Creative brief development

Treat brief development as a v0 → v1 → v2 cadence, not a one-time event

Most teams write the brief once and ship. Brief development means versioning. v0 is the rough draft; v1 is the brief that ships the first campaign; v2 incorporates what the first campaign taught you; v3 incorporates the second campaign's lessons. After 6-8 versions the brief converges into a brand-and-category operating doc. Set up the version cadence at the start; don't try to retrofit it later.

Develop v0 in 60 minutes - the rough draft

v0 is permissive. Goal, audience, angle, archetype, do-nots, references, deliverables - all present but not yet refined. Don't over-invest in v0; you'll learn more from shipping it than from polishing it. The cost of imperfect v0 is small; the cost of NOT shipping the first campaign is large. Ship v0 fast.

Run the first sprint, then capture what you learned in a 'lessons' block

After sprint 1, write a 'lessons' block at the bottom of v0: what worked, what didn't, what surprised you. Top-3 performers. Bottom-3 performers. The audience cut that converted vs the one that didn't. This block is the raw material for v1; without it, the brief never improves.

# Lessons block format:
Sprint 1 lessons (Jul 14 - Jul 28):
- Top performer: customer-testimonial archetype (CTR 2.1%, CVR 4.3%)
- Bottom performer: founder-to-camera (CTR 0.6%) — KILL
- Surprise: 'taste' angle outperformed 'performance' 3:1
- Audience: 'currently uses generic' cut converted at 2x base

Develop v1 by tightening the angle and adding 1-2 do-nots

v1 is the lessons-informed brief. The angle that overperformed becomes a stronger constraint ('lead with taste, not performance'). The archetype that underperformed becomes a do-not ('no founder-to-camera - confirmed cohort fatigue'). The audience cut that converted gets sharpened; the one that didn't gets removed. v1 should be SHORTER than v0 - tighter constraints mean fewer words.

Build a 'do-nots library' that survives across briefs

Some do-nots are campaign-specific ('no Christmas references in a summer campaign'). Others are brand-level ('we never use stock photography'). Separate the two. The brand-level do-nots go into a 'do-nots library' that gets pulled into every brief automatically. The campaign-specific do-nots stay in the brief. Over 6-12 months, the library becomes one of the most load-bearing brand assets you have.

Develop archetype expertise by tracking which archetypes win across briefs

After 4-6 sprints, you'll have data on which hook archetypes consistently win for YOUR brand. Maintain an archetype scorecard: archetype name, win rate, average CTR, average CVR, last shipped date. Brief v3+ should default to the proven archetypes and only experiment with new ones when there's a specific reason. The archetype scorecard is the highest-leverage piece of compounding intelligence.

# Archetype scorecard:
- Customer testimonial · 9 shipped · win rate 67% · avg CTR 1.8%
- Problem→solution · 12 shipped · win rate 58% · avg CTR 1.6%
- Founder-to-camera · 4 shipped · win rate 0% · DO NOT USE
- Day-in-the-life · 2 shipped · insufficient data

Develop AI-readability into the brief from v3 onward

Once the brief is stable (v3+), engineer it for AI receivers. Three changes: explicit naming (no 'you know our voice'), heavier do-nots (AI defaults to category-average without them), named archetypes (no 'engaging hook'). The AI-readable brief is a brief humans also read better - the discipline of naming everything forces decisions out into the open.

Develop the brief into team operating knowledge by v6+

After 6-8 sprints, the brief stabilizes. New team members read it and produce on-strategy work without a meeting. AI generators produce closer-to-finished output. At this point, the brief IS the brand's creative operating system. Maintain it like infrastructure: version it, audit it quarterly, retire stale do-nots, add new archetype data.