SERP clustering: let the search results decide which keywords share a page
SERP clustering is a way of grouping keywords by comparing the pages Google ranks for each one. If two keywords return many of the same URLs in their top results, the search engine is treating them as one need, and they usually belong on one page. It replaces guesswork about meaning with evidence from the results themselves.
On this page — 4 sections
What is SERP clustering, and why use results instead of meaning?
Quick answer
SERP clustering groups keywords whose top search results share URLs. It uses results because they show how the engine actually interprets each query. Two phrasings that look different can share a page, and two that look alike can need separate pages.
Grouping by meaning works well for spelling variants. It fails on pairs where the words hide a different need. "مكيف سبليت" and "سعر مكيف سبليت" look like one topic, but in Riyadh the first may return guides and comparisons while the second returns shop listings. The results tell you that; the words don't.
The reverse also happens. On an Egyptian cooking site, "طريقة عمل الملوخية" and "الملوخية الخضراء بالفراخ" may share most of their results, because the engine sees both as a request for the standard recipe.
How does SERP clustering work, step by step?
Quick answer
Collect the top ten results for every keyword in the same market and language. For each pair of keywords, count the URLs they share. Pairs above your chosen threshold are grouped together. Each resulting group becomes one topic, and each topic gets one page.
| Keyword pair (Dubai) | Shared URLs in top 10 | Grouped at 3+? |
|---|---|---|
| "تصليح مكيف" / "صيانة مكيفات" | 6 | Yes: one page |
| "تصليح مكيف" / "فني مكيفات" | 4 | Yes: one page |
| "تصليح مكيف" / "تنظيف دكت المكيف" | 1 | No: separate page |
| "تصليح مكيف" / "تركيب مكيف سبليت" | 0 | No: separate page |
Collect all the results for a project at the same time and with the same location settings. Results move, and comparing a keyword checked last month with one checked today mixes two different snapshots.
How many shared results should it take to group two keywords?
Quick answer
There is no official number. Many SEO specialists start at three or four shared URLs out of ten, roughly 30 to 40 percent. A lower threshold merges more keywords into fewer pages; a higher one splits them into more pages. Adjust it after reading the groups.
- Too low: unrelated keywords end up together. An oud store in Riyadh would see "دهن عود" grouped with "بخور" because a few big retailers rank for both.
- Too high: close variants stay apart, and you plan two pages that will compete. "كنب زاوية" and "كنبة ركنة" in Cairo end up as separate pages.
- About right: most groups read as one need when you scan them, with a handful of borderline cases to decide by hand.
When should you overrule what the clusters say?
Quick answer
Overrule them when the overlap comes from something other than intent: a few giant sites ranking for everything, a map pack or marketplace dominating local results, or keywords with too few results to compare. In those cases, read the results yourself and decide.
In the Gulf, large marketplaces and directories often fill several top positions for many shopping keywords. Two keywords can share four URLs simply because the same marketplace ranks for both, not because the needs are the same.
Business sense also overrules the data. A Kuwait car-service shop may find that "تغيير زيت" and "فحص سيارة" overlap slightly, but they are separate services with separate prices, and each deserves its own page and booking button.
This article is part of the Semantic SEO series — Writing for meaning: entities, definitions, attribute-value facts, semantic distance and pages that actually answer.
About the author
Mohamed Youns
Semantic SEO Engineer · Author & system developer
Mohamed Youns writes about how search engines understand content — the same standards he applies when building semantic systems at Nut Hub. nut-hub.org