AND Creative Strategy Engine — The 4-Week Roadmap Process

AND builds the system that pairs production fluency with advertising fluency at scale. The engine handles everything mechanical. The strategist owns the calls that matter. One visual 4-week roadmap per client, every month, same shape every account.

Most agencies sell production. We sell strategy backed by production. The leverage is at the strategist's desk. The engine strips away the work that doesn't require taste or judgment so the strategist can spend their time on the work that does.

The AND Approach — Six Principles

01 · The Gap

Advertising fluency is the edge

Production fluency is table stakes. The agencies that win pair production with the ability to read a buyer, pick a hook, and know what makes someone stop, lean in, and convert.

02 · The Territory

Cover the psychological map

Every macro persona contains 3 distinct micros, 4 angles each, 4 vehicles each. 48 distinct ads per macro at full coverage. The Coverage Map is the long-arc territory we build over months.

03 · The Math

Tier sizing is economics-driven

20 / 50 / 100 assets per month are not arbitrary. Each tier maps to the coverage pace, persona depth, and testing capacity the account economics support. Throughput follows strategy.

04 · The Levers

Coverage and vehicle mix, equally

Coverage expands psychological territory. Vehicle mix scales what's working in market right now. Engine surfaces both every cycle. Strategist picks which to push this month based on signal.

05 · The Leverage

Strategist judgment is the asset

The engine doesn't replace the strategist. It strips the mechanical work so judgment becomes the whole job. The Bet, the test pick, the kill-or-coach call, the taste on the final hook.

06 · The Cadence

Read every week, ship every day

L90 sets direction, L30 confirms trend, L7 surfaces leading edge. Production runs five days a week, staggered across clients. One visual 4-week roadmap per client per month.

The Coverage Model — Math and Terms

Macro Persona

broad archetype

Micro Personas

distinct cognitive biases

Angles

strategic frames per micro

Vehicles

formats per angle

Distinct Ads

full coverage of 1 macro

Macro = the broad buyer archetype. Most accounts have 3-5. Micro = a sub-segment of a macro with a distinct cognitive bias, trigger, and hook. Angle = the strategic frame (price comparison, study count, mechanism, peer proof). Vehicle = the format (podcast cut, tweet static, UGC, animation, founder VO). Coverage Map = the full matrix per client, maintained by the engine in Milanote, built incrementally over months. Creative Jobs = the three things any creative does: Borrow Trust (TOF), Show the Habit (MOF), Give Them a Reason Today (BOF).

Tier Architecture — Per-Month Coverage Activation

Tier	Assets / Mo	Hero Macro Coverage	Net New Test	Pace
T1 · Iterate	20	1 macro, partial (~12-14 ads · 2 micros × 2 angles × 3 vehicles) + 0-1 supporting (3-5 ads)	3 (new micro inside hero)	Rotate macros monthly · full Coverage Map cycled in ~6 months
T2 · Scale	50	1 macro, full coverage (48 ads · 3 × 4 × 4)	2-3 (new macro or micro)	One macro per month · cycled in 3-5 months
T3 · Velocity	100	2 macros, full coverage (96 ads)	4 (2 micros or 1 macro)	Two macros per month · cycled in ~2.5 months

Data Windows — Read All Three, Every Cycle

L90

Direction

For: Account ROAS / CPA baselines. Top spend creatives. Format mix at scale.

L30

Trend

For: Confirming direction. Which L90 winners are still working, which fatigued.

Leading Edge

For: Emerging signal. What's breaking through this week. What just died.

The 4-Week Process — Week 4 Presents Next Month's Roadmap

Week 1

Execute Month N · Wave 1 (brief locked in Wk 4 of Month N-1)

Read

L7 of last week's ships. Engine surfaces the summary.

Decide

Pod call locks Wave 1 production. Brief is already approved.

Produce + Ship

Production runs Mon-Fri, staggered across clients. Ship day varies per client cadence. Motion-tagged on ship.

Week 2

Reinforce · double down on Wave 1 signal

Read

L7 read. Engine flags hook performance.

Decide

Pod call adjusts Wave 2. Add to No New Production list if patterns are clear.

Produce + Ship

Mon-Fri parallel. More variants of Wave 1 winners. Mid-month visual recap sent to client.

Week 3

Scale + Test · net new persona test ships

Read

L7 + L30 read. Engine flags trend direction and persona signal.

Decide

Lock Net New test vehicle (winning format from L30). Scale top Wave 1-2 variants.

Produce + Ship

Mon-Fri parallel. Net new persona assets ship in Wave 3.

Net New Test Protocol

Pick a new micro inside a Hero macro, or a net new macro from the Coverage Map. Use a winning vehicle from L30 (hold one variable constant). Size at 15% of monthly assets (T1=3, T2=8, T3=15). Ship in 2 waves (Wk 1 baseline, Wk 3 double down). Verdict Wk 4: Scale (CPA within 30% of hero, ROAS within 20%) · Extend (close, 1 more month) · No New Production (CPA >2x or ROAS <0.5x with adequate spend).

Week 4

Push final wave · Present Month N+1 roadmap

Read

Full month review. Engine pulls L90 + L30 + L7 + Coverage Map signal.

Decide

Net New verdict. Strategist writes Month N+1 Bet, picks next macro/micro to activate, updates No New Production list.

Build

Month N+1 roadmap assembled. Engine fills template. Strategist edits human-owned sections.

Present

Last review of the month. Strategist walks client through Month N+1 roadmap. Approved so Month N+1 starts running Day 1.

Ship

Wave 4 of current month + month-end visual report.

Automation Map — What's Auto, What's the Strategist

Step	Owner	Tool · Trigger	What Happens
Motion signal pull (L90/L30/L7)	Auto	Motion MCP · Mon 9am	Three CSVs per client land in /Claude Outputs/<client>/motion/.
Brand inputs ingest (reviews, social, support)	Auto	Python scraper · monthly	Reviews, Reddit, comments, support tickets pulled to /inputs/.
Coverage Map build (macros/micros/angles/vehicles)	Semi	Claude prompts · quarterly	Engine drafts the 48-ad map. Strategist edits voice, confirms distinctness, locks in Milanote.
Signal read (winners, kill candidates, trends, emerging)	Auto	Python · weekly	Top performers, No New Production candidates, L90→L30→L7 deltas, L7 emerging. Output to brief draft.
Strategic read of the signal	Human	Strategist · pod call Wk 4 Mon	Interprets meaning. What does this signal mean for the brand right now.
The Bet for the month	Human	Strategist · Wk 4 Tue	One sentence. What we're betting on this month. Strategist writes from scratch.
Coverage Map activation + asset math	Semi	Claude + Python · Wk 4 Tue	Engine recommends slice to activate, applies tier matrix to get per-week counts. Strategist confirms.
Hook curation per week	Semi	Claude + strategist · Wk 4 Wed	Engine suggests 3 hooks per week tied to micros activated. Strategist picks final.
Net New Test + No New Production picks	Human	Strategist · Wk 4 Tue	Strategist picks persona to test, decides what's truly out vs needs re-brief.
Visual brief assembly + presentation	Semi	HTML template + strategist · Wk 4 Wed-Thu	Template auto-populates. Strategist picks visual assets and presents to client.

No New Production

Replaces "Kill List"

We do not turn off existing ads. We do stop producing new variants of macros, micros, angles, or vehicles that L30 signal says aren't earning more investment. Existing inventory keeps running until the buyer decides otherwise. Each roadmap lists what's on No New Production this month with L30 evidence. Strategist can move a category back into production the following month with a re-brief.