Keyword Clustering: Mapping Many Keywords to the Right Pages

Keyword clustering is the process of grouping search terms that can be answered by the same page, so each page targets one need and no two pages compete. The most reliable way to do it is to compare the search results: if two keywords return mostly the same URLs on Google, they belong together. This guide explains the SERP overlap method step by step, with a worked example for a service business.

Why keyword clustering matters before you write or design

A keyword list on its own does not tell you how many pages to build. Without grouping, teams tend to make one of two mistakes:

• Too many pages. Separate pages for "office fit out Dubai", "office fit out company Dubai" and "office interior fit out" end up thin and nearly identical. They split links and signals, and Google may show none of them prominently.
• Too few pages. One broad service page tries to rank for informational questions, comparisons and location searches at once, and satisfies none of them well.

Google's own documentation notes that when many matching pages are very similar, its systems show only the most relevant ones, and that it generally avoids showing more than two listings from the same site in top results, as set out in the guide to Google Search ranking systems. Building near duplicate pages for every variant works against that.

Keyword grouping by topic versus by results

Keyword grouping by topic, done by eye, puts terms together because they sound related. That is a reasonable first pass but it is often wrong. "Website maintenance" and "website support" sound alike, yet the results may show different kinds of pages. Grouping by the results uses Google's current view of intent as the deciding factor, which is closer to what you need for ranking.

The SERP overlap method, step by step

SERP overlap compares the top organic results for two keywords and counts how many URLs they share. The more they share, the more likely one page can rank for both.

Step 1: build and clean the list

Collect keywords from Search Console, keyword research tools, sales conversations and People Also Ask. Our guide to finding long tail keywords UAE buyers search covers those sources. Remove duplicates, misspellings and terms for services you do not offer.

Step 2: collect the results for each keyword

For each keyword, record the top ten organic URLs for the location you care about, such as the UAE, in English or Arabic as appropriate. For small lists, search manually in a private window with the right location. For larger lists, rank tracking and SEO tools can export result URLs in bulk.

Step 3: compare pairs and count shared URLs

Take the keyword with the most demand as the anchor. For every other keyword, count how many of its top ten URLs also appear in the anchor's top ten. A working rule used by many SEO teams is:

These thresholds are a practical convention, not a Google rule. Adjust them if your results are dominated by directories or marketplaces, which can inflate the overlap.

Step 4: look at page types, not just URLs

Two keywords can share few URLs but show the same kind of page, such as agency service pages, or the reverse. Note the dominant format for each cluster: service page, comparison article, list, guide, product category or local listing. That format becomes the brief for the page.

Step 5: name each cluster and assign it to one URL

Give each cluster a primary keyword, usually the term with the clearest commercial intent or highest demand, and map it to one existing or planned URL. Every other keyword in the cluster becomes a secondary term, a heading or an FAQ on that page.

Worked example: a fit out company in Dubai

Hypothetical example: an interiors and fit out firm in Dubai has twelve keywords it wants to target. After collecting the top ten UAE results for each and counting shared URLs against the anchor terms, it ends up with four clusters.

Two findings changed the plan. First, "office interior fit out" initially looked like a candidate for its own design page, but it shared most of its results with the office fit out anchor, so it became a section rather than a separate page. Second, "cafe fit out" overlapped heavily with restaurant terms, so the firm did not need a separate cafe page. The approvals cluster pointed to government and landlord pages, so the guide was written to explain the process in general terms and tell readers to confirm current requirements with the relevant authority.

Content mapping: from clusters to a site structure

Content mapping is where clusters become a sitemap. Put the clusters in a sheet with these columns: cluster name, primary keyword, secondary keywords, intent, page type, target URL, status (exists, update, create) and the pages it should link to.

Map existing pages first

Before creating anything, match clusters to pages you already have. Search Console shows which URLs already get impressions for terms in each cluster. Improving a page that Google already associates with the topic is usually faster than starting a new one.

Plan hub and supporting pages

Commercial clusters usually map to service or category pages. Informational clusters map to guides that link back to the service they relate to. This structure gives the service page internal links from relevant content and gives readers a clear next step.

Bilingual sites

Run the overlap check separately for Arabic keywords. Arabic searches can return different result types from their English equivalents, so the clusters may not match one to one. Our guide to bilingual SEO for Arabic and English pages covers how to pair the two versions.

Keyword cannibalisation: finding and fixing clashes

Keyword cannibalisation happens when two or more of your pages target the same cluster and compete for it. Signs include rankings that swap between URLs, two pages sharing impressions for the same query in Search Console, and neither page holding a stable position.

Choosing the fix

• Merge and redirect when both pages cover the same need. Combine the best content on the stronger URL and permanently redirect the weaker one.
• Differentiate when the pages could serve different intents. Rewrite one to target a separate cluster and adjust its title and headings.
• Canonicalise when near duplicate pages must exist for practical reasons, such as filtered product listings.
• Fix internal links so menus, breadcrumbs and in content links point to the chosen page with consistent anchor text.

On large sites, cannibalisation often appears during a rebuild, when old and new templates overlap. The governance points in our article on SEO and UX governance in enterprise rebuilds help prevent that.

Tools and their limits

Several SEO platforms offer automated keyword clustering based on SERP overlap, and they save time on lists of hundreds or thousands of terms. Check a sample of their clusters manually, because results change by location, device and date, and tools may not collect UAE results by default. Semantic clustering, which groups by meaning using language models, is quick but does not reflect what Google ranks, so use it as a first pass only. For how AI tools fit into this kind of work more broadly, see how AI is changing web design and SEO work.

Turn your keyword list into a site plan

Codeeo's SEO team clusters keywords using search result data, maps them to existing and new pages and flags cannibalisation, and the web development team can build the resulting structure into a new or rebuilt site. If you are not sure where to start, a website audit reviews your current pages against search demand. Contact Codeeo to discuss your keyword list.