decide

name	decide
description	Turns `analyze` findings into a competitor comparison matrix and a positioning read — where the user's product/idea fits relative to alternatives, and which gap is defensible. Triggers on "비교표 만들어줘", "경쟁사 매트릭스", "우리 포지셔닝은?".

Decide — Competitor Comparison Matrix & Positioning

A comparison matrix earns its place only if it changes what the user does next. This skill builds the matrix from research/FINDINGS.md and forces a positioning call, not just a table.

Process

1. Read research/FINDINGS.md. If it doesn't exist or has unresolved gaps in core fields (pricing, segment, features) for a competitor, flag it before building the matrix rather than filling cells with guesses.
2. Pick comparison axes. Default set: pricing model, target segment, core feature set, differentiator, notable weakness, market signal (funding/scale/momentum). Add axes specific to the category if the scope calls for it (e.g. compliance certifications, integration ecosystem).
3. Build the matrix — one row per competitor + one row for the user's product/idea (even if hypothetical) + one row for "status quo / do nothing."
4. Find the gap. Look across the matrix for an axis where every competitor is weak or absent and the target segment cares. State it as a testable hypothesis, not a certainty — the gap must tie to evidence from analyze, not intuition.
5. Name the risk of the gap. Is it a gap because it's hard to defend (regulatory, technical), or because nobody's validated demand there yet? Say which.

Write to research/COMPARISON.md:

---
findings_ref: research/FINDINGS.md
created: {ISO-8601}
---

# Competitor Comparison Matrix

| | Pricing | Target Segment | Core Features | Differentiator | Weakness | Market Signal |
|---|---|---|---|---|---|---|
| **{Our product}** | {...} | {...} | {...} | {...} | {...} | {...} |
| {Competitor A} | {...} | {...} | {...} | {...} | {...} | {...} |
| {Competitor B} | {...} | {...} | {...} | {...} | {...} | {...} |
| Status quo / do nothing | {...} | {...} | {...} | {...} | {...} | {...} |

## Positioning Read

**The gap**: {axis where competitors are collectively weak, tied to a
specific evidence point from FINDINGS.md}

**Why it's defensible / why it's risky**: {one or the other, stated plainly}

**Who this positioning is NOT for**: {segment/use-case where the user should
not compete head-on}

Anti-Rationalization

• "We're better across the board" is not a positioning — pick the one axis that matters most to the target segment and lead with it.
• A matrix cell filled with "N/A" for every competitor on your own differentiator axis is worth double-checking — either the axis is mis-chosen or the gap claim is real; verify against FINDINGS.md before either believing it or discarding it.
• Don't compare against strawman competitors — use the ones with real overlapping demand, even if that makes the matrix less flattering.

Evidence

The comparison is complete when the matrix has no ungrounded cells (each traces to a FINDINGS.md entry), and the positioning read names one specific, falsifiable gap plus who it excludes.

Red Flags

• Every competitor axis filled but the "our product" row is vague or aspirational with no evidence behind it.
• A positioning claim that isn't falsifiable ("we're more innovative").
• Gap identified but never checked against whether competitors avoided it deliberately (regulatory/margin reasons) vs. by oversight.