| name | keyword-research |
| description | Standalone keyword research — expands seeds via the brand's connected keyword MCP (Ahrefs, Semrush, SE Ranking, or GSC), classifies search intent, maps keywords to content types, surfaces competitor content gaps, long-tail and SERP-feature opportunities, and delivers a prioritized keyword strategy document. Volume and difficulty come from the connected provider and are never fabricated. Triggers on "/digital-marketing-pro:keyword-research", "what keywords should we target", "find content gaps versus competitors", "expand these seed keywords", "which queries have buying intent". Reads the brand profile, guidelines, and campaign history; hands 20+ raw keywords to /digital-marketing-pro:keyword-cluster for pillar+spokes clustering. |
| argument-hint | [topic or seed keywords] |
/digital-marketing-pro:keyword-research
Purpose
Standalone keyword research tool — expansion, search-intent classification, and competitor gap analysis. Produces a prioritized, intent-classified keyword list with content recommendations. Volume and keyword-difficulty figures come from the brand's connected keyword MCP (Ahrefs / Semrush / SE Ranking / GSC) — this skill surfaces and interprets them, it does not fabricate them. Clustering into a pillar+spokes plan is delegated to /digital-marketing-pro:keyword-cluster (the keyword_cluster.py engine); this skill produces the seeds that skill consumes.
Input Required
The user must provide (or will be prompted for):
- Seed keywords or topic: Starting keywords, a topic area, or a URL to extract keyword themes from
- Target audience: Who the content is intended to reach (demographics, expertise level, pain points)
- Industry: The vertical or niche to contextualize volume and difficulty estimates
- Competitor domains: Optional -- 1-3 competitor domains to run content gap analysis against
- Target market/language: Geographic and language targeting for volume estimates
- Content goals: Traffic, leads, thought leadership, product sales, or brand awareness
- Existing content inventory: Optional -- URLs or topics already published to avoid duplication
Process
- Load brand context: Read
~/.claude-marketing/brands/_active-brand.json for the active slug, then load ~/.claude-marketing/brands/{slug}/profile.json. Apply voice, compliance, industry context. Check guidelines/_manifest.json for restrictions, messaging, channel styles, voice-and-tone rules, and templates. If a template matching this command exists in ~/.claude-marketing/brands/{slug}/templates/, apply its format. If no brand exists, prompt for /digital-marketing-pro:brand-setup or proceed with defaults.
- Check campaign history: Run
python "${CLAUDE_PLUGIN_ROOT}/scripts/campaign-tracker.py" --brand {slug} --action list-campaigns to identify previous keyword research and content campaigns to build upon rather than duplicate.
- Load reference files: Consult
skills/content-engine/ for content strategy context and skills/context-engine/industry-profiles.md for industry-specific keyword benchmarks and search behavior patterns.
- Expand the seed set: Use the brand's connected keyword MCP (Ahrefs
getRelatedKeywords, Semrush, SE Ranking, or GSC query mining) to expand seeds into a candidate list, pulling provider volume and keyword-difficulty figures where available. Record the provider and pull date — volume/KD are provider estimates, not measurements, and providers disagree by 20-50%. Do not claim volume/KD/trend numbers the connected tools didn't return.
- Classify search intent: Categorize every keyword into intent buckets -- informational (how-to, what-is), navigational (brand, product names), commercial (best, reviews, comparison), and transactional (buy, pricing, demo, free trial).
- Map keywords to content types: Assign each cluster a recommended content format -- blog post, landing page, pillar page, comparison page, FAQ, video, tool, or interactive content -- based on intent and SERP feature analysis.
- Identify content gaps vs competitors: If competitor domains were provided, cross-reference their ranking keywords against the brand's current coverage to surface missed opportunities and underserved topics.
- Discover long-tail opportunities: Expand each cluster with long-tail variants, question-based keywords (People Also Ask patterns), and related search modifiers that represent lower-difficulty entry points.
- Assess SERP feature opportunities: For each primary keyword, identify which SERP features are present (featured snippets, People Also Ask, knowledge panels, image packs, video carousels) and note which are attainable.
Output
A structured keyword strategy document containing:
- Keyword clusters organized by topic theme, each with individual keywords listed
- Estimated monthly search volume and keyword difficulty per keyword
- Search intent classification (informational, navigational, commercial, transactional) per keyword
- SERP feature opportunities per cluster (featured snippets, PAA, video, image pack)
- Recommended content type and format for each cluster
- Priority score (high/medium/low) with rationale for sequencing
- Content gap analysis showing competitor-owned keywords the brand is missing
- Long-tail keyword opportunities with lower difficulty and high relevance
- Question-based keyword list for FAQ and People Also Ask targeting
- Recommended content creation roadmap based on priority ranking
- Quick-win keywords (low difficulty, decent volume, high relevance) flagged for immediate action
- Seasonal or trending keyword opportunities with timing recommendations
- Internal linking opportunities between keyword clusters and existing content
Tips & caveats
- Search volume from any provider is an estimate. Ahrefs, Semrush, GSC, SE Ranking all disagree by 20-50% on the same keyword. Use ranges, not point estimates.
- Keyword difficulty (KD) is a heuristic, not a measurement. A KD of 60 means "competitive" — not "impossible". A small brand with niche authority can rank for KD-70 keywords against generalist KD-30 sites.
- Long-tail isn't always lower-volume. With AI search rewriting queries, the actual click-driving query may differ from the seed. Always check the resulting query a user typed via GSC, not the rank-tracker assumption.
- Hand off to
/digital-marketing-pro:keyword-cluster once you have ≥ 20 raw keywords. Clustering before writing is what produces topical authority, not keyword lists.
- Don't research the same keyword set quarterly. Re-research only when business model, target market, or competitive landscape changes. Otherwise the deltas are noise.
- Intent classification beats volume. A "buy [product]" query at 200/mo is worth more than "what is [product]" at 5000/mo for most commercial brands.
Agents Used
- seo-specialist -- Keyword research, volume and difficulty estimation, SERP analysis, content gap identification, and priority scoring
- content-creator -- Content type mapping, content angle recommendations, and editorial planning for keyword-targeted pieces
