Gold Standard Reference Template¶

Your gold standard is the benchmark for every piece of writing AI generates for your book. Find your best 300-500 words -- a blog post, email, speech, article -- where your voice is unmistakable. Paste it in. Analyze it sentence by sentence. Extract what makes it you.

The real system extracted 19 techniques from 385 words. Yours might surface 8 or 25. The number doesn't matter. The specificity does.

How to Use: 1. Paste your best writing sample below 2. Analyze each technique you find (the more specific, the better) 3. Build the density test from your sample's metrics 4. Extract quotable lines as benchmarks 5. Reference this file from every other voice system file

# Gold Standard Reference

> The benchmark for your book's writing quality. Every AI output gets measured against this.

## Your Best Writing (~400 Words)

> [PASTE YOUR BEST PIECE HERE -- blog post, email, speech, article.
> Should be 300-500 words where your voice is unmistakable.
> Pick something where people said "that sounds like you."
> Not your most polished piece -- your most *you* piece.]

---

## Technique Analysis

Analyze your piece sentence by sentence. What makes it yours? Be ruthlessly specific -- "good writing" isn't a technique. "Short declarative sentence after a long compound one" is.

### 1. Opening Technique
**What it does:** [How does your piece open? Specific detail? Question? Story? Provocation?]
**Pattern:** [Describe the replicable pattern. e.g., "Concrete sensory detail before any abstraction"]
**Example from your piece:** [Quote the opening]

### 2. Metaphor Usage
**What it does:** [Single sustained metaphor? Brief analogies? No metaphors at all?]
**Pattern:** [e.g., "One sustained metaphor throughout" or "Brief analogies per section, never mixed"]
**Example from your piece:** [Quote]

### 3. Personal to Universal
**What it does:** [How do you move from personal experience to universal insight?]
**Pattern:** [e.g., "Specific anecdote -> 'The more I think about it...' -> universal truth"]
**Example from your piece:** [Quote the transition]

### 4. Sentence Rhythm
**What it does:** [How do you vary sentence length? Where do short punches land?]
**Pattern:** [e.g., "Short punch after long setup" or "Builds intensity toward a one-line close"]
**Example from your piece:** [Quote a short/long pair]

### 5. Vulnerability Balance
**What it does:** [How much do you reveal? How do you handle admissions?]
**Pattern:** [e.g., "States vulnerability once, moves on. Never dwells or seeks sympathy."]
**Example from your piece:** [Quote]

### 6. Closing Technique
**What it does:** [How do you end? Callback? Declaration? Question?]
**Pattern:** [e.g., "Callbacks to opening image" or "Single declarative sentence, no summary"]
**Example from your piece:** [Quote the close]

### 7. [YOUR TECHNIQUE]
**What it does:** [...]
**Pattern:** [...]
**Example from your piece:** [...]

### 8. [YOUR TECHNIQUE]
**What it does:** [...]
**Pattern:** [...]
**Example from your piece:** [...]

[Add as many as you find. The real system extracted 19 from 385 words.
Common ones to look for: hedging habits, humor style, authority signals,
specificity patterns, tense preferences, voice (active/passive), contrast
structures, list rhythms, paragraph length patterns.]

---

## Density Test

Measure your gold standard sample. These become your targets.

| Question | Your Piece | Draft Target |
|----------|-----------|-------------|
| Word count | [X] | Match this density |
| Sentences that could be cut | [should be 0] | 0 |
| Distinct insights | [count them] | Match or exceed |
| Memorable phrases | [count them] | Match or exceed |
| Specific details (names, numbers, places) | [count] | Match or exceed |
| Hedging words (maybe, perhaps, sort of) | [should be 0] | 0 |
| Filler words (basically, actually, very) | [should be 0] | 0 |
| Contrasts/tensions that create energy | [count] | Match or exceed |
| Callbacks (ending references opening) | [yes/no] | Yes |

**The Rule:** If your gold standard does it in [X] words, your AI drafts don't get [2X] to say the same thing. Depth comes from insight, not length.

---

## Quotable Lines

List 3-5 lines from your piece that someone might highlight, screenshot, or quote back to you.

1. **"[YOUR LINE]"** -- Why it works: [brief note on technique]
2. **"[YOUR LINE]"** -- Why it works: [brief note]
3. **"[YOUR LINE]"** -- Why it works: [brief note]
4. **"[YOUR LINE]"** -- Why it works: [brief note]
5. **"[YOUR LINE]"** -- Why it works: [brief note]

Use these as your benchmark for punchy, memorable prose. If a draft section has zero quotable lines, it's not landing hard enough.

---

## Apply This Standard

When reviewing any AI-generated draft:

1. **Is the opening as specific as yours?** If generic, rewrite.
2. **Is there one sustained metaphor, or three mixed ones?** Pick the strongest.
3. **Does it move from personal to universal?** Your experience -> reader's takeaway.
4. **Could you cut 20%?** If yes, you're not at gold standard density.
5. **Would anyone quote a line?** If not, the prose isn't landing.

---

*This is your bar. Hit it.*

Adapted from the gold standard reference used for Blueprint for An AI-First Company. The original extracted 19 techniques from 385 words and was referenced in every writing and editing session across 81,000 words.