Skip to content

The Gold Standard Method

Every voice system needs a benchmark. Not a description of good writing -- an actual piece of YOUR writing that you can point to and say: "This. This is what it should sound like."

The Gold Standard Method extracts that benchmark from your existing work and turns it into something AI can reproduce.

Pick Your Best 400 Words

Find one piece where your voice is unmistakable. Blog post, newsletter, email to a friend, speech transcript -- format doesn't matter. What matters: if someone who knows your writing read it blind, they'd recognize it as yours.

Keep it short. Under 400 words. You need a piece small enough to analyze sentence by sentence, not skim paragraph by paragraph. The constraint forces depth. A 2,000-word essay hides patterns in volume. A 385-word blog post has nowhere to hide.

One more requirement: the piece should demonstrate range. Not just one note played well, but the full spectrum -- humor and seriousness, personal and universal, specific and abstract. If it only shows one gear, it's a sample, not a standard.

The Analysis Process

Go through the piece line by line. For each sentence, ask: what is this doing, and how is it doing it?

The dimensions to extract:

  • Opening technique. Does it start specific or generic? "Russ rapping through my AirPods on a plane to Toronto" vs. "I was recently thinking about productivity." Specific openings signal a distinct voice. Generic openings signal AI.
  • Metaphor usage. Is there one metaphor sustained throughout, or three competing for attention? Sustained metaphors are a voice signature. Mixed metaphors are a noise signal.
  • Personal-to-universal moves. Where does the piece shift from "my experience" to "your takeaway"? That transition point is often where your voice is strongest.
  • Sentence rhythm. Are lengths varied? Short punches after long setups? Monotone sentence length is the first thing that makes writing feel AI-generated.
  • Vulnerability balance. Does it show humanity without wallowing? "Recovering perfectionist" earns trust. "I was devastated by my failures" loses it.
  • Filler word count. "Basically," "actually," "really," "just," "very." Count them. The target is zero.
  • Hedging language. "I think," "maybe," "perhaps," "sort of." Count them. Also zero.

19 Techniques from One Piece

Here's what this analysis actually produces. From a single 385-word blog post, we extracted 19 distinct techniques:

  1. Specific openings, not generic
  2. One metaphor, sustained (basketball/court through the entire piece)
  3. Personal-to-universal transition
  4. Punchy sentences that land ("Overthinking keeps me benched." -- 4 words.)
  5. Vulnerable without being confessional
  6. Zero filler words
  7. Sentence length variation creating rhythm
  8. Playful-to-serious pivot ("Jokes aside" -- two words, clean gear shift)
  9. Recursive demonstration (the piece about starting IS a start)
  10. Bookend structure (opens with MVP concept, closes with "I'm finally playing")
  11. Tension through contrasts (thinking vs doing, benched vs playing)
  12. Zero hedging language
  13. Specifics over vague descriptors ("18+ hours" not "a long flight")
  14. Active voice dominance
  15. Present tense for urgency
  16. Rule of three in closing
  17. Stakes without melodrama ("where potential goes to die")
  18. Preview with specificity (names exact topics, not "interesting things")
  19. Self-deprecation as credibility builder

Nineteen techniques. One short piece. That's the density of insight you're looking for.

Building the Density Test

From your gold standard analysis, create a scorecard. This becomes the concrete benchmark for every section you write or edit.

Dimension Gold Standard Your Draft
Word count 385 ?
Sentences that could be cut 0 ?
Distinct insights 3-4 ?
Memorable phrases (quotable?) 4-5 ?
Specific details creating mental images 3 ?
Hedging words 0 ?
Filler words 0 ?
Contrasts/tensions 5 ?
Callback to opening in close Yes ?
Numbers instead of vague descriptors 2 ?

The scorecard converts "write well" into "match these specific targets." AI can work with targets. It cannot work with vibes.

What This Produces

A concrete benchmark that replaces subjective judgment with measurable criteria. Instead of "this doesn't sound right," you can say: "the opening is generic (technique #1 violated), there are 3 filler words (technique #6 violated), and every sentence is the same length (technique #7 violated)."

The gold standard doesn't make your voice better. It makes your voice transferable -- to AI, to future writing sessions where you're tired and your instincts are off, to editing passes where you need to know whether something meets the bar or not.

The analysis is a one-time investment. The benchmark lasts the entire book.

Next Steps