meta-ads-campaign-builder

Files (3)

• SKILL.md
• audience-strategy.md · references/
• meta-campaign-types.md · references/

SKILL.md (first lines)

## Process

1. **Define business goal** — Sales, leads, app installs, awareness, or traffic
2. **Select objective and conversion location** — Use the ODAX Selector below
3. **Choose campaign structure** — ASC vs manual (see Decision Tree)
4. **Configure ad set** — Audience, placements, budget, schedule, optimization goal
5. **Build creative** — Ad format, copy, media, CTA
6. **Launch and protect learning phase** — No major changes for 5-7 days

Read `references/meta-campaign-types.md` for detailed setup per objective × conversion location.
Read `references/audience-strategy.md` for complete audience targeting guidance.

## ODAX Objective Selector

**Meta's 6 objectives (ODAX system):**

| Objective | Use When | Optimize For | Typical Conversion Location |
|-----------|---------|-------------|---------------------------|
| **Awareness** | Brand reach, video views, store traffic | Reach, Ad Recall Lift, ThruPlay | N/A (upper funnel) |
| **Traffic** | Driving website/app visits | Link Clicks, Landing Page Views | Website, App |
| **Engagement** | Post engagement, video views, messages | Post Engagement, Video Views, Messages | Facebook, Instagram, Messenger |
| **Leads** | Lead form submissions, calls, sign-ups | Leads, Conversions | Instant Forms, Website, Messenger, Calls |
| **App Promotion** | App installs, app engagement | App Installs, App Events | App (iOS/Android) |
| **Sales** | Purchases, catalog sales, ROAS-driven | Conversions, Value, ROAS | Website, App, Shop, Messenger |

**Common mistake:** Using Traffic when you want Sales or Leads. Traffic optimizes for clicks, not conversions. Meta's algorithm will find clickers, not buyers. Always optimize for the deepest meaningful conversion event.

**Objective alignment rule:** If the goal is revenue → Sales. If the goal is leads → Leads. Only use Traffic if you genuinely just want page visits (rare for performance marketers).

## Campaign Structure Decision Tree

**Q1: Ecommerce with 100+ conversions/week?**
- Yes → **Advantage+ Shopping Campaigns (ASC)** — typically reduces cost per purchase ~12% vs manual
- No → Manual campaign structure

**Q2: How should the manual account be structured?**

Use the **simplified 3-campaign architecture** (2025 best practice):

| Campaign | Purpose | Audience | Budget Share |
|----------|---------|----------|-------------|
| **1. Prospecting** | New customer acquisition | Broad / Lookalikes / Advantage+ | 60-70% |
| **2. Retargeting** | Convert warm audiences | Website visitors, engagers, cart abandoners | 20-30% |
| **3. Testing** | Test new creative, audiences, offers | Varies per test | 10% |

**Why consolidated?** Meta's algorithm in 2025 performs better with fewer, larger ad sets that accumulate 50+ conversions/week. Hyper-segmentation starves ad sets of data.

## ASC (Advantage+ Shopping) Setup

| Setting | Configuration |
|---------|-------------|
| Targeting | Country-level only (no interest/behavior targeting) |
| Customer ratio | Set existing vs new customer budget split (e.g., 70% new / 30% existing) |
| Creative | Up to 150 ads per campaign (use this ceiling strategically) |
| Catalog | Required for dynamic product ads |
| Budget | Enough for 50+ conversions/week minimum |
| Pixel + CAPI | **Must be strong** — ASC relies entirely on algorithm learning from signals |

**When ASC works best:** Ecommerce with 100+ conversions/week, strong Pixel + CAPI, diverse creative library, sufficient budget.

**When to skip ASC:** Low conversion volume (<50/week), weak tracking, need granular audience control, non-ecommerce.

## Ad Set Configuration

### Budget & Bid Strategy

| Strategy | When to Use | How It Works |
|----------|------------|-------------|
| **Highest Volume** (default) | No specific CPA/ROAS target, want maximum results | Meta spends full budget to get most conversions |
| **Cost Cap** | Have a target CPA | Sets maximum average CPA — may underspend if target is tight |
| **Bid Cap** | Have a hard maximum per-result bid | Caps individual auction bids — most restrictive |
| **ROAS Goal** | Have a target ROAS (ecommerce) | Optimizes for return on ad spend target |

**Budget sufficiency rule:** Each ad set needs enough budget for 50+ conversions/week (optimal) or 25+ (minimum). If your target CPA is $50, that means $250-$350+/week per ad set minimum.

**60/30/10 allocation:**
- 60% to best-performing audiences/ad sets
- 30% to scaling tests
- 10% to experimental

**Scaling rule:** Increase budgets by 20% increments maximum. Larger jumps risk resetting learning phase.

### Placement Strategy

| Approach | When | Configuration |
|----------|------|--------------|
| **Advantage+ Placements** (default) | Most campaigns | Let Meta optimize across all placements. Recommended default |
| **Manual placements** | Specific creative constraints | Select specific placements (Feed only, Reels only, etc.) |

**Monitor placement performance:** Even with Advantage+, review placement-level ROAS. If Audience Network drives clicks but no conversions, consider excluding it.

**Creative specs by placement:**
- Feed: 4:5 vertical (best) or 1:1 square
- Stories/Reels: 9:16 full vertical
- Right Column: 1:1 square
- In-stream video: 16:9 landscape
- 4:5 vertical outperforms 1:1 by ~15% in Feed (Billo data)

### Attribution Windows

| Window | Best For |
|--------|---------|
| 7-day click, 1-day view (default) | Most campaigns. Standard attribution |
| 1-day click | Short purchase cycles, low-ticket items |
| 7-day click | High-ticket, longer consideration (removes view-through noise) |

**Critical (October 2025):** `7d_view` and `28d_view` windows removed from Insights API. Don't build reporting dependencies on these.

## Ad Creative Setup

### Format Selection

| Format | Best For | Specs |
|--------|---------|-------|
| Single image | Simple messaging, product shots | 1080x1350 (4:5), <30MB |
| Single video | Storytelling, demos, UGC | 4:5 or 9:16, 15-30 sec, <4GB |
| Carousel | Multiple products, feature walkthrough | Up to 10 cards, 1:1 or 4:5 |
| Collection | Ecommerce catalog browsing | Cover image/video + catalog products |
| Reels-native | Mobile-first vertical video | 9:16, 15-60 sec |

From narayan-metaflow/metaflow-marketing-skills, MIT.

More from narayan-metaflow/metaflow-marketing-skills