division of labor between the main page, sub-pages and anchor text.
When sorting out brand protection words, group brand words, competitive product words, brand + function words, and brand + price words separately to avoid mixing them with common words.
When processing regional SEO vocabulary, separate city words, province words, store words and local service words to facilitate planning of regional page levels.
When planning the structure of a topic page or collection page, first check whether there are enough related words in a group to support the combination of the main page and sub-pages.

How to Use

First organize the keywords into a "one word per line" format, and then delete obvious duplicate words, irrelevant words and pure noise words
Paste the cleaned word list into the tool, and let the page do the first round of lightweight clustering based on shared word fragments.
Review the results group by group, focusing on whether a group contains price words, tutorial words, brand words, regional words, contrast words and other modifiers that need to be further separated.
Add the main page, secondary pages, target intent and internal link direction to each clustering group, and organize the results into a truly executable keyword map
After the content is online, continue to review it with Search Console or on-site data: which groups need to be split, which groups can be merged, and which groups still lack pages to take over?

Features

Both Chinese and English keywords can be divided into buckets first: it is suitable for the first round of grouping of mixed word lists exported from Baidu, Google, Search Console or advertising platforms.
Quickly cluster by shared word fragments: Prioritize the repeated dominant words, subject words and combination words to facilitate the first cleaning and then segmentation of the vocabulary
More suitable for page planning rather than pure duplication removal: in addition to seeing which words are the same, you can also quickly find which words should be grouped into the same topic page, column page or FAQ page
Helps discover page grabbing words in advance: If there are tutorial words, tool words, price words, and brand words in the same group, it usually means that the page will continue to be split later.
The result is a copyable Markdown structure: after clustering, it can be pasted directly into documents, tables, Feishu or Notion, and then assigned to content, SEO or delivery teams
Browser local processing: The keyword list is calculated on the current page and does not need to be uploaded to the server first. It is suitable for internal thesaurus and unpublished topic selection.

Comparison of common practices for keyword clustering

Which clustering method to use first depends on whether you are cleaning the vocabulary, determining search intent, or have entered the page online stage.

method	most suitable stage	Advantages	limitations
Lightweight clustering of shared word fragments (this page)	Preliminary screening of vocabulary, draft structure, and topic selection	Fast, low threshold, suitable for bucketing large quantities of keywords first, and quickly discovering topic clusters and potential word clusters.	It cannot alone replace SERP judgment; words with the same theme but different intentions may still need to be manually split.
SERP overlapping clustering	Determines whether multiple words share the same page	It is closer to the real search results and suitable for judging whether two words can be jointly carried out on one page.	The cost is higher, search results or ranking page data need to be collected, and the implementation complexity is also higher.
Artificial keyword map	Finalization of page division, internal links and columns before going online	It is most suitable for business goals and can truly implement the main page, sub-pages, conversion intentions and content structure.	Time-consuming, dependent on operator experience, usually needs to be completed based on the first two clustering results

Best Practices

Clean the vocabulary first and then do clustering

Remove obviously repeated words, pure brand words, noise words and irrelevant words first, and then import them into the clustering tool. The earlier you clean, the higher the readability of each subsequent group, and the easier it is to see the true page structure.

Each group will answer only one question on the page first.

If modifiers such as "tools, tutorials, prices, brands, downloads, and comparisons" appear in a group at the same time, it usually means that the group is too thick and the pages should continue to be split instead of allowing one page to bear multiple purposes at the same time.

Transform clustering results into keyword maps

Don’t stop at “break into groups and it’s over.” It is recommended to continue to fill in at least four columns: main page, auxiliary page, goal intention, and next action. In this way, the clustering results can truly guide writing, revision, and internal links.

Prioritize repairing the groups with the most severe page grabbing.

If a group corresponds to multiple online pages, first check whether these pages are competing for the same subject. Word-grabbing problems usually hurt rankings before "long-tail words haven't been covered yet".

Separately look at question words, transaction words, and navigation words

Words like "what" and "how to do" are more like description pages or FAQs; words like "price", "buy" and "comparison" are more commercial; brand words and navigation words often need to be undertaken independently. Don’t merge them just because the topics are similar.

After going online, use real data to review whether the clustering is reasonable.

After the content is online, you can continue to review the display words, click words and ranking page performance of Search Console: whether a group needs to be split into multiple pages, or whether multiple groups can actually be merged into the same content cluster.

FAQ

What is the difference between keyword clustering and keyword deduplication?

Duplication removal only solves "how many times the same word appears", while keyword clustering solves "which words should fall on the same page or the same content group". When actually doing SEO planning, clustering is more valuable than simple deduplication, because it directly affects the column structure, topic page splitting and internal link layout.

Is this tool suitable for SERP overlap clustering?

Not exactly the same. This page is more suitable for the "first round of lightweight clustering", that is, quickly bucketing the vocabulary based on shared word fragments to facilitate structural organization and manual review; if you want to do strict SERP overlap clustering, URL co-occurrence clustering or ranking page overlap analysis, you usually need to combine search result data and manual judgment.

Is it suitable to organize long-tail words, question words and modifiers?

suitable. In particular, modifiers such as "how to do it, what it is, price, recommendation, tutorial, template, case, download" often determine whether a word should ultimately be classified into a tutorial page, a tool page, a transaction page, or an information page. Clustering first and then splitting pages is usually more stable than writing the content directly.

Can Chinese keywords also be used?

Yes. This page supports the first round of quick grouping of Chinese and English keywords according to shared word fragments. However, it should be noted that Chinese search terms often have stronger semantic differences. For example, modifiers such as "tutorial", "tool", "price" and "comparison" may determine completely different page types, so the clustering results are more suitable for preliminary screening first and then manual review.

Why are some words grouped together but I think they should be separated?

Because the first round of clustering is more biased towards “theme similarity”, and the actual page planning also depends on search intent, SERP ranking pages, business stages and conversion paths. For example, "Keyword Clustering Tool" and "Keyword Clustering Tutorial" may be classified into the same topic, but they should usually be split into a tool page and a tutorial page when going online.

How many pages should a clustering group correspond to?

There is no fixed answer. Generally, the same subject words are put into the same group first, and then check whether brand words, price words, comparison words, tutorial words, and navigation words appear in the group at the same time. If the modifiers are significantly different, it is usually better to continue splitting the group into 2 to 4 pages rather than shoehorning it into the same page.

Can keyword clustering help deal with page competition for words?

Can. Page word grabbing is often caused by multiple pages covering the same topic at the same time without a clear division of labor. First use the clustering results to group similar words, and then assign each group a main page, several auxiliary pages and clear anchor text. You can usually see where there is a risk of word grabbing more quickly.

In addition to SEO, can it also be used for advertising and information architecture?

Yes. In advertising, the keyword clustering results can be used to split ad groups and landing pages; in content operations, they can be used to select quarterly topics; and in website revisions, they can be used to rearrange the structure of columns, topic pages, FAQ pages, and help centers.

What should be the next step after clustering?

Usually, instead of writing an article immediately, you first fill in 4 fields for each group: main page, secondary page, target intent, and whether to include internal links. This way you can turn the clustering results into a real keyword map, rather than just a "neat-looking" grouping list.

How to improve the usability of clustering results when the vocabulary library is large?

It is most effective to do three steps of preprocessing first: the first step is to remove duplicate lines and obviously irrelevant words; the second step is to pull out brand words, regional words, and model words separately; the third step is to import in batches according to product lines or business lines. This makes it easier to get an executable page structure than mixing thousands of rows together and clustering them all at once.

Glossary

Keyword clustering: The process of organizing keywords with the same theme, similar modifiers, or keywords that may be carried out by the same page into the same group for the purpose of lexicon cleaning, page planning, and content structure design.
long tail keywords: More specific, longer, and clearer search terms usually have lower search volume but clearer intent, and are often used for tutorial pages, FAQ pages, scenario pages, and conversion pages.
search intent: What users really want to accomplish when initiating a search include information acquisition, navigation search, tool use, price comparison, purchase decision, etc.
keyword map: Further mapping the keyword groups to the planning table on the main page, sub-page, column level and internal link structure is the next step after clustering.
Page grabbing words: Multiple pages compete for the same subject or the same intent at the same time, making it difficult for search engines to determine which page should be ranked best. Common manifestations include ranking fluctuations and page substitutions.
Topic page: A main page built around a larger core topic usually inherits the main keyword, and continues the more detailed long-tail keywords through sub-pages, FAQ or tool pages.
Column page: A relatively stable category or channel page on the site is suitable for hosting a set of ongoing keywords, rather than one-time event words or short-term hot words.
anchor text: The clickable text portion of an internal or external link. After clustering is completed, the anchor text should be consistent with the page division to avoid conflicts with multiple pages using the same main anchor text.
SERP overlap: See whether two keywords often appear on the same ranking page on the search results page to determine whether they are more suitable to share a page or build separate pages.
Content Cluster: A combination of main page and sub-page built around a core theme. The common form is pillar page + supporting pages, which strengthens the topic relevance through internal links.

Which pages are more suitable for different keyword types?

After the clustering results come out, first determine what kind of page the subject in the group is more like, and then decide whether to continue splitting pages.

Keyword type	Common signal words	A more suitable page	Handling suggestions
Tool word	Tools, generators, checkers, calculators, converters	Tool page/online usage page	The page should first meet the requirements of "can be used immediately", and then add usage scenarios and comparisons in the FAQ and description areas.
tutorial words	Tutorials, how-tos, steps, methods, guides	Tutorial Page / Blog Page / Help Documentation	Emphasize more on operating procedures, examples and pitfall avoidance, and do not hard-merge them with pure tool pages.
Definition word	What is it, meaning, principle, difference	Explanation page / Glossary page / FAQ topic	Suitable for educational content and guiding traffic to subsequent tool pages or service pages.
business words	Price, cost, purchase, recommendation, which one is better, comparison	Plan page/landing page/comparison page	Usually closer to conversion, should be processed separately from pure information words
brand words	Brand name, product name, official website, login, download	Brand page / Navigation page / Product page	Don’t mix it with common words, otherwise it will easily dilute the brand intention
regional words	Beijing, Shanghai, local, nearby, city name	Region page/store page/service page	It should be looked at together with the regional level, service scope and duplicate content risk.
scene word	Templates, cases, checklists, examples, applicable scenarios	Case page/template page/resource page	Suitable for expanding from the main page to supporting content clusters
question word	Why, error, failure, failure, repair	FAQ Page / Troubleshooting Page / Support Documentation	More suitable for problem-oriented content, often paired with tutorial pages or help centers

Clustering result review checklist

Just because a group of words "looks like a group" doesn't mean they should share the same page; the table below is more suitable for a secondary check.

checkpoint	What to focus on	Next action
Is the subject consistent?	Does the group all revolve around the same core theme, rather than just sharing a general word?	If only general words are the same, continue to subdivide into smaller topic groups
Are search intent mixed?	Whether there are obviously different intentions such as tutorials, tools, prices, brands, downloads, comparisons, etc. at the same time	Split different modifiers into different page types
Whether the existing page has been captured?	Whether the same group of keywords has been covered by multiple old pages	Determine a main page and change the remaining pages to support pages or adjust anchor text
Is there enough words to support a single page?	Does a group have only one or two isolated words, or has it formed a stable topic cluster?	When there are insufficient words, merge them into the upper-level topic page first. There is no need to force a separate page.
Is column level required?	Is this group more like a topic page or a sustainably expanding column?	Stable and long-term scalable themes are more suitable for column pages or main topic pages
Is it suitable to be a FAQ/subpage?	Whether there are a large number of question words, case words, template words or scene words in the group	Split them into supporting content and strengthen internal links for the main page.

Privacy & Security

All keyword clustering on this page is done locally in your browser. The keyword list you paste will not be uploaded to an external server, nor will it be written to server logs, databases, or long-term storage. After closing or refreshing the page, the current input content is cleared from the page state. This page is more suitable as a first-round sorting tool for undisclosed topics, internal thesaurus, advertising word lists, and brand monitoring words.

Authoritative References

Google Search CentralGoogle Search Central: SEO Starter Guide
Google Search CentralGoogle Search Central: Link best practices
Google Ads HelpGoogle Ads Help: Use Keyword Planner
Google Search CentralGoogle Search Central Blog: Google does not use the keywords meta tag