---
name: Brand Audit
category: business
works_with: [claude, chatgpt, gemini]
difficulty: plug-and-play
version: 1.1
source: original
tested: yes
---

# Brand Audit

## What it does
Audits your brand the way a stranger experiences it — name, promise, voice, visuals, consistency — and tells you where it's confused, generic, or contradicting itself.

## When to use it
- Before a launch or redesign
- Your brand grew in pieces and might not add up anymore
- You can no longer see your own brand clearly (nobody can, after long enough)

## The skill
```
Act as a brand strategist auditing me as a skeptical outsider. You
have no loyalty to what I've built.

My brand: [NAME + WHAT IT IS]
The promise / one-liner: [WHAT I TELL PEOPLE IT DOES]
Who it's for: [AUDIENCE]
Where it lives: [SITE, CHANNELS, PRODUCTS]
Voice/tone I aim for: [DESCRIBE]
Other brands/projects I run: [LIST — for conflict checking]
Samples: [PASTE REAL EXAMPLES — homepage text, posts, product names]

Audit:
1. FIRST IMPRESSION — Seeing only the name and one-liner cold: what
   would a stranger think this is? Where would they guess wrong?
2. PROMISE CHECK — Does the one-liner make a claim that's specific,
   believable, and different? Rewrite it if it's any of: vague,
   generic, or overclaiming.
3. VOICE CONSISTENCY — Compare my samples against my stated tone.
   Quote lines that break character.
4. ECOSYSTEM FIT — Where my other projects help this brand, where
   they confuse it, and whether they should be visibly connected or
   kept separate.
5. THE GENERIC TEST — Which parts of this brand could belong to any
   competitor if you swapped the logo? Those parts are dead weight.
6. VERDICT — The single biggest brand problem, the single biggest
   strength to double down on, and the 3 fixes in priority order.

Rules: no compliment sandwiches. Quote my own material as evidence
for every criticism. If the name itself is the problem, say so.
```

## Example output
Run for real on TemplatedAI itself, July 2026 — the library auditing its own brand. Condensed excerpt:

> **FIRST IMPRESSION:** "TemplatedAI" cold reads as *yet another AI SaaS* — a stranger's first guess is "a tool that generates templates with AI," not "a curated, honest prompt library." The `[Word]AI` naming pattern is the same pattern every guru product uses.
>
> **VOICE CONSISTENCY:** The About lede said we rebuild prompts "into clean, **tested**, honest templates" — while the homepage stat counter said **0 tested**. The honesty brand's own copy claimed the thing its flags admit hasn't happened.
>
> **THE GENERIC TEST:** Logo-swapped, "clean templates" and "plain language" could be any prompt site. What survives the swap: the tested/untested flags, the "What we stripped" sections, and "Curator, not guru." That's the brand. The name isn't.
>
> **VERDICT:** Biggest problem — copy overclaims, fatal if a critic finds them first, trivially fixable today. Biggest strength — "What we stripped": proof-of-honesty you can't fake.

The overclaims it caught were fixed the same day. The audit works — it drew blood on its own house.

## Tweaks
- Feed it screenshots/exports of real touchpoints, not descriptions
- Rerun after implementing fixes — brands drift back toward mush
- Works on personal brands too; use your channel as the "brand"