What Is a Keyword Grouping Tool?

A keyword grouping tool turns a flat list of keywords into a plan for pages. Without grouping, a keyword list is just a pile of phrases, and every phrase looks like it deserves its own page. With grouping, related phrases sit together, each group maps to one page, and the writer knows which keyword leads the page and which keywords support it. That single change prevents a large share of the muddle that weakens small sites, including pages that compete with each other, thin pages that say almost the same thing, and strong topics split across too many weak URLs. This guide explains what a keyword grouping tool does, how grouping differs from clustering and from topic clusters, how the free LinkTrixx tool groups keywords by the words they share, and how to turn its groups into pages that cover a topic fully without competing with each other.
Grouping is a planning step, and it belongs before writing, not after. You group first, then you brief, then you write, then you check the copy. The Keyword Grouping Tool on this site is the grouping step in that chain. It is free, needs no signup, and runs in your browser. You paste a keyword list, one keyword per line, and groups form live as you type, based on the words and two word phrases the keywords share. The companion guides on this site cover the next steps. What Is a Keyword Position Checker? and How to Use a Keyword Position Checker show how to check where a keyword sits inside the copy once the page is drafted, and the grouping examples in this guide use only the sample and the test list actually observed on the live grouping tool, quoted exactly, so every count and every group name in this guide is a real output from the live tool, not an invented illustration.
What keyword grouping means
Keyword grouping is the process of sorting keywords that belong together into groups, so that each group can be targeted by one page. Semrush defines keyword clustering, the close cousin of grouping, as an SEO technique centred on grouping search terms that share the same search intent and targeting them together on a single page. Ahrefs defines keyword clustering in similar terms, as the process of grouping keywords with the same or similar intent, so that you can target them with one page instead of multiple pages. Grouping uses the same idea at a simpler level. It asks a practical question about each keyword. Does this keyword belong on the same page as that keyword, because a reader who wants one would be satisfied by a page that covers both, or does it need a page of its own, because the reader wants something different.
Each group gets a primary keyword and secondary keywords. The primary keyword is the main term the page is built to rank for. It appears in the title, the heading, the opening paragraph and the URL slug, and it names the page in plain language. Secondary keywords are the synonyms, longer phrasings and related variants that the same page also covers, naturally, in headings and body text. Semrush makes the same distinction in its clustering guide, and its worked example shows one page ranking for a large number of related keywords once the cluster is covered well. The point is not to stuff every secondary keyword into every paragraph. The point is to write one comprehensive page that answers the whole group, so the secondary keywords appear because the coverage is complete, not because a list was pasted into the text.
Grouping also answers a counting question that keyword research alone cannot answer. A list of 200 keywords is not a plan for 200 pages. It may be a plan for 20 pages, or 40, depending on how the keywords group. A list of 2,000 keywords, in one example cited by a clustering guide, can become 150 clusters, and those clusters become a realistic editorial calendar. That reduction is the value. It turns an overwhelming list into a finite set of pages, each with a clear job, and it stops the commonest planning error in SEO, which is to publish a new thin page for every new phrase that research returns.
Grouping, clustering and topic clusters are three levels
Three terms are used loosely online, and keeping them apart makes the rest of this guide far easier to follow. Keyword grouping is the level this tool works at. It groups keywords by shared words and phrases, so you can see at a glance which phrases belong together in wording. Keyword clustering is the level above it in rigour. It groups keywords by shared search intent, usually by comparing the search results Google returns for each keyword, so that keywords go together when Google already treats them as the same need. Topic clustering is the site level. It connects multiple pages around a broader theme, typically with a pillar page that covers the theme broadly and supporting pages that cover subtopics in depth, all linked together.
Semrush draws the same line in its clustering tool guide. Keyword clustering is page level and determines which keywords belong on a single page. Topic clustering is site level and connects multiple pages around a broader theme. The two work together, but they are not the same thing. A topic cluster may contain many keyword groups, one for each page inside it. A keyword group, on the other hand, is one page. Confusing the levels leads to two opposite errors. Treating every keyword group as a topic cluster produces a site with too many pillar pages and no depth. Treating a whole topic cluster as one keyword group produces a single page so broad that it answers nothing well.
The LinkTrixx tool works at the grouping level, by shared words. That is a strength and a limit, and this guide is honest about both. Shared word grouping is fast, transparent and easy to check by eye. You can see exactly why two keywords grouped together, because the shared phrase is visible in the group name. It needs no search data, no account and no crawl, which is why it can run free in a browser. Its limit is that shared words are not the same as shared intent. Two keywords can share words and still want different pages, and two keywords can want the same page while sharing few words. The sections below show how to use the shared word groups as a first sort, and then apply an intent check, using search results, before you commit a group to a page. That two stage approach, group by words first, check by intent second, is the correct way to use a grouping tool, and it is how this guide uses the observed examples.
Why grouping matters for rankings
Grouping matters for four reasons, and each one is a practical problem you have probably met, even if you did not name it grouping at the time. The first is keyword cannibalisation. Cannibalisation happens when two or more pages on your own site target the same keyword and the same search intent, so they compete instead of reinforcing one page. Google generally shows one URL per site prominently for a query, so your backlinks, internal links and engagement data get split across URLs instead of concentrating on one. Grouping prevents that at the planning stage, by assigning every keyword to exactly one page before any page is written. If two keywords are in one group, they get one page. If they are in different groups, they get different pages, with different intents, and the risk of competition falls.
Google own documentation supports the consolidation idea, even though Google does not use the word cannibalisation in its guidance. Its page on consolidating duplicate URLs explains that consolidation helps search engines gather the signals for individual URLs, such as links to them, into a single preferred URL, that it simplifies tracking for a piece of content, and that it avoids spending crawl time on duplicate pages. The Google Search Central page on consolidating duplicate URLs, read on 22 November 2026, is the authoritative source for that reasoning. Grouping is the planning version of the same idea. Instead of consolidating pages after they compete, you group keywords before they are written, so the competing pages are never created.
The second reason is topical authority. Topical authority is the idea that a site which covers every part of a topic, each part on its own dedicated page, is recognised as a reliable source on the overall subject. Semrush puts it plainly in its clustering tool guide. When your site covers all the major subtopics within a theme, each subtopic on its own dedicated page, search engines recognise it as a reliable source on the overall subject. Grouping builds that coverage deliberately. Each group becomes a subtopic page, the groups together cover the theme, and internal links between them show how the parts fit. A site built that way is easier for search engines to understand than a site where coverage is accidental, with gaps where no group was ever planned and overlaps where two groups were never separated.
The third reason is content efficiency. Writers produce fewer, stronger pages when they work from groups, instead of churning out near duplicates that dilute each other. A group gives a writer a primary keyword, secondary keywords, and the questions the page must answer, so the brief starts from the best keywords rather than from guesses. That also helps content read more naturally, because the writer covers subtopics the group implies, rather than repeating one phrase to hit a count. The fourth reason is cleaner internal linking. Every group gives you a natural structure to link against, with a pillar page linking to its supporting pages and supporting pages linking back, and related groups linking sideways where a reader would benefit. That structure is far easier to build when the groups exist before the pages do.
There is also a longer term benefit that Ahrefs notes in its clustering guide. Clustering helps you understand search intent better, because the keywords in a cluster reveal what searchers actually want. Its example uses a coffee maker cluster, where keywords about taste, home use and value together tell the writer that searchers want recommendations for home use that make tasty coffee and will not break the bank. That insight comes from reading the group as a whole, not from any single keyword in it. The same reading skill applies to the grouping tool outputs in this guide. Read each group as a statement about what the searcher wants, not just as a list of phrases that share words.
How grouping works: three methods
Grouping and clustering tools use three broad methods, and knowing which method produced a group tells you how much to trust it. The first method, and the most reliable for ranking decisions, groups by search results overlap. Two keywords go in the same cluster when Google returns substantially the same top ranking pages for both. Ahrefs explains the logic in its clustering guide. If Google ranks the same pages for many keywords, they have the same or similar intent and should generally be targeted on the same page. Semrush Keyword Strategy Builder uses the same logic. It compares the top 10 organic results for each keyword, and when two keywords return similar URLs, they get grouped into a cluster. Semrush states explicitly that this separates SERP based clustering tools from tools that rely on word similarity alone, because the groups reflect how search engines actually organise results, not just which keywords look alike.
The second method groups by meaning. Semantic and natural language methods group keywords by topical similarity, using language models rather than live search data. This method is faster, and it works where search data is thin, such as for new or low volume queries with no stable results page. Its weakness is that meaning and intent are not the same thing. Two phrases can mean similar things while searchers want different outcomes from them, and a meaning based group can merge pages that should stay apart. The third method groups by shared words and phrases. This is term clustering, or pattern matching. It groups keywords that share literal words or substrings, such as bucketing everything containing a phrase together. Ahrefs describes term clustering as grouping by common words or phrases rather than search result similarity, and notes that it is useful for understanding trends and finding niches. Its coffee example shows term clusters for hotels, including near, beach, city, downtown and airport, which immediately reveal that many people search for hotels near places. A term method is transparent and fast, and it is the method the LinkTrixx tool uses.
Each method has a place, and the best plans combine them. Use shared word grouping to sort a raw list quickly and to spot trends. Use search results overlap to decide the borderline cases, where shared words suggest one page but intent may suggest two. Use meaning based grouping to fill gaps where no search data exists yet. A practical rule from the overlap literature is that four or more shared URLs in the top 10 results suggests one page, while zero or one shared URL suggests separate pages, with two or three shared URLs as a judgement call. That rule is a working rule from clustering guides, not an industry standard, and different tools set the bar in different places. The principle behind it is sound even where the exact number is debated. Let Google own ranking evidence decide the borderline, because Google is the system that will rank the page.
Ahrefs adds a useful shortcut for checking a single pair of keywords without running a full clustering job. Enter one keyword in Keywords Explorer, scroll to the SERP overview, use the Compare with option, enter the second keyword, and read the SERP similarity score out of 100. A high score means cluster, a low score means do not cluster, and a middle score means take your best guess. That manual check is the right tool for the handful of borderline pairs in any grouping exercise, and it costs far less time than clustering a whole list twice. It is also the check this guide recommends after you run the LinkTrixx grouping tool, for any group where the shared words look convincing but the intent is unclear.
How the LinkTrixx keyword grouping tool groups
The free tool is at the Keyword Grouping Tool page. The page title is Free Keyword Grouping Tool, the badge reads Free tool, No signup, and the heading reads Free Keyword Grouping Tool. The introduction states the job in one sentence. Paste a keyword list and watch similar keywords group together by the words they share. The page states that the tool is free, needs no signup, and that nothing leaves the browser. The field is labelled Your keywords, with the helper text One keyword per line. Groups update live as you type. The output heading reads Grouped keywords, and the empty state reads Paste keywords on the left to see them grouped here. There are three buttons, Load sample, Copy and Download. Those details come from a logged out observation of the live tool page on 22 November 2026, and they are the only behaviour this guide claims for the tool.
The grouping rule is stated in the tool page frequently asked questions section, and it is the key to reading every output correctly. The tool looks for words and two word phrases that keywords share. The most specific shared phrase wins, so link building services and link building agency group under link building, not under services. Unmatched keywords go to a group named Ungrouped. That most specific phrase rule is what keeps the groups useful. Without it, a generic word such as services would swallow keywords from unrelated services into one meaningless group. With it, the group name tells you the shared idea that actually binds the keywords together, and the generic word stays in the background where it belongs.
The five steps listed on the tool page are Paste keywords one per line, Watch groups form live, Review Ungrouped keywords, Copy or download, and Build a page per cluster. The Copy button changes its label to Copied briefly. No separate toast message was observed for Copy in the page content during the observation. Download showed no separate message in page content during the observation. The download file name, stated in the tool page text, is grouped-keywords.txt. This guide does not claim a toast message exists for either button, because none was observed. The distinction matters, because a guide that invents a confirmation message will confuse a reader who looks for it and does not find it. What the reader will see, on the evidence of the observation, is the Copy label changing briefly, and a downloaded file, with no separate message recorded in the page content for Download.
Nothing in the live observation showed search volume, keyword difficulty, search intent labels, or search results data on the tool page. This guide therefore does not describe the tool as a keyword research tool, a rank tracker, or a clustering tool based on search results. It is a grouping tool based on shared words, and its value is in the first sort of a list, the visible group names, and the Ungrouped review step, which forces you to decide what to do with keywords that share no phrase with any other keyword. That decision, which the next sections work through, is where much of the planning value sits.
Worked example one: the sample list
The Load sample button fills the tool with five keywords, in this order. Link building services, link building agency, guest posting services, guest posting agency, and seo audit. The observed output summary reads 3 groups, 5 keywords. The groups, in the order observed, are as follows.
- guest posting, 2 keywords: guest posting services, guest posting agency
- link building, 2 keywords: link building services, link building agency
- Ungrouped, 1 keyword: seo audit
Read that output slowly, because it teaches the whole method in miniature. The first group is named guest posting, not services, because guest posting is the most specific shared phrase. The two keywords in it share that phrase, and they differ only in the final word, services or agency. A reader searching for either phrase wants a provider of guest posting, and one page that covers guest posting services and explains the agency behind them can satisfy both. The second group works the same way. Link building services and link building agency share the phrase link building, and one page about link building services, with a clear account of the agency offering them, can satisfy both. The most specific phrase rule is visible in both groups. Services appears in two keywords, but it does not become a group, because it is generic and it would merge guest posting with link building, which are different services with different buyer questions.
The Ungrouped group holds seo audit, because seo audit shares no word or two word phrase with any other keyword in the list. That does not mean seo audit is unimportant, and it does not mean it should be deleted. It means that, on the evidence of this list alone, seo audit belongs to no group yet. It may deserve a page of its own, it may belong with keywords that are not in this list, or it may fold into a nearby page as a secondary topic. The tool flags the decision. It does not make the decision for you. That is the correct division of labour between a grouping tool and a planner, and it is why the tool page lists Review Ungrouped keywords as a separate step, rather than hiding unmatched keywords or forcing them into the nearest group.
Notice also what the sample does not show. It does not show search volume for any keyword, it does not rank the groups by opportunity, and it does not tell you which group to build first. Those judgements need research data from elsewhere, such as volume and difficulty from a keyword research tool, and business judgement about which service matters most to you. The grouping tool gives you the structure. Research data and business judgement give you the order. Confusing the two, by treating the first group in the output as the first page to build, is a mistake. The output order follows the grouping logic, not a priority ranking.
Worked example two: the test list
A second list was tested on the live tool during the same observation, with six keywords, in this order. Link building, guest posting, link building services, guest posting services, seo audit, and seo audit checklist. The observed output summary reads 3 groups, 6 keywords. The groups, in the order observed, are as follows.
- guest posting, 2 keywords: guest posting, guest posting services
- link building, 2 keywords: link building, link building services
- seo audit, 2 keywords: seo audit, seo audit checklist
There is no Ungrouped group in this output. Every keyword found a group, because every keyword shares a phrase with at least one other keyword. The change from the sample is instructive. In the sample, seo audit stood alone because no other keyword shared its phrase. In the test list, seo audit checklist shares the phrase seo audit with seo audit, so the two form a group named seo audit, and the Ungrouped group disappears. Adding one keyword changed the structure, which is exactly how grouping should behave. Groups are not fixed categories. They are the result of the list you feed in, and a list that is missing a bridge keyword will strand a phrase that a fuller list would group.
The test list also shows the most specific phrase rule working at a different level. Link building, on its own, is both a keyword and a group name. Link building services joins it, because the two share the phrase link building. Guest posting and guest posting services behave the same way. In each case, the short keyword is the broad term and the longer keyword is the specific service phrasing, and one page can cover both by leading with the broad term and answering the specific phrasing in a section. That page plan is stronger than two thin pages, one for the broad term and one for the service phrasing, because the broad page would be vague and the service page would repeat the broad explanation. One page, built from the group, avoids both problems.
Use these two examples as your reference for what the tool actually outputs, and do not expect outputs this guide has not quoted. If you paste a longer list, you will see more groups, larger counts, and a larger Ungrouped group where phrases stand alone. The summary will state the number of groups and the number of keywords, in the same form as the examples above. The group names will be the most specific shared phrases the tool finds. Anything beyond that, such as a ranking of groups, a difficulty score, or a volume figure, is not part of the observed behaviour, and you should not wait for it or assume it will appear.
From groups to pages: a step by step method
Turning groups into pages takes five decisions for each group, and the method below works through them in order. Work group by group, and do not start writing until all five decisions are made for the group in front of you. A group with an undecided primary keyword, an undecided intent, or an undecided relationship to existing pages will produce a draft that drifts, however well it is written.
Step one, name the primary keyword. In most groups, the primary keyword is the group name itself, or the shortest, clearest keyword in the group. In the sample, link building is the natural primary keyword for its group, and guest posting is the natural primary keyword for its group. In the test list, the same names work, because the short keywords are present in the list. If a group has no short keyword, choose the clearest phrasing as primary and treat the rest as secondary. The primary keyword goes in the title, the main heading, the opening paragraph and the URL slug, following the on page practice Semrush describes in its clustering guide, which is to include the primary keyword in the slug, title tag, meta description and main heading, and to mention it within the first paragraph.
Step two, list the secondary keywords and assign each one a home inside the page. A secondary keyword should map to a heading, a section, a list, or a frequently asked question, not to a passing mention. If a secondary keyword has no natural home, ask whether it really belongs in the group, or whether the group is too broad. In the link building group, services and agency are not separate pages. They are answers inside one page. The services phrasing belongs in a section that explains what is offered, and the agency phrasing belongs in a section that explains who offers it and how the work is done. That mapping is the brief, and a writer who receives it can write with confidence, while a writer who receives only a keyword list must guess.
Step three, check intent with search results. Open Google for the primary keyword and for each secondary keyword, and compare the pages that rank. If the same pages rank for both, the group is sound. If different pages rank, with different kinds of content, the group may need splitting. Use the Ahrefs shortcut described above for a single pair, by comparing SERP similarity for the two keywords. For a whole group, scan the top results for the primary keyword and ask whether a page that answers the primary keyword would also satisfy a searcher who typed each secondary keyword. If the answer is yes for all of them, keep the group. If the answer is no for one of them, split that keyword into its own group or move it to a different group where the intent matches.
Step four, check existing pages before you create a new one. Semrush clustering guidance makes this decision explicit. If you already have a page that corresponds to a cluster, optimise the existing page. If you do not, create and optimise a new page. Many sites already have a page that partially covers a new group, and improving it is faster and safer than publishing a second page that will compete with it. Search your own site for the primary keyword, read the page you find, and decide whether the group extends that page, splits off a new page, or replaces a weak page that should be consolidated. Record the decision in the group plan, so the writer does not have to rediscover it.
Step five, plan the internal links before you write. The pillar page for a theme links to each group page, each group page links back to the pillar, and related group pages link sideways where a reader would benefit. The keyword combination generator guide and the keyword combination use case guide show how combination lists feed grouping, and the combination comparison guide shows how free and paid tools divide the same work. Link the new page to the most relevant older pages in the same theme, and plan one or two older pages that will link to the new page once it is live, inside a relevant paragraph, so the new URL is discovered through pages search engines already know. That backward linking step is part of publishing, not an afterthought, and it should be planned at the grouping stage, when the relationships are fresh in mind.
What to do with Ungrouped keywords
Ungrouped keywords need a decision, and there are only three sensible decisions. The first is to give the keyword a page of its own. That is right when the keyword names a distinct service, topic or intent, and when you can write a substantial page about it without repeating an existing page. In the sample, seo audit could deserve its own page, because an audit is a distinct service with its own process, deliverables and buyer questions, even though it shares no words with link building or guest posting. A standalone page is justified by the topic, not by the grouping output, and the grouping output merely flags that no other keyword in the list belongs with it.
The second decision is to fold the keyword into a nearby page as a secondary topic. That is right when the keyword is a subtopic, a variant, or a question that a broader page should answer, even though it shares no phrase with the group. An audit section inside a broader services page, for example, might be the right home for a keyword that is too narrow for its own page. The test for folding is the same as the test for grouping. Would a reader who searched for the folded keyword be satisfied by the page it joins. If yes, fold it, and give it a heading or a section so it is easy to find. If no, do not fold it to save effort, because a dissatisfied reader and a confused search engine are the result.
The third decision is to set the keyword aside for now. That is right when the keyword is vague, off topic, or a phrase you cannot yet write about well. Setting aside is not deleting. Keep the keyword in your list, with a note on why it was set aside, and revisit it when the list grows or the site grows. A fuller list may group it, as the test list grouped seo audit once seo audit checklist was added. A new service, a new page, or a new theme may give it a home. The worst decision is the fourth one, which is to force an Ungrouped keyword into the nearest group because an empty Ungrouped section feels untidy. Forced grouping creates the very competition and confusion that grouping exists to prevent, and it hides the decision instead of making it.
Review the Ungrouped group every time you run the tool, not just the first time. Lists grow, services change, and a keyword that stood alone last month may group cleanly this month once bridge keywords are added. The tool page lists Review Ungrouped keywords as a step for every run, not as a one off task, and that repetition is deliberate. Ungrouped keywords are often the most interesting part of a list, because they mark the edges of your current coverage, where new pages and new services are waiting to be planned.
When shared words mislead
Shared word grouping is transparent, which makes its failures visible, and that visibility is an advantage if you use it. The classic failure is the chocolate cake example from the Ahrefs clustering guide. The keyword best chocolate cake recipe with coffee was grouped under the topic of chocolate cake by both Ahrefs and a dedicated clustering tool, because there is some overlap in results. Yet Google ranks a mix of pure chocolate cake recipes and chocolate cake recipes with coffee for that keyword, and the top pages for the two topics differ sharply in authority, with average Domain Rating and linking domain counts far lower for the coffee variant. Ahrefs concludes that clustering, as the tools suggested, would probably be a mistake, because ranking for the coffee variant looks easier than ranking for the pure recipe, and a combined page would miss that distinction. The lesson is general. A group that looks right by words or by partial result overlap can still be wrong by intent and by competition, and the only way to catch it is to look at the results and the metrics behind them, not just at the group name.
Semrush gives a subtler example in its clustering guide, using two phrases about apple cider vinegar and dog shampoo. One searcher wants ideas on how to use apple cider vinegar to bathe a dog, an informational intent, with a featured snippet and People Also Ask results. The other wants commercial dog shampoo that contains apple cider vinegar, a commercial intent, with product listings and popular products results. The phrases share words, and a naive grouping would merge them, but Google understands the difference, and the results show it. The searcher intents are different, the result types are different, and the pages that satisfy them must be different. Always check intent for groups where the shared phrase is a product, an ingredient, or a modifier that can change the job the searcher wants done.
A third failure mode is the generic word trap, which the most specific phrase rule in the LinkTrixx tool is designed to avoid. Words such as services, agency, best, free, and online appear in many keywords, and grouping by them would create huge, meaningless groups. The tool avoids that trap by letting the most specific shared phrase win, so services stays in the background and link building or guest posting names the group. When you group by hand, apply the same rule. Name each group by the most specific phrase that binds its keywords, and be suspicious of any group named by a generic word. If you find yourself with a group called services, or best, or guide, split it by the specific phrases inside it, and let each specific phrase become a group of its own.
A fourth failure mode is grouping by words when the real distinction is audience. Link building for SaaS companies and link building for law firms share the phrase link building, and a shared word group would merge them, but the buyer questions, the publisher types, and the examples that persuade each audience are different. The long tail keywords guide and the long tail generator guide show how longer phrasings reveal audience and intent, and those phrasings should drive the split, even when the shared words suggest one group. Audience is an intent signal, and intent outranks wording whenever the two disagree.
Grouping for campaign planning and outreach
Grouping is not only for blog pages and service pages. It is also a planning step for campaigns, where the groups become angles, landing pages and outreach lists. The guide Digital PR Campaign Ideas That Earn Backlinks works through campaign ideas that earn coverage, and grouping helps you plan those campaigns without repeating the same angle under different names. Group the keywords and phrases that describe a campaign theme, such as the service, the audience, the data angle and the story hook, and each group becomes a distinct campaign idea with its own landing page and its own pitch. If two campaign ideas share a group, they are the same campaign with different wording, and running both will split coverage and confuse journalists. If they sit in different groups, with different intents and different audiences, they are genuinely distinct campaigns, and they can run in sequence without cannibalising each other.
The same logic applies to outreach. The guide LinkTrixx Digital PR Outreach explains how outreach pitches are built around a clear topic and a clear reason to cover it. A pitch that buries its subject, or that mixes two subjects, will be ignored, however good the underlying story is. Grouping the phrases for a pitch, then checking the keyword position inside the pitch copy with the position checker guide and the position checker use case guide, is a short chain that prevents both errors. Grouping decides the subject. The checker verifies that the subject appears early in the copy. The free checker versus paid tools comparison, published the same day as this guide, explains why the free checker is the right tool for that copy check, and why a paid rank tracker is the wrong tool for it, because a tracker reads Google results and never reads your pitch.
For service businesses, grouping also clarifies the offer before any campaign begins. The sample groups in this guide, link building and guest posting, are two services with different buyer questions, and a campaign that merges them into one vague offer will answer neither buyer well. Separate groups, separate pages, and separate campaign angles, linked together where a buyer might want both, is the structure that serves readers and search engines alike. That structure is also the structure that makes internal linking natural, because the links follow the buyer journey, from a broad service page to a specific campaign page, to a guide that explains the method, and back again.
Grouping at scale: from a handful of keywords to a full site plan
The method above works for five keywords, and it scales to thousands, with one change. At scale, you do not check every pair by hand. You group in passes, and you spend human judgement on the borderline groups, not on the obvious ones. The first pass is shared word grouping, using a tool like the LinkTrixx grouping tool for a first sort, or a spreadsheet sort for a list that is too long to paste in one run. The second pass is an intent check, using search results overlap, for the groups where the shared words are suggestive but the intent is unclear. The third pass is a business check, where you rank the groups by combined volume, difficulty, resources and business goals, following the prioritisation factors Semrush lists in its clustering guide, which are resources, combined volume, personal keyword difficulty, efficiency of optimising an existing page versus creating a new one, and business impact.
At scale, the Ungrouped group grows, and that is expected. A list of 2,000 keywords will contain many phrases that share no phrase with any other keyword in the list, including long, specific questions, brand names, and emerging terms. Do not force those phrases into groups to tidy the output. Keep them in a separate list, review them for distinct intents, and let the strongest become single keyword pages or sections, while the weakest are set aside. A clustering guide that collapses 200 keywords into 15 to 20 clusters in minutes still requires human review of search overlap before publishing, and the same caution applies at any scale. Automation sorts. Humans decide. A plan that reverses that order, by publishing every automated group without review, will publish borderline groups that should have been split, and split groups that should have been merged.
Keep a keyword map as you scale. A keyword map is a simple record that states, for each keyword, which page owns it. It prevents the commonest scaling error, which is for two writers, or the same writer months apart, to target the same keyword on two pages without realising it. The map is the long term memory of your grouping decisions. When a new keyword arrives, check the map before you write. If the keyword belongs to an existing group, add it to that page. If it belongs to no group, create a new group and a new page, or fold it into an existing page with a new section. The map, the groups and the pages stay in step, and cannibalisation is prevented by process, not by luck.
Grouping at scale also needs a rule for when to stop splitting. A group that is too broad produces a page that tries to answer everything and satisfies no one. A group that is too narrow produces many thin pages that compete and dilute each other. The test for the right size is the same test used for a single group. Can one substantial page answer every keyword in the group well, with a heading or a section for each secondary keyword, without becoming a list of unrelated answers. If yes, the group is the right size. If the page would need unrelated sections that no reader would read together, split the group. If two groups would need pages that repeat the same introduction, the same explanation and the same examples, merge them. Size is judged by the page, not by the count of keywords.
Grouping, internal linking and topical authority
Groups and links belong together, because a group plan is also a link plan. Once groups are decided, the link structure follows naturally. The pillar page for a theme links to each group page, using the group primary keyword as the link text where it fits naturally in a sentence. Each group page links back to the pillar, and to the neighbouring group pages that a reader would want next. A reader who lands on the link building group page and wants guest posting next should find that link in the text, not only in a menu. A reader who lands on a narrow group page and wants the broader theme should find the pillar link early, so the context is clear.
That structure is what Semrush describes when it says the end goal of clustering is a dedicated page for each keyword cluster, with related pages connected through internal links. It is also what builds topical authority in practice, rather than in theory. Topical authority is not a badge awarded for using a phrase. It is the result of covering a theme completely, with each subtopic on its own page, linked in a way that shows the relationship between the parts. A site with complete coverage and clear links is easier for search engines to understand, easier for readers to navigate, and easier for writers to extend, because the next group has an obvious home in the structure.
Link the new page from older pages in the same theme, inside relevant paragraphs, when it is published. That backward linking step, planned at grouping time and executed at publish time, helps search engines discover the new URL through pages they already know. It also helps readers, who may arrive at an older page and need the newer, more specific page that the group plan created. Do not add new links only in a list at the end of an older page. Place them where the older page discusses the topic the new page covers, so the link serves the reader at the moment the question arises. The grouping plan should name those older pages, so the linking step is not left to memory after publishing.
Common grouping mistakes
The first mistake is grouping by words alone and never checking intent. Shared words are a clue, not a verdict. Always check search results for borderline groups, and split groups where intent differs, even when the wording is similar. The second mistake is the opposite, which is to create one page per keyword. That produces thin pages, cannibalisation, and a site that is hard to navigate. Group first, and let one strong page serve a whole group. The third mistake is to ignore the Ungrouped group, either by deleting its keywords or by forcing them into the nearest group. Ungrouped keywords are decisions, not leftovers, and they deserve the three way decision described above, own page, fold in, or set aside with a note.
The fourth mistake is to group without checking existing pages. A new group that duplicates an existing page will compete with it, however well the new page is written. Always search your own site before you create a page for a group. The fifth mistake is to treat group order as priority order. The order in a tool output follows grouping logic, not opportunity. Rank groups by volume, difficulty, resources and business impact, using research data from elsewhere, before you decide what to build first. The sixth mistake is to expect a grouping tool to supply data it does not supply. The LinkTrixx tool shows groups, counts and group names, based on shared words. It does not show volume, difficulty, intent labels or search results. Use a research tool for those data, and use the grouping tool for the structure. Expecting one tool to do both jobs will leave you waiting for data that never arrives, while the structure you already have goes unused.
The seventh mistake is to group once and never revisit the groups. Keyword lists change, services change, and search results change. A group that was sound last year may need splitting when a new intent emerges, and two groups that were separate may merge when Google starts returning the same pages for both. Revisit the keyword map when you add a significant batch of keywords, when you launch a new service, or when Search Console shows two pages drawing impressions for the same query. Grouping is a living plan, not a one off exercise, and the sites that keep it current are the sites that avoid the slow drift into overlap that grouping exists to prevent.
A complete workflow: grouping to published page
Put the whole method together and it runs as one workflow, from research to published page, using free tools for the steps they serve well and paid or research tools only where their data is needed. Research first, using a keyword research tool or Google Keyword Planner to build a list with volume and difficulty data. Group second, using the Keyword Grouping Tool for a first sort by shared words, and checking borderline groups by search results overlap. Plan third, by naming a primary keyword and secondary keywords for each group, checking existing pages, and planning internal links, including backward links from older pages. Brief fourth, by turning each group into a brief with headings mapped to secondary keywords, questions the page must answer, and the internal links it must carry. Write fifth, with the brief in front of you, covering the group completely and naturally, following the people first standard in the Google Search Central guidance on helpful content, which asks whether content is substantial, complete, original and satisfying for the reader.
Check sixth, using the free checker to verify keyword placement inside the draft, following How to Use a Keyword Position Checker, and using the Keyword Density Checker guide if you want a density diagnostic as well. The long tail comparison guide and the density comparison guide show how the free tools in this chain compare with paid suites, and the combination comparison guide does the same for combination building. Publish seventh, with a clear title, meta description, headings, and the internal links planned at the grouping stage, including the backward links from older pages. Monitor eighth, in Search Console, by query and by page, watching for overlap between pages, and revisit the grouping decision when overlap appears. That loop, research, group, plan, brief, write, check, publish, monitor, is the whole discipline in one paragraph, and it works because each step feeds the next and no step is skipped.
For service pages, the same workflow applies, with one addition. A service page must answer buyer questions, not only keyword questions. Link the group plan to the service offer, so the page explains what is offered, how it works, who it is for, and what happens next, in language a buyer can act on. The service pages for link building services, SaaS link building, white label link building and guest posting are the destinations that group plans for service keywords should lead to, where the group describes a service rather than a guide topic. A grouping plan that ends at a guide page when the searcher wants a service, or at a service page when the searcher wants a guide, has grouped by words and missed the intent, and the intent check in step three exists to catch exactly that error.
Frequently asked questions
What is a keyword grouping tool?
A keyword grouping tool sorts keywords that belong together into groups, so each group can be targeted by one page. The LinkTrixx keyword grouping tool groups by the words and two word phrases keywords share, with the most specific shared phrase winning as the group name, and unmatched keywords placed in a group named Ungrouped. A group gives a page one primary keyword and a set of secondary keywords, which prevents thin pages, reduces competition between your own pages, and turns a flat keyword list into a plan for pages.
How does the LinkTrixx Keyword Grouping Tool decide a group?
The tool looks for words and two word phrases that keywords share, and the most specific shared phrase wins as the group name. Link building services and link building agency group under link building, not under services. Unmatched keywords go to a group named Ungrouped. Groups update live as you type, with one keyword per line in the field labelled Your keywords. The sample list of 5 keywords gives 3 groups, 5 keywords, and the test list of 6 keywords gives 3 groups, 6 keywords, with no Ungrouped group.
Is keyword grouping the same as keyword clustering?
Grouping and clustering are closely related, but clustering is usually more rigorous. Grouping by shared words, as the LinkTrixx tool does, sorts by wording. Clustering by search results overlap groups keywords when Google returns the same top pages for both, which is a stronger test of shared intent. Semrush and Ahrefs both describe clustering as grouping by shared intent so one page can target the whole group. Use shared word grouping as a first sort, then check borderline groups by search results overlap before you commit a group to a page.
What should I do with Ungrouped keywords?
Decide for each Ungrouped keyword whether it deserves a page of its own, whether it should fold into a nearby page as a secondary topic, or whether it should be set aside for now with a note. A distinct service or topic with enough substance for a full page deserves its own page. A narrow subtopic that a broader page should answer should fold in, with a heading or section. A vague or off topic phrase should be set aside and revisited when the list grows, because a fuller list may group it, as the test list grouped seo audit once seo audit checklist was added. Do not force Ungrouped keywords into the nearest group.
Can one page rank for all the keywords in a group?
Yes, when the group shares one intent and the page covers the group completely. Ahrefs found, in a study of 3 million searches, that the average top ranking page also ranks in the top 10 for nearly 1,000 other relevant keywords, with a median of around 400, which shows how many related terms one strong page can cover. That result depends on the group being sound. If the keywords in a group want different outcomes, one page will not satisfy them all, and the group should be split. Check intent with search results before you assume one page will cover a group.
Is the LinkTrixx Keyword Grouping Tool free?
Yes. The tool page carries the badge Free tool, No signup, and states that the tool is free, needs no signup, and that nothing leaves the browser. The observation on 22 November 2026 recorded no account step and no payment step. The Copy button changes its label to Copied briefly, with no separate toast message observed in page content, and Download showed no separate message in page content. The download file name stated on the tool page is grouped-keywords.txt.