What this Skill helps with

Keyword Clustering Intent Mapper groups a keyword list into related topics, classifies likely search intent, maps terms to funnel stages, and proposes pillar and supporting pages. It helps teams turn an unstructured export into a content architecture that can be reviewed before briefs are written.

Who this Skill is for

It is intended for SEO practitioners, content teams, agencies, and founders planning a new topic area or reorganizing an existing site. It works best when the list includes search volume, current ranking URL, clicks, impressions, or conversion context.

What you provide

  • A keyword list, CSV, or Search Console export
  • Market, product, audience, and country context
  • Existing page URLs and target terms
  • Business priorities and conversion goals
  • Brand or compliance constraints

What it produces

  • Keyword groups with representative topics
  • Informational, navigational, commercial, and transactional intent labels
  • Awareness, consideration, and decision-stage mapping
  • Suggested pillar and supporting pages
  • Notes on ambiguous terms and possible negative keywords
  • A CSV-ready output for further planning

How the Skill works

The workflow cleans duplicated and malformed terms, then classifies intent before grouping semantically related queries. It separates terms that look similar but imply different tasks. The final architecture maps each group to one proposed page and flags places where an existing URL may already satisfy the need.

Installation and first use

  1. Download and add the Skill to Claude.
  2. Attach the keyword file and explain the business context.
  3. Identify the market and intended audience.
  4. Ask Claude to flag ambiguous terms before clustering.
  5. Review page recommendations against the existing site.

Example workflow and expected output

Upload 300 keywords for a B2B reporting product with current ranking URLs. The expected output is a set of named clusters, intent and funnel labels, a proposed pillar page, supporting pages, and a list of conflicts with existing URLs. The output should keep uncertain classifications visible rather than forcing them into a group.

Limits and review guidance

Intent cannot be confirmed from wording alone. Review important terms against current search results and customer language. Search volume does not equal business value, and a generated cluster should not create a new page when an existing page can satisfy the same need. Check for cannibalization before publication.

Frequently asked questions

How many keywords can I provide?

Large CSV files can work, but splitting very large exports by market or product area often makes review easier.

Does the Skill use live search results?

Only when live access and research are explicitly available. Otherwise its intent labels are reasoned estimates that need SERP review.

Will each cluster become a new page?

Not necessarily. Some clusters belong on an existing page, and closely related groups may be better served together.