Back to .md Directory

Content Research Skill

> Research topics, trends, competitors, and audience questions to generate content briefs. Combines web search, community mining, and keyword analysis into actionable research documents.

May 2, 2026
0 downloads
0 views
ai mcp claude workflow
View source

Content Research Skill

Research topics, trends, competitors, and audience questions to generate content briefs. Combines web search, community mining, and keyword analysis into actionable research documents.

When to Use

Use this skill when the user wants to:

  • Research a topic before writing content
  • Find trending subjects in their niche
  • Discover what questions their audience is asking
  • Analyze competitor content for gaps and opportunities
  • Build a content brief with sources and angles
  • Find statistics, quotes, or references to support a piece

Trigger phrases: "research this topic", "what's trending in", "content ideas for", "what are people asking about", "competitor content analysis", "find sources for", "research before writing"

Prerequisites

  1. Web search access — Claude Code's WebSearch tool must be available
  2. (Optional) Ahrefs API for keyword data — if the Ahrefs MCP server is connected
  3. (Optional) Notion database for saving research briefs — see content-writing-notion.md skill

No API keys are strictly required — the core workflow uses web search. Additional tools enhance depth.

Workflow

Phase 1 — Define the Research Scope

  1. Ask the user:
    • Topic or keyword: What area are we researching?
    • Purpose: Blog post? Video script? Social content? Product positioning?
    • Audience: Who are we writing for? (role, experience level)
    • Depth: Quick scan (10 min) or deep dive (comprehensive)?
  2. Set the research plan based on depth:
DepthSourcesOutput
Quick scanTop 5 web results + 1 community sourceBullet-point brief
StandardTop 10 results + 3 communities + keyword dataStructured research doc
Deep dive15+ sources + competitors + keyword + trendsFull research report

Phase 2 — Web Research

Search for the topic across multiple angles:

Core topic search:

  • "{topic}" best practices 2025
  • "{topic}" tutorial guide
  • "{topic}" common mistakes

Audience questions:

  • "{topic}" reddit — find real questions people are asking
  • "{topic}" site:stackoverflow.com — technical questions
  • "{topic}" "how do I" — beginner intent
  • "{topic}" vs — comparison intent (auto-suggest reveals alternatives)

Trend signals:

  • "{topic}" trends 2025
  • "{topic}" new features
  • "{topic}" announcement

Expert perspectives:

  • "{topic}" site:news.ycombinator.com — Hacker News discussions
  • "{topic}" conference talk
  • "{topic}" case study

For each source, extract:

  • Key claims (with attribution)
  • Data points (statistics, benchmarks, numbers)
  • Unique angles not covered elsewhere
  • Controversies or debates (great for engagement)

Phase 3 — Community Mining

Dig into community sources for authentic audience questions and pain points:

Reddit: Search relevant subreddits for the topic. Focus on:

  • Top questions (sorted by upvotes)
  • Pain points mentioned repeatedly
  • "I wish..." or "Why can't..." patterns
  • Advice threads (what practitioners actually recommend)

Dev.to / Hashnode / Medium: Search for existing content on the topic. Note:

  • Which articles performed well (claps, reactions, comments)
  • What angle they took
  • What's missing or outdated in existing coverage

Stack Overflow / GitHub Discussions: For technical topics:

  • Most common errors or confusion points
  • Frequently asked questions
  • Popular workarounds (sign of a gap in documentation)

Phase 4 — Keyword & SEO Data (if Ahrefs available)

If the Ahrefs MCP server is connected, pull keyword data:

  1. Keyword overview: Search volume, difficulty, CPC
  2. Related terms: Long-tail variations and question keywords
  3. SERP overview: Who currently ranks and what type of content
  4. Content gap: What competitors rank for that you don't

Without Ahrefs, use web search to estimate:

  • Google autocomplete suggestions (search and note suggestions)
  • "People also ask" boxes
  • "Related searches" at the bottom of Google results

Phase 5 — Competitor Content Audit

Identify the top 3-5 existing pieces on this topic:

For each piece, document:

### [Article Title](URL)
- **Author / Site**: Who published it
- **Word count**: Approximate length
- **Format**: Tutorial / listicle / opinion / case study
- **Key points covered**: Bullet list
- **Strengths**: What they did well
- **Gaps**: What's missing, outdated, or shallow
- **Engagement signals**: Comments, shares, backlinks (if visible)

Synthesize the audit into:

  • Table stakes: Points every article covers (you must include these)
  • Differentiators: Angles no one has taken yet
  • Outdated info: Things that have changed since competitors published

Phase 6 — Assemble the Research Brief

Compile everything into a structured document:

# Content Research Brief: {Topic}

**Date**: {today}
**Researcher**: Claude Code
**Depth**: {quick / standard / deep}

## Topic Overview
{2-3 sentences summarizing the landscape}

## Target Audience
- **Who**: {role, experience level}
- **What they need**: {information, guidance, validation}
- **Where they are**: {platforms they frequent}

## Key Findings

### What the audience is asking
1. {Question from Reddit/community} — [{source}]({url})
2. {Question from Stack Overflow} — [{source}]({url})
3. {Question from forums} — [{source}]({url})

### Data points & statistics
- {Stat 1} — [{source}]({url})
- {Stat 2} — [{source}]({url})
- {Stat 3} — [{source}]({url})

### Expert opinions & quotes
- "{Quote}" — {Expert Name}, {Context} [{source}]({url})

### Trends & signals
- {Trend 1}: {why it matters}
- {Trend 2}: {why it matters}

## Competitor Content Audit
| Article | Site | Strengths | Gaps |
|---|---|---|---|
| {title} | {site} | {strengths} | {gaps} |

## Keyword Data
| Keyword | Volume | Difficulty | Intent |
|---|---|---|---|
| {keyword} | {vol} | {diff} | {intent} |

## Recommended Content Angles
1. **{Angle 1}**: {Why this works — gap it fills}
2. **{Angle 2}**: {Why this works — unique perspective}
3. **{Angle 3}**: {Why this works — audience demand}

## Suggested Outline (for top recommended angle)
- H1: {Title}
- H2: {Section 1}
- H2: {Section 2}
- H2: {Section 3}
- H2: {Conclusion / CTA}

## Sources
1. [{Title}]({url}) — {one-line summary}
2. [{Title}]({url}) — {one-line summary}
...

Phase 7 — Save & Handoff

  1. Save the research brief locally:
    mkdir -p ./research
    # Save as: research/{slugified-topic}-{date}.md
    
  2. (Optional) Push to Notion if the content-writing-notion skill is available
  3. If the user wants to proceed to writing, hand off to the content-writing skill with the brief as context

Customization Points

VariableDescriptionWhere to Set
Niche / industryChanges which communities and keywords to focus onPhase 1 input
Competitor listPre-define competitors to always checkCreate competitors.json
Keyword toolAhrefs (MCP), or manual Google suggestionsDepends on available tools
Output locationLocal ./research/ dir or Notion databasePhase 7 config
Research depthQuick / Standard / Deep — controls thoroughnessPhase 1 input

Troubleshooting

IssueFix
Web search returning irrelevant resultsNarrow the query with quotes and site-specific operators
No community discussions foundTry broader terms or adjacent topics. Some niches are quieter.
Ahrefs not availableFall back to Google autocomplete and "People also ask" mining
Research brief too longUse "Quick scan" depth. Limit to top 5 sources.
Competitor content is paywalledNote it as a gap — your free content can capture that audience
Topic too broadAsk the user to narrow to a specific subtopic or angle

Related Documents