Develop Data-Driven Content Strategies

Provides a framework for planning content strategy, topic clusters, and an editorial roadmap around one governing rule: every piece of content should be searchable, shareable, or both, prioritized in that order since search traffic is the foundation. Searchable content matches existing search intent (a specific keyword, comprehensive coverage, keywords in title/headings/URL); shareable content creates demand via a novel insight, contrarian take, or honest story. It defines concrete content types on each side - use-case pages ([persona] + [use-case]), hub-and-spoke topic clusters, and template libraries for searchable; thought leadership, data-driven content, expert roundups, case studies, and "meta" behind-the-scenes posts for shareable - and a content-pillar structure (3-5 core topics identified product-led, audience-led, search-led, or competitor-led) that each spawn subtopic clusters of individual articles. Keyword research is mapped to four buyer stages by modifier pattern: awareness ("what is," "how to"), consideration ("best," "vs," "alternatives"), decision ("pricing," "reviews," "demo"), and implementation ("templates," "tutorial," "setup"). Six ideation sources are named with what to extract from each: keyword exports (topic clusters, buyer stage, quick wins), sales/support call transcripts (questions, objections, verbatim customer language), survey responses, forum research (site:reddit.com/site:quora.com searches for FAQs and debates), and competitor content analysis (site:competitor.com/blog for top posts and coverage gaps). Ideas are then scored on a weighted rubric: Customer Impact 40%, Content-Market Fit 30%, Search Potential 20%, Resource Requirements 10%.

/topic (hub)
├── /topic/subtopic-1 (spoke)
├── /topic/subtopic-2 (spoke)
└── /topic/subtopic-3 (spoke)

Use it when deciding what content to create and in what order - building topic clusters, content pillars, or an editorial roadmap - not for writing the actual copy, which routes to a copywriting skill, technical or on-page SEO (seo-audit), AI-search optimization (ai-seo), scaled or templated content generation (programmatic-seo), page hierarchy and URL structure (site-architecture), email content (email-sequence), or social posts (social-content). It explicitly warns against defaulting to elaborate hub/spoke URL structures for ordinary posts - most content works fine under /blog, and dedicated pillar or hub URL paths are reserved for major topics with genuinely layered depth. Before planning, it checks for an existing .agents/product-marketing-context.md file and only asks about business context, customer research, current content state, and competitors not already covered there.

Inputs are business and customer context plus whatever raw research material is available - keyword exports from Ahrefs/SEMrush/GSC, sales or support call transcripts, survey responses, or nothing at all, in which case forum and competitor web searches substitute. Output is a three-part content strategy: content pillars with rationale and their subtopic clusters, a prioritized list of specific topics each tagged searchable/shareable/both, content type, target keyword, buyer stage, and the customer-research evidence behind it, and a topic-cluster map showing interconnection. A bundled reference (references/headless-cms.md) extends the skill into CMS selection and content modeling: a headless-vs-traditional decision checklist, seven common marketing content types (Landing Page, Blog Post, Case Study, Testimonial, FAQ, Author, CTA Block) with their key fields, a required SEO-fields checklist per page type, a draft-review-approve-schedule-publish editorial workflow, and a platform comparison across Sanity (GROQ, real-time collaboration), Contentful (enterprise multi-locale), and Strapi (self-hosted, no per-seat pricing).

Names concrete tools throughout: Ahrefs, SEMrush, and Google Search Console for keyword data; Reddit, Quora, Indie Hackers, Hacker News, and Product Hunt for forum-sourced ideation; and Sanity, Contentful, or Strapi, plus Hygraph, Keystatic, Payload, Builder.io, and Prismic as further options, for the content backend. It hands off to seven sibling skills for adjacent work: copywriting, seo-audit, ai-seo, programmatic-seo, site-architecture, email-sequence, and social-content.

Content strategists and marketers deciding what to publish, in what order, and for whom - building a pillar and cluster structure and a prioritized editorial roadmap grounded in customer research rather than guesswork - who also need CMS content-modeling guidance once the strategy moves into implementation.