---
name: competitor_teardowns
description: Runs a structured competitive analysis: captures positioning, pricing, messaging, and product facts on competitors, scores them on a fixed rubric, and writes the findings as a one-page teardown brief the team can actually act on. Use when asked to compare competitors, analyze a market, audit a rival's offering, or answer "how do we stack up".
---

# Competitor Teardowns

## Purpose

Most competitor analysis dies in a doc nobody reads. This skill fixes that by forcing two things: a fixed rubric so competitors get scored on the same axes, and a one-page output contract so findings stay short enough to use. It is for teams deciding how to position, price, or message against real rivals, not for market-sizing reports.

## Workflow

1. **Scope first.** Ask for the user's product (or read it from context) and the 3 to 5 competitors to analyze. Cap at 5. More than that and scores blur together. Also confirm the decision this teardown serves: pricing change, repositioning, feature gap, sales battle card. That decision picks the scoring weights.
2. **Capture facts, same template each time.** For each competitor, gather these five blocks. Use web search to verify pricing and product claims; do not write a price from memory. Note the source and date for every fact.
   - Positioning: the one sentence the competitor uses to describe itself (usually the homepage hero). Quote it verbatim.
   - Pricing: tiers, headline price, free tier or trial, any usage-based component. Mark anything you could not verify as unverified.
   - Messaging: the three claims repeated most across their site, ads, and docs. Look for the pattern, not one-off lines.
   - Product: top 3 features they push hardest, plus any obvious gaps reviewers or changelogs reveal.
   - Proof: logos, case studies, review scores (with source and date), funding or size signals.
3. **Score on the rubric.** Rate each competitor 1 to 5 on six axes: positioning clarity, price competitiveness, messaging consistency, feature depth, proof strength, moat (switching costs, network effects, integrations). Weight the axes for the decision at hand. Example: a pricing decision weights price competitiveness and moat at 2x, everything else at 1x. Show the weights.
4. **Find the gaps.** For each competitor, write exactly one sentence: "They are vulnerable where [specific weakness], because [evidence]." If you cannot back it with captured evidence, drop it.
5. **Write the brief.** Produce the one-page output described in Output Contract.

## Output Contract

The deliverable is one page, in this order, no exceptions:

1. **The matchup** (3 lines): our product in one sentence, the competitors by name, the decision this teardown serves.
2. **Score table**: competitors as rows, six axes as columns, weighted total. Mark the winner per axis.
3. **Three findings**: the three most decision-relevant facts, each one sentence plus the evidence behind it. If a finding does not change the decision, cut it.
4. **What to do**: two or three concrete moves (reword a page, test a price point, build or highlight a feature), each tied to a finding above. No move may be introduced here that has no finding behind it.
5. **Fact ledger**: every price, claim, and review score with source and capture date.

Everything goes in the chat or one markdown file, whichever the user asked for. Never split findings across multiple documents.

## Operating Rules

- Verify pricing and feature claims with a fresh search, every time. Sites change plans and pages quietly; a number you remember is a number you invented.
- Quote competitor copy verbatim for positioning and messaging, then analyze. Paraphrase loses the signal.
- Score every competitor on every axis. A blank cell is a skipped judgment, not honesty.
- Cap findings at three and moves at three. The point of a teardown is focus, not coverage.
- No market-overview talk, no trend essays, no TAM paragraphs. If a sentence does not help the stated decision, delete it.
- State the weights before scoring, and show them in the output. Weighted scores without shown weights are just opinions with extra steps.
- Date the fact ledger. A teardown has a shelf life; the ledger tells the next reader whether this one is still fresh.