SEO is not just about creating pages that mention specific keywords. It’s about understanding the distinct topics that keywords or groups of keywords represent and uncovering the users’ search intent in making those searches. This is where keyword clustering comes in, helping you organise the content on your website into logical sections, as well as being able to quantify your SEO investment.
Read on to understand my approach to keyword clustering, how it works, how it can help you identify what content your website needs to get in front of your ideal customers, and how to use it to create a traffic and revenue model to help you prioritise and decide what content to invest in. Plus, learn about AI tools to help you streamline clustering, and download my FREE template to help you get started!
👉 Get the Template (Go to File > Make a Copy so you can edit!)
What is keyword clustering?
Keyword clustering is the process of grouping related keywords into distinct pages based on search intent and mapping out groups of pages covering specific topics. Clustering is important because it helps ensure you have one page per distinct topic, making your site easier to use and improving SEO performance.
Let’s take a look at an example, using the top personal loan keywords in the UK and search volume data taken from SEMrush:
| Keyword | Monthly Search Volume |
| Personal loan | 27.1K |
| Personal loans | 27.1K |
| Personal loan calculator | 14.8K |
| Personal loans UK | 6.6K |
| Tesco personal finance loan | 5.4K |
| Low interest personal loans | 4.4K |
Let’s imagine we’re building a personal finance website that offers lending products, how might we group these keywords into distinct pages?
The first two keywords, personal loan and personal loans, are the same – They serve the same purpose and the identical search volume is an indicator that they are indeed the same thing. These keywords can clearly be grouped onto the same page.
If our website is UK-specific, we could add the personal loans uk keyword to the same URL, since the page is only applicable to the UK anyway. Alternatively, if it was a multinational website, we might want a separate UK-specific personal loans page alongside other loan pages for other specific countries.
Next, the personal loan calculator keyword. This could be grouped in with these three, provided we can provide a loan calculator tool on the landing page for our users. Or we might decide that a separate calculator tool is a better experience for the user and spin it off as a separate page.
The Tesco personal finance loan keyword probably doesn’t make sense to be grouped in with the four non-brand keywords we’ve just looked at. Our personal loans page needs to cater for non-brand searches, so the Tesco keyword should probably be mapped to a page specifically about Tesco personal loans, if it’s indeed relevant to create a page about this provider.
Finally, we have the low interest personal loans keyword. We might see this as a variation of the existing non-brand keywords we’ve looked at. After all, isn’t everyone after a low-interest loan? Alternatively, we might interpret this as a distinct topic from our general personal loans page.
The way to get a feel for this is to compare the search results between these keywords and assess the similarity of the results. If the results between the two are very similar, then we may conclude that low interest personal loans is the same topic as personal loans, and group this onto the same page.
We can now map our six keywords to two or three distinct pages on our site based on our decisions:
| Keyword | Monthly Search Volume | URL path |
| Personal loan | 27.1K | /personal-loans |
| Personal loans | 27.1K | /personal-loans |
| Personal loan calculator | 14.8K | /personal-loan-calculator |
| Personal loans UK | 6.6K | /personal-loans |
| Tesco personal finance loan | 5.4K | /tesco-personal-loans |
| Low interest personal loans | 4.4K | /personal-loans |
Simple, right? We now have an idea of the pages we’ll need to create if we’re to compete for these keywords in organic search.
However, if you want to build a large site that’s a real authority on this topic, you’ll need to assess many more keywords and potential pages.
How to find keywords
There are several keyword research tools out there that you can use to source keywords:
- KeywordsEverywhere: This is one of my favourite keyword research tools because it injects search keyword ideas directly into your search results. It’s also very cost-effective, making it the perfect tool for marketers on a tight budget.
- SEOmonitor: SEOmonitor’s keyword research tool aggregates misspellings and close variants, and helps to filter out irrelevant keywords. I find that this helps to reduce the complexity of keyword clustering projects.
- SEMrush: SEMrush is almost ubiquitous in the SEO industry. If you already have a SEMrush account, the Keyword Magic tool is the place to go for keyword data.
Whatever tool you choose, the first step is to source keywords relevant to the main topic you want to research.
Expanding on the personal loans topic example, here’s how you can approach it:
- Search for personal loans in the keyword tool and export all the relevant keywords provided, excluding any that are clearly unrelated. Consider filtering out low-search volume keywords, for example, those with fewer than 30 searches per month. This helps to reduce complexity, and we can account for long-tail searches later in our SEO strategy process.
- Think of 4-5 or more sub-topics of personal loans that potential customers might search for. For our example, these might include unsecured loans, bad credit loans, personal loan interest rates, or pre-approved personal loans. Repeat and export keyword lists for all the topics you feel are relevant to your customers.
- Consolidate all your keyword lists into one sheet ready for step two.
Clustering keywords manually or with AI tools
After sourcing your keywords, the next step is to cluster them into relevant pages based on the topic, as we did in the example. This works for smaller sets of keywords but becomes very time-consuming if you’re trying to cluster a large topic.
Recently, I’ve been using keywordinsights.ai to help streamline the process. Keyword Insights includes a keyword clustering tool that groups keywords into pages based on Natural Language Processing and pulls in Google search results so that clusters are based on what’s ranking.
Some people also use general LLMs like ChatGPT or Gemini, but beware – these tools can provide incomplete or inaccurate results because they lack the context of the keyword clustering use case, such as by not having an awareness of other pages that are ranking in search. It’s a better idea to invest in keyword clustering-specific tools or do it manually.
If you’d prefer to map your keywords manually, you can use the Keyword List tab in our free template:
- Paste your keywords into column B and search volumes into column E
- If you have it, paste your current ranking and ranking URL data in columns G and H. This will help you determine if you have relevant pages for the topics you want to focus on.
- Add the names of relevant topic groupings in the Topic List tab, then assign these to relevant keywords in column A
- Map related keywords to pages using column C (page name)
- Read on below for suggestions on how to map your pages to specific URLs.
If you choose KeywordInsights.ai, then steps 1-4 will be automated. You can export the output to a CSV or Google Sheets, then paste the relevant data into the relevant columns of the template. Be sure to manually review the page recommendations and decide if you agree with the choices. If not, now’s the time to tweak the page and topic selection.
Map topic clusters to URLs and page types
After defining your page names, you can define relevant URLs in the Keyword List tab. While you can define your URLs at the same time as the keyword clustering stage, when it comes to URL structure, you likely have several options on how you could organise your site.
Holding back and doing this at the end means you can see the bigger picture of exactly what pages you need. This can help spark ideas about the information architecture of your site, meaning how you should structure your pages into relevant folders. It also means you can correct any issues with the site’s current structure, for example by creating new pages on certain topics that weren’t appropriately covered before.
Often, I’ll draw out a site architecture diagram to help me visualise the site’s information architecture before assigning individual URLs. This helps me ensure my selections are logical and relevant to the overall goals of the site.
Once you’ve mapped your URLs on the Keyword List tab, you can move over to the Page List tab, which should now be automatically populated based on your choices. In column D, you can define which pages are new or existing, helping you to clearly identify the amount of net-new content you need to produce vs existing pages that are either working well or may require optimisation.
Next, define what type of page experience is required for each URL, to suit the the search intent. The template includes several common page types for you to use, or you can create your own. Here’s what I typically use:
- Articles: For blog posts or other long-form content serving information search intent
- Landing Pages: For bottom-of-funnel pages geared around lead generation and serving a transactional/commercial search intent
- Product Pages: For eCommerce product pages or lead generation pages that need to be distinct from other types of landing pages
- Category Pages: For eCommerce category pages serving non-brand transactional/commercial keywords
- Tools: For calculators and other interactive tools that help users in their buying journey
Factor in traffic goals
You should now have a list of keywords clustered into topics and assigned to relevant URLs on your site and knowledge of what pages need work and which don’t. However, this isn’t the only thing you can do with keyword clustering. You can also use this process to create a model for potential traffic growth, followed by revenue estimates, helping you understand the potential impact of embarking on your content strategy.
Using the template, you can estimate three potential traffic goals on a per-page basis. These traffic goals are based on taking the total search volume for each page and multiplying them by a Click Through Rate (CTR) estimate. I typically use these CTR estimates as starting points:
- Baseline: This is intended to estimate the potential traffic if all you did was write the new pages called for in your keyword clustering plan. To estimate this, multiply the search volume by the average CTR in your Google Search Console performance data.
- Headroom: This is to estimate the total size of the market – the maximum organic traffic you are likely to achieve. This is based on a CTR of 20%, which is around what you might expect if your site ranked with a mix of positions 1, 2 and 3 for all your target keywords.
- Intermediate: This is where you can estimate something a bit more adventurous than your baseline, but not as aggressive as the headroom. A good starting point is somewhere halfway between your baseline and the headroom CTR, which can be revised up or down based on any pre-existing research into what other optimisations need to be done, such as fixing technical SEO issues or improving user experience.
These values can be set up in cells B2-4 of the Variables & Summary tab on the spreadsheet, which will automatically populate traffic data cross the traffic summary and Page List tab.
Build in revenue estimates
After setting up your traffic goals, you can add revenue estimates. This step will depend on how mature your web analytics set-up is. For the template, you’ll need the following pieces of data:
- Page conversion rate: The % of sessions on a page turn into conversions, which could be form fills or transactions. The template allows you to set conversion rates for each specific page type, allowing you to factor in that some pages may convert at different rates to others.
- Earnings per Click/Cost per Acquisition: Essentially, how much one conversion worth is to you. For lead generation, this is your average deal value divided by the number of form completions you need to acquire one lead. For eCommerce, this is your average basket value.
Fill out conversion rate values in cells B10-14 of the Variables & Summary tab, and add your average goal value to cell B7. This will automatically populate revenue data on the summaries and Page List tab.
Some websites will cover different topics that may have different conversion rates or goal values. If this is the case for you, consider creating a separate document for each vertical. This will allow you to set a different goal value for each vertical to ensure your revenue data is relevant and realistic.
You’ll also see the summary includes annual traffic and revenue summary as well as month. This is a tip I learnt from Tom Critchlow’s excellent SEO MBA course, which is now sadly closed. Tom’s rationale for incorporating annual data is that it’s a more relevant time span for C-suite executives, who typically use annual or quarterly figures. This is a good way to speak your audience’s language, highlight the potentially long-term nature of your SEO initiative, and of course, make the size of your SEO opportunity look bigger and more enticing compared to using monthly data.
Sense check your work with other stakeholders
The hard work is now done, but chances are you’ll need to convince at least one other person in the business why your new content strategy is a worthwhile investment.
This is now a good point to check your work with other stakeholders and get their input before submitting the final document to the CEO or primary stakeholder. Again, it’s a tip Tom Critchlow advocates and comes from the concept of “pre-wiring” used at consulting firms like McKinsey. It helps ensure all other stakeholders are on board with your recommendations by making them a part of the process.
To get started, consider all the relevant stakeholders who will be affected by your recommendations, and schedule time to review the document independently with each to ensure that their feedback and insights can be incorporated.
Consider some of the following talking points:
- Content goals: How do your recommendations tie into the overall business goals? Consider the page types and topics you’ve chosen
- Site architecture: How do the recommended pages fit into the site?
- Page list: Are all page recommendations relevant or can the list be filtered down to exclude less relevant ones?
- Conversion rate and revenue data: Validate that these data points are realistic, refine with the stakeholder’s input if necessary
Because all the main fields of the template are connected with formulas, you can tweak different values under columns A and B of the Variables & Summary tab, and fields will update.
Next steps
You’re done! The keyword mapping is complete and you now have a way to quantify the annual traffic and revenue that your SEO initiative could generate, grouped by page types and topics.
At this point, you may want to present the data back to your CEO or key stakeholder, if you’re embarking on a purely content-led SEO initiative. Alternatively, you may be aware of one or two technical SEO issues on the site that need to be investigated and resolved to be able to achieve your goals.
If this is you, now’s a good time to evaluate what technical work needs to be done via a technical SEO audit. You can then use your keyword mapping sheet’s revenue projections to help quantify the opportunity of these technical SEO fixes alongside the proposed content work.
Get help and support
Content strategy is our bread and butter at Atkinson Smith Digital. If you’d prefer a content strategy specialist to complete the keyword mapping process explained in this guide so you don’t have to, get in touch with us with details about your project and what support you need.
Advertising Disclosure: We’ve partnered with the KeywordInsights.ai affiliate scheme. If you choose to purchase from KeywordInsights via our link, we may earn a small commission which gets reinvested into producing free guides like this article.