Competitive Research

Competitive Research Skill

This skill produces a competitive landscape document — a structured audit of 2–4 named or inferred competitors covering messaging, positioning, content strategy, and market gaps. It registers the output as project context for reuse by positioning, gtm-plan, channel-advisor, editorial-strategy, and cc-content skills.

This skill follows the shared step sequence in ../_shared/skill-contract.md. Steps below add only competitive-research-specific behavior.

Step 0: Recall learnings

If .claude/learnings.md exists, read it silently. Apply relevant entries — in particular, prior competitor selections, frameworks tried, or market-specific constraints. Do not announce this step. If the file is absent, continue normally.

Step 1: Load context and identify competitors

Read the ## Context files table in CLAUDE.md. If present, load context/organization-background.md (or equivalent Summary match: organization, company, business model, background, profile, mission, values, size, market position, customer, constraints) — competitive teardowns are grounded in what the organization actually does.

Ask once to identify 2–4 competitors:

• If $ARGUMENTS includes competitor names (e.g. /competitive-research Acme, BigCorp, StartupX) → parse and use them directly; skip the question.

• If organization background is loaded and names obvious candidates (e.g., "competes against Salesforce, HubSpot, Pipedrive") → confirm and use them.

• Otherwise, ask once:

  "Name 2–4 of your main competitors or market alternatives. (You can list companies, product categories, or delivery models — e.g. 'Salesforce, HubSpot, custom-built solutions'.) I'll research their recent public content to audit messaging, positioning, and market gaps."

• If the user doesn't name any and org background yields no hints → proceed with a light audit scope ("I'll work with what you've given me; if you add competitor names later, we can run this again").

Step 2: Briefing check

Confirm the audit has at least one named competitor to research. If none are named and no organization background exists, flag this as a non-degraded case and offer to proceed anyway (ask: "Should I proceed with what you have, or would you like to add competitor names now?").

Step 3: Calibrate

Consult the brief in $ARGUMENTS, if present. Set audit depth and scope:

• Narrow brief (e.g., "focus on messaging only" or "compare social strategy") → scope the audit tightly to those dimensions.
• Broad brief (e.g., "full competitive landscape") → develop a complete teardown across all dimensions.
• No brief → develop a full competitive audit across all dimensions.

Step 4: Generate

For each competitor (up to 4), conduct a structured teardown using the framework from docs/research/competitive-research.md § 1 (repeatable competitive-content teardown structure). The framework covers:

A. Identity & positioning

• Overview (company, target market, target audience)
• Positioning statement / category anchor
• Core value proposition
• Primary differentiator / "only-we" claim
• Positioning claims made (verbatim taglines/headlines)

B. Messaging & voice

• Messaging themes / pillars
• Tone & voice (formal/casual, technical/accessible)
• Consistency of voice across pages
• Emotional vs. rational appeal

C. Target-segment signals

• Personas addressed (job titles, industries)
• Segment-specific content/landing pages
• Buyer-committee roles targeted

D. Content coverage & cadence

• Topic clusters covered (breadth) and depth per cluster
• Format mix (blog, video, whitepaper, webinar, tools, interactive)
• Publishing frequency / cadence
• Content freshness
• Funnel-stage distribution (TOFU/MOFU/BOFU)

E. Distribution & amplification

• Social presence (platforms, follower counts, posting frequency, engagement rate)
• Backlink/referring-domain profile
• Earned media / press mentions

F. Conversion architecture

• Primary and secondary CTAs
• Lead-capture mechanisms
• Proof elements (case studies, testimonials, ROI calculators, comparison pages)

G. Assessment

• Standout/differentiating features
• Observed strengths and weaknesses
• Gap/opportunity notes

For each competitor, use WebSearch and WebFetch to pull recent public content (site copy, blog archives, social feeds, press). Summarize findings per dimension (score 1–5 or Harvey Ball notation where applicable). Flag standout strengths, execution gaps, and underserved angles each competitor misses.

After the teardowns, synthesize a content-focused SWOT per competitor (3–4 items per quadrant), referencing only marketing/content/messaging factors:

• Strengths — dominant topic clusters, high-authority profile, consistent voice, clear differentiation.
• Weaknesses — thin topic coverage, stale content, weak messaging, missing funnel stages, format monotony.
• Opportunities — uncovered keywords/topics, underused formats, underserved segments, unaddressed trends.
• Threats — competitors with stronger resources, content easily replicated, rivals narrowing focus into your niche, algorithm/CTR changes.

Share-of-Voice (SOV) sample. Optionally include a light SOV estimate:

• SEO SOV: For 5–10 key category keywords (free via Google Keyword Planner), record each competitor's top 3 ranking positions and apply a documented CTR curve (position 1 ≈ 30%, position 2 ≈ 15%, position 3 ≈ 10%, etc.) to estimate search traffic share. Document your keyword set and CTR model.
• Share of Search (the most rigorous free proxy): Use Google Trends to compare search volume for each competitor's brand over the last 12 months. Record as a simple ratio (e.g., "3x more searches for Competitor A than Competitor B").

Content-gap analysis. Identify white space across competitors:

• Keywords or topics one or more competitors rank for that none of the others do.
• Content stages missing (commonly MOFU/BOFU — comparison pages, case studies, ROI calculators for specific verticals).
• Formats no competitor uses for a given topic (e.g., only long-form, no FAQ or video).
• Audience segments no competitor serves with dedicated content.

Prioritize gaps by impact (search volume, business relevance) and effort.

Step 5: Internal quality check

Review the competitive landscape for:

• Completeness: All named competitors covered? All seven dimensions attempted?
• Grounding: Findings tied to actual public content, not speculation?
• Consistency: Scoring logic (1–5 or Harvey Balls) consistent across competitors?
• Clarity: Can someone unfamiliar with each company understand the summary?
• Actionability: Are gaps and opportunities concrete enough to feed positioning or content strategy work?

Fix any gaps before proceeding.

Step 6: Delimited output

Return the competitive landscape inside clear delimiters:

─────────────────────────────────────────────────────────────────
## Competitive Landscape Audit
─────────────────────────────────────────────────────────────────

[Generated competitive landscape document here]

─────────────────────────────────────────────────────────────────

Step 7: Register output

Write the competitive landscape document to context/competitive-landscape.md by convention. Apply collision-check logic to ensure the file does not overwrite or collide with an existing competitive analysis or market research file:

1. Check if context/competitive-landscape.md exists.
2. If it exists, generate a distinct filename by appending a -N suffix before .md (e.g., context/competitive-landscape-2.md). Keep incrementing N until a filename that does NOT already exist is found. Never overwrite an existing file.
3. Write to the final, collision-checked filename.

Confirm the file was created, then add a row to the ## Context files table in CLAUDE.md (or create the table if it does not exist). Use this format:

| Competitive landscape | `context/competitive-landscape.md` | Competitive landscape: competitor teardown, messaging audit, SWOT, share-of-voice, and content-gap analysis across 2–4 competitors; market gaps and underserved angles |

The Summary must include keywords "competitor", "competitive landscape", and "market gap" so that gated skills (positioning, gtm-plan, channel-advisor) recognize and load it via semantic match — do not hand-edit the Key Config Files table; the pre-commit hook owns that sync.

Step 8: Feedback

Store learnings tagged [cc-concept:competitive-research] in .claude/learnings.md. Examples:

[cc-concept:competitive-research] "share-of-voice" confuses clients; prefer "search visibility" — 2026-07-20
[cc-concept:competitive-research] industry leads with 5 key vendors, not 2-4; scope as broad audit when user lists >4 — 2026-07-20

If the session revealed a preferred competitor set, market dynamics, or a framework that worked or didn't, record it now.