funnels-build-campaign

# Funnels Build Campaign - The End-to-End Composite

> *"This is the AI-replaces-the-agency moment. One product brief in. A full launch-ready campaign out."*

This is the **composite** skill that builds an entire direct-response campaign end-to-end. It orchestrates the funnels family for architecture (offer + ladder + sequence) + delegates to the convert-* family for asset authoring (sales page, VSL, OTO, etc.) + ships back the full campaign deliverable.

Expected runtime: 15-30 minutes wall-clock with parallel delegation. Output: ~80-120 pages of campaign artifacts.

## When to Use

Trigger phrases: "build my campaign", "build full campaign", "full campaign build", "launch a new product", "build me a launch", "build me a sales engine", "agency replacement", "end-to-end campaign", "campaign architecture", "build the whole thing", "one-shot launch", "30-minute campaign build", `/funnel build-campaign <product>`.

Use when:
- The user has a product (or a clear product concept) and wants the full launch-ready campaign delivered
- The user has 15-30 minutes of compute time and wants to fan out parallel work
- The user needs the "agency-replacement" deliverable - copy, page structure, email sequences, the works

Do NOT use for:
- Single-skill needs (use the specific skill - `funnels-offer`, `convert-sales-page`, etc.)
- URL audit / CRO teardown (use `market-funnel`)
- Iterative one-piece-at-a-time work (use the relevant single skill)
- A full-launch *campaign strategy* across channels (use `market-launch` for channel-level orchestration)

## Inputs

1. **Product description** - what's being sold, who buys it, what problem it solves
2. **Audience** - the One Person at Rung 2 (core offer level)
3. **Existing assets** - anything the user already has (lead magnets, course content, swipe files, recordings)
4. **Sales model preference** - webinar-led OR launch-led (informs Phase 1 path)
5. **Target price for core offer** - informs ladder pricing
6. **Time horizon** - when does the campaign go live?
7. **Capacity** - can the user fulfill DFY services (capacity-limited) or only digital products?

## The 4-Phase DAG

The composite runs in 4 phases, executed STRICTLY SEQUENTIALLY at the phase level. Within each phase, internal parallelism is capped at `max_concurrent_children: 4` (per-phase cap, not global). Phases 2 and 3 do NOT run concurrently with each other - Phase 3 starts only after Phase 2 completes. This avoids the global-pool starvation issue where P2 (4 children) + P3 (concurrent) would consume the entire delegation pool.

### Phase 1 - Architecture (sequential, ~5 min)

**Job:** lock the offer, ladder, and conversion sequence model. Phase 2 + 3 depend on Phase 1.

| Step | Skill | Output |
|------|-------|--------|
| 1.1 | `funnels-offer` | The Irresistible Stack - core offer with value-equation engineering, mechanism, deliverable |
| 1.2 | `funnels-offer-stack` | Bonus stack - 5–7 named bonuses with objection mapping |
| 1.3 | `funnels-offer-pricing` | Pricing levers - anchor + contrast + payment plan + decoy if multi-tier |
| 1.4 | `funnels-offer-guarantee` | Guarantee statement in Donahoe Voice |
| 1.5 | `funnels-ascension-ladder` | 5-rung ladder map with bridge offers |
| 1.6 | `funnels-lead-magnet` | Rung 0 magnet design |
| 1.7 | `funnels-tripwire` | Rung 1 tripwire design + front-end revenue stack (bump + OTO + downsell) |
| 1.8 | **Branch decision:** webinar-led OR launch-led | |
| 1.8a | If webinar: `funnels-conviction-webinar` | 4-stage webinar script |
| 1.8b | If launch: `funnels-cart-cycle-launch` | 3-stage cart-cycle launch architecture |

**Phase 1 output:** the architecture document - offer + ladder + sequence model, all locked.

### Phase 2 - Asset Authoring (parallel via delegate_task, ~10-15 min)

**Job:** convert-* pack delegations for each asset. Fans out 7-9 parallel children.

Phase 2 receives the locked architecture from Phase 1 and delegates to convert-* skills via `delegate_task`. Each delegate gets:
- The locked offer (from Phase 1)
- The relevant section of the architecture
- The audience + voice rules
- Any existing assets the user supplied

Cap parallelism at 4 children at a time (per playbook risk-mitigation: avoid context-overflow on parallel orchestration).

| Asset | Skill (delegated) | Receives |
|-------|------------------|---------|
| Lead-magnet opt-in page | `convert-lead-magnet-page` | magnet name, audience, opt-in form fields |
| Bridge page (advertorial) | `convert-bridge-page` | ad-to-offer warm-up copy |
| Webinar registration page (if webinar-led) | `convert-webinar-reg-page` | webinar promise, presenter credibility |
| Sales page | `convert-sales-page` | full offer + stack + guarantee + price |
| VSL script + page | `convert-vsl` + `convert-vsl-page` | 18-min VSL aligned to the offer |
| OTO page | `convert-oto-page` | OTO offer (post-tripwire AND post-core) |
| Downsell page | `convert-downsell-page` | OTO recovery offer |
| Thank-you page | `convert-thank-you-page` | post-purchase delivery + referral CTA |

Each asset comes back with structured brief + clean HTML/CSS reference + Donahoe Method audit footer.

### Phase 3 - Sequence Authoring (parallel via delegate_task, ~5 min - runs AFTER Phase 2 completes)

**Job:** funnels family email sequences.

| Asset | Skill | Output |
|-------|-------|--------|
| Welcome / nurture sequence (5-day Story-Drip) | `funnels-story-drip` | 5 emails with Donahoe Voice |
| Daily pivot ongoing sequence (10-15 emails) | `funnels-daily-pivot` | rolling daily emails for mature-list cadence |
| Cart-window emails (if launch-led) | `funnels-daily-pivot` (cart-window mode) | 5-7 emails through cart cycle |
| Webinar sequence (if webinar-led) | `funnels-daily-pivot` | reg → reminder → live → replay → cart-close |

### Phase 4 - Audit Pass (sequential, ~3-5 min)

**Job:** verify the campaign is structurally sound + every asset passes the Donahoe Method.

| Step | Skill | Output |
|------|-------|--------|
| 4.1 | `funn