Competitor Analysis skill for Claude Code

Good competitor analysis answers a decision (“how do we position the launch?”, “what do we build next?”, “which keywords can we win?”) with dated, sourced evidence, and ends in specific moves. A feature checklist with green ticks for us and red crosses for them is not analysis; it is wishful thinking.

Ground rules

• Start with the decision the research serves. Scope everything else down to it.
• Every claim gets a source and a date: URL, screenshot, review link, or tool export. Competitor sites change weekly; undated findings rot.
• Use public information only. No logging into competitors’ products under false pretenses, no scraping that violates terms or robots.txt, no social engineering of their staff. Signing up for a public free trial as yourself is fine.
• Be fair. Describe competitors as their best customers would. Underestimating rivals is the most expensive analysis error.
• Separate observation (“pricing page lists 3 tiers starting at $29”) from inference (“they are moving upmarket”) and label inferences with confidence.
• Third-party traffic and keyword numbers (Similarweb, Semrush, Ahrefs) are estimates; use them for relative comparison, not as facts.

Workflow

- [ ] 1. Define the decision and scope
- [ ] 2. Build the competitive set
- [ ] 3. Capture positioning and messaging
- [ ] 4. Compare product and pricing
- [ ] 5. Mine customer sentiment
- [ ] 6. Analyze SEO, content, and channels
- [ ] 7. Synthesize: gaps, threats, moves
- [ ] 8. Deliver the report (and battlecards if needed)

1. Decision and scope

Write: the decision, the audience for the report (founders, sales, marketing, product), the segment (e.g. “US SMB accounting firms”), and the time budget. A launch-positioning question needs messaging and pricing depth; an SEO question needs keyword and content data.

2. Competitive set

Tier	Definition	How many
Direct	Same job, same customer, same category	3–5
Indirect	Same job, different approach (agency, spreadsheet, platform feature)	2–3
Status quo	Doing nothing / manual process	Always include
Emerging	New entrants, open-source projects, AI-native tools	1–3

Sources for discovery: G2/Capterra category pages, “alternatives to X” SERPs, customer interviews (“what else did you consider?”), lost-deal notes, Product Hunt, GitHub topics, investor announcements.

3. Positioning and messaging

For each competitor, capture from the home page, pricing page, and top landing pages (screenshot with date):

• Headline and subhead verbatim; the target audience they name.
• Category they claim (“AI-native CRM”, “open-source Firebase alternative”).
• Top 3 differentiators they claim; proof they use (logos, numbers, awards).
• Primary CTA and motion (self-serve trial, demo-led, open source → cloud).
• Tone and visual style in one line.

Plot positioning on two axes that matter to buyers in this segment (e.g. simple ↔ powerful, SMB ↔ enterprise, price). Empty quadrants are candidate positions — if customers actually want them.

Wayback Machine (web.archive.org) snapshots of the home and pricing pages over 12–24 months reveal direction changes.

4. Product and pricing

Feature comparison only for capabilities buyers weigh in decisions (from interviews, reviews, and RFPs), rated honestly:

Capability (buyer language)	Us	Comp A	Comp B	Evidence
Import from QuickBooks	Full, 1-click	Full	CSV only	Docs URLs, trial test 2026-10-12

Use a 4-level scale (none / partial / full / best-in-class) with evidence, not ticks.

Pricing: tiers, price points (monthly and annual), value metric (per seat, per usage, flat), free plan/trial terms, limits that force upgrades, enterprise gating (SSO, audit logs), add-ons. Compute the price for 2–3 realistic customer profiles — list prices hide what customers actually pay.

Changelogs, release notes, public roadmaps, job postings (hiring for ML, sales in EMEA), and docs reveal investment direction.

5. Customer sentiment

Mine G2, Capterra, TrustRadius, App Store/Play reviews, Reddit, Hacker News, and community forums. Read the 3-star reviews first; they are the most balanced. Code themes:

Theme	Comp A mentions (+/−)	Example quote (link)
Onboarding complexity	4+ / 19−	“Took our team 3 months to set up”

Look for: switching triggers (why people leave them), unmet needs, praised strengths you must match, and segment fit (who loves them vs. who churns).

6. SEO, content, and channels

• Organic: with an SEO tool, compare estimated organic traffic, top pages by traffic, and keyword overlap. Run a content gap (keywords where 2+ competitors rank top 10 and you do not). Filter to keywords relevant to your product and intent; hand the list to keyword-research for prioritization.
• Content types: which page types drive their traffic (comparison pages, templates, integrations, programmatic pages, docs, free tools).
• Paid: Google Ads Transparency Center and Meta Ad Library show active ads and messaging.
• Community and social: where they show up (Discord, YouTube, LinkedIn founders, newsletter), posting cadence, engagement level.
• Partnerships and integrations: marketplace listings, co-marketing, affiliates.
• Tech stack hints: BuiltWith or Wappalyzer, used sparingly for relevant signals (e.g. moving to a new billing system).

7. Synthesize

Summarize per competitor in a few lines: who they win with, why, where they are weak, and what they will likely do next. Then:

• Gaps: customer needs that are poorly served across the set (from sentiment + positioning map).
• Threats: where a competitor is better for our core segment, or moving into it.
• Moves: 3–7 specific recommendations, each tied to evidence and owner — positioning line changes, comparison pages to create, features to prioritize, pricing adjustments, keyword clusters to target, sales talk tracks.

8. Battlecards (for sales)

One page per competitor: when we win, when we lose, their claims and our honest responses, landmine questions to ask the prospect, proof points, pricing comparison for typical deal size. Keep it factual; inaccurate battlecards destroy sales credibility in one call.

Comparison and “alternatives” pages

When the output feeds public “Us vs Them” pages: state facts verifiable on the competitor’s own site as of a stated date, acknowledge where they are a better fit, avoid trademark misuse in logos, and have the claims reviewed. Misleading comparative claims create legal exposure (e.g. FTC and Lanham Act in the US, comparative advertising rules in the EU).

Gotchas

• Feature-parity thinking: copying competitor roadmaps produces a worse version of them. Look for underserved jobs, not missing checkboxes.
• Treating estimated traffic numbers as precise; compare ranks and trends instead.
• Researching only direct competitors misses the status quo, which is usually the biggest rival.
• Reviews skew toward extremes and toward vendors who solicit them; weight by volume and recency.
• Reports without dates or sources become untrustworthy within a quarter.

Output: report template

# Competitor analysis — <segment>, for <decision> (<date>)

## Summary
<5 bullets: the landscape, where we win, biggest threat, top 3 moves>

## Competitive set
| Competitor | Tier | Target customer | Positioning line | Pricing from | Motion |

## Positioning map
<axes, placements, open space>

## Product and pricing comparison
<capability table with evidence; price for profiles S/M/L>

## Customer sentiment
<theme table per competitor with quotes and links>

## SEO and channels
<traffic estimates (relative), top content types, content gap clusters, paid messaging>

## Gaps, threats, and recommended moves
| # | Move | Evidence | Expected impact | Owner | Effort |

## Sources
<URL — accessed date — what it supports>