# Brand Research Skill

Version 1.0.0 · CC BY 4.0

## Purpose

A staged operating instruction for evidence-led brand research.

## When to use

Use this before naming a positioning territory, drafting a competitive summary, or preparing a brand workshop. It is most useful when source material is scattered across a site, a short interview set, and a few approved documents.

## Inputs

- A decision the research must inform.
- Approved source links, documents, and interview notes.
- The audience, market, and time boundary for the work.
- Privacy and confidentiality limits.

## Process

1. Make a source ledger. Give each source a stable ID and record what it directly supports.
2. Separate observations from interpretations. A repeated product claim is an observation; “the brand owns simplicity” is an interpretation to test.
3. Ask the model to group only the supplied observations into themes. Require source IDs after every theme.
4. Challenge each theme with one disconfirming source, missing-evidence note, or alternative reading.
5. Write a decision brief with the evidence, tensions, recommendation, and open questions kept separate.

## Check

- [ ] Every consequential statement points to an opened source or is marked interpretation.
- [ ] The recommendation names a trade-off, not only a benefit.
- [ ] The output says what evidence would change the conclusion.
- [ ] Confidential material and credentials are absent from the tool input.

## Expected output

A one-page brand-research brief: decision, evidence table, themes, tensions, recommendation, and open questions.

## Limitations

This method improves traceability; it does not establish market truth, replace interviews, or grant permission to use confidential material.