Two kinds of playbook — match the ask.

1) AUTOMATION PLAYBOOKS (continuous rules — apply via apply_playbook / enable_autopilot) These run every cron tick within hard guardrails (daily action ceilings, ≤20% budget steps, learning-phase skip): - Bleeder Killer — pause ads whose CPL is ≥2x average with no path to recovery. - Winner Scaler — auto-raise budget ~15% on ads with CPL < 70% of average and healthy engagement. - Creative Refresher — flag/queue replacements as fatigue sets in. - Budget Optimizer — rebalance spend toward the better-performing ad sets. - A/B Auto-Graduate — promote the winning variant once it's statistically ahead. - Full Autopilot — the complete package (pause + scale + refresh + rebalance + graduate). Pick the narrowest playbook that solves the user's problem; default new automations to "suggest"/ask mode unless the user explicitly wants auto-execute. Never bypass the guardrails with a force override.

2) TACTICAL CALENDARS (time-boxed, multi-week — schedulable templates) Dated checklists for a campaign window, not always-on rules: - BFCM / peak season — pre-season audience warming and creative build, in-event budget ramp and bid strategy, post-event retargeting and winback. - Product launch — teaser -> launch -> sustain sequencing with creative volume by phase. - Lead-gen ramp — broad prospecting -> proven offer -> scale ladder. - Recovery — for an account that broke: diagnose first (see the diagnose skill), stabilize spend, rebuild on what survives. - Scale ladder — the $1k -> $10k staged plan from the scale skill, as a week-by-week checklist. Each stages the right move at the right time and references the underlying skill (scale, targeting, creative) for the how. Use a calendar when the goal has a deadline/window; use an automation playbook when the goal is ongoing efficiency.

The templates

BFCM / Peak Season

A 4-week build-up, in-event ramp, and post-event harvest for Black Friday / Cyber Monday.

Week 1 · Warm & build (T-3 weeks)

→Launch broad prospecting to seed pixel signal and refill retargeting pools.
→Brief and produce 4-6 offer-led concepts (the offer is the hook at BFCM).
→Exclude existing customers from prospecting; build/refresh LAL from purchasers.

Week 2 · Test offers (T-2 weeks)

→A/B the actual BFCM offer angles cheaply on statics before committing video spend.
→Lock 3-5 winning creatives; kill anything below the 7/10 bar.
→Confirm pixel/CAPI + catalog sync are healthy (run the audit skill) — a break now is catastrophic.

Week 3 · Ramp (event week)

→Scale winners vertically +20% every 1-2 days (faster cadence is OK during peak demand).
→Duplicate winners horizontally into fresh audiences as frequency climbs.
→Raise daily budget caps ahead of the spike so delivery isn't throttled mid-event.

Week 4 · Harvest (post-event)

→Pivot budget to retargeting the warm BFCM traffic that didn't convert.
→Pull losers fast; reallocate to the few creatives still under target CPA.
→Capture learnings: which offer/angle won — feeds next season.

Product Launch

Teaser → launch → sustain sequencing for a new product or offer over 4 weeks.

Week 1 · Teaser

→Run awareness/engagement creative to build a warm audience before launch.
→Produce launch creative across 3-4 distinct angles (problem, transformation, proof, novelty).

Week 2 · Launch

→Go live with conversion campaign, one concept per ad set, broad prospecting.
→Retarget the teaser audience with the strongest offer creative.
→Watch learning phase — don't touch budgets until ~50 conversions land.

Week 3 · Prove & scale

→Confirm winners (CPL < 0.7x avg, freq < 2.5) and scale vertically +20%.
→Cut underperformers; double production on the winning angle.

Week 4 · Sustain

→Introduce fresh creative to fight fatigue; rotate in new hooks.
→Expand horizontally to new audiences/geos with the proven concept.

Lead-Gen Ramp

Broad prospecting → proven offer → scale ladder for a lead-gen account over 4 weeks.

Week 1 · Establish baseline

→Confirm the real conversion event (promoted_object) and target CPL from unit economics.
→Launch 2-3 broad ad sets, one concept each; budget sized to exit learning in ~1 week.

Week 2 · Find the offer

→Test angle > offer > persona; statics first to find the message cheaply.
→Identify the ad set beating target CPL with healthy frequency.

Week 3 · Scale the winner

→Run the scale eligibility gates; scale vertically +20% every 2-3 days.
→Add 1-2 broader ad sets to keep prospecting wide.

Week 4 · Diversify

→Expand horizontally into LAL/geo; introduce a bid cap once CPA is proven.
→Queue creative refreshes against the first signs of fatigue.

Account Recovery

Diagnose-first stabilization for an account that broke — 3 weeks back to health.

Week 1 · Diagnose & stop the bleed

→Run the diagnose skill — rule out tracking breakage FIRST before touching creative.
→Pause clear bleeders (zero leads at €20+, CPL > 3x avg); do NOT mass-pause on a tracking break.
→Stabilize spend at a level the surviving winners can hold.

Week 2 · Rebuild on what survives

→Consolidate budget onto the ad sets still hitting KPI.
→Produce diverse, pain-led concepts to replace fatigued creative.

Week 3 · Re-scale carefully

→Re-enter scaling only after 3 stable days; +20% steps, no jumps.
→Monitor frequency and CPM for the saturation that caused the original break.

$1k → $10k Scale Ladder

The staged spend ladder from the Scale skill, as a week-by-week checklist.

Week 1 · Stage 1 ($1k–$2.5k)

→Vertical 20% scaling on confirmed winners; add 1-2 broader ad sets.
→Lock 3-5 winning creatives; halt scaling if CPA crosses 1.5x target.

Week 2 · Stage 2 ($2.5k–$5k)

→Horizontal scaling via LAL expansions and geo duplication.
→Introduce Advantage+ Shopping alongside manual campaigns; track MER company-wide.

Week 3 · Stage 3 ($5k–$10k)

→Bid-cap (Tichenor) and Target-ROAS (Faris) diversification.
→4+ new creative concepts weekly; layer brand/awareness at 10-15% of budget.