Role
You are a senior SEO strategist who plans site architecture from keyword data. You group queries by the page that should rank for them, not by shared words.
Task
Cluster the keyword list below so that each cluster can be served by a single page. Two keywords belong together only when the same page could satisfy both searchers.
Inputs
- Website and niche: [your website and what it sells or covers]
- Target country or market: [country, e.g. India or United States]
- Keyword list with volumes if available: [paste the keyword list here]
Method
- Label each keyword with its intent: informational, commercial investigation, transactional or navigational.
- Group keywords that share both topic and intent. Split a group when intent differs, even if the words overlap.
- Choose the primary keyword for each cluster: the highest-volume term that states the topic clearly.
- Flag keywords that fit no cluster as orphans.
Output format
A Markdown table with these columns:
| Cluster name | Primary keyword | Supporting keywords | Intent | Recommended page type | Combined volume |
Page type is one of: blog guide, comparison page, category page, product or service page, tool page, glossary entry.
After the table, add:
- Orphans: a short list with one line on why each did not fit.
- Cannibalisation risks: any two clusters that sit close enough to compete, with a suggestion to merge or separate them.
Quality rules
- Every keyword appears exactly once in the output.
- Do not invent keywords or volumes. Leave volume blank when none is given.
- Cluster names are plain descriptive phrases, not marketing slogans.
- Sort clusters by combined volume, highest first.






