AGENT SKILL

meta-ads-campaign-builder

Build and configure Meta Ads campaigns across all ODAX objectives — from objective selection through audience strategy, placement, budget, and creative launch.

适用职业市场营销财务
使用场景营销推广
License 许可
MIT
最近更新
2026-09-16
Stars
★ 20
Forks
2

01 前言 Front Matter

name
meta-ads-campaign-builder
description
Build and configure Meta Ads campaigns across all ODAX objectives — from objective selection through audience strategy, placement, budget, and creative launch.

02 资源文件 2

03 SKILL.md 正文

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

Ad Copy Structure

Element Guideline
Primary text Under 125 characters (truncation point). Hook → benefit → CTA
Headline Under 40 characters. Clear value prop or offer
Description Under 30 characters. Supporting detail
CTA button Match objective: Shop Now, Learn More, Sign Up, Get Offer

Copy framework: Hook (stop the scroll) → Problem/Pain → Solution/Benefit → CTA

Video rules:

  • First 3 seconds = hook (determines if user watches)
  • Bold, high-contrast captions mandatory (85% of Facebook videos watched without sound)
  • Front-load the value proposition
  • Graphic CTAs outperform text-only CTAs

For creative testing frameworks, use the meta-ads-creative-engine skill.

Output Format

## Meta Ads Campaign Build Plan

### Campaign Overview
- Campaign name: [naming convention]
- Objective: [ODAX objective]
- Conversion location: [Website/App/Instant Forms/etc.]
- Structure: [ASC / Manual 3-campaign / Custom]
- Total monthly budget: $[X]

### Ad Set Configuration
| Ad Set | Audience | Budget | Optimization | Attribution |
|--------|----------|--------|-------------|-------------|
| [name] | [targeting] | $[X]/day | [event] | [window] |

### Audience Strategy
- Prospecting: [approach]
- Retargeting: [approach]
- Exclusions: [who to exclude]

### Creative Plan
| Ad | Format | Hook | CTA | Landing Page |
|----|--------|------|-----|-------------|
| [name] | [format] | [hook concept] | [CTA] | [URL] |

### Launch Checklist
- [ ] Pixel + CAPI verified (EMQ ≥ 6.0)
- [ ] Conversion event configured and firing
- [ ] Audience exclusions set (existing customers from prospecting)
- [ ] Creative meets spec requirements (4:5 vertical, captions on video)
- [ ] Budget sufficient for 50+ conversions/week/ad set
- [ ] Attribution window set (7d click / 1d view default)
- [ ] No changes planned for 5-7 day learning phase

Example

Input: "I sell premium candles DTC, average order value $45, target CPA $18. I'm spending $4K/month on Meta. Currently have Pixel + CAPI active with 7.2 EMQ. Getting about 55 purchases/week. Want to set up a proper campaign structure."

Output:

Meta Ads Campaign Build Plan

Campaign Overview

  • Objective: Sales (Purchase optimization)
  • Conversion location: Website
  • Structure: ASC + Manual retargeting — 55 purchases/week qualifies for ASC. Run ASC for prospecting + manual for retargeting
  • Total monthly budget: $4,000

Campaign Architecture

Campaign 1: ASC — Prospecting (70% = $2,800/mo / ~$92/day)

Setting Configuration
Customer ratio 80% new / 20% existing
Creative 15-20 best-performing ads (diverse formats: static, video, carousel)
Catalog Connected — enable dynamic product ads
Optimization Purchase
Budget $92/day (enough for ~35+ purchases/week at $18 CPA)

Campaign 2: Manual — Retargeting (20% = $800/mo / ~$26/day)

Ad Set Audience Budget Optimization
Hot retarget Website visitors 0-14 days + cart abandoners $18/day Purchase
Warm retarget Website visitors 15-60 days + engagers (50%+ video viewers) $8/day Purchase

Exclude: past purchasers (30 days) from both ad sets to avoid wasted spend.

Campaign 3: Testing (10% = $400/mo / ~$13/day)

Ad Set Purpose Budget Duration
Creative test Test 2-3 new creative concepts against control $13/day 7-10 days per test

Audience Strategy

  • Prospecting (ASC): Country-level (US). Algorithm handles targeting using your strong Pixel + CAPI signals (7.2 EMQ — solid)
  • Retargeting: Website visitors by recency (14-day warmer than 60-day), cart abandoners, video engagers (50%+ view)
  • Exclusions: Upload customer purchase list → exclude from prospecting. Exclude 30-day purchasers from retargeting. Note: detailed targeting exclusions removed March 2025 — use customer list suppression instead

Creative Plan

Ad Format Hook CTA Placement
Hero product 4:5 static image Lifestyle shot — candle in styled room Shop Now Feed, Stories
UGC unboxing 9:16 video (20 sec) "This is the best candle I've ever owned" Shop Now Reels, Stories
Scent carousel Carousel (5 cards) "Find your signature scent" — one scent per card Shop Now Feed
Social proof 4:5 static Customer quote + product image Shop Now Feed
Process video 9:16 video (15 sec) "Hand-poured in small batches" — behind-the-scenes Shop Now Reels

Copy example (UGC ad):

  • Primary text: "This candle completely transformed my living room. The scent lasts 60+ hours." (89 chars)
  • Headline: "Premium Hand-Poured Candles"
  • CTA: Shop Now

Launch Checklist

  • [x] Pixel + CAPI verified (EMQ 7.2 — above 6.0 threshold)
  • [ ] Purchase event configured as primary conversion
  • [ ] Customer list uploaded → excluded from ASC prospecting
  • [ ] 30-day purchaser exclusion on retargeting ad sets
  • [ ] All creative in 4:5 vertical (Feed) and 9:16 (Reels/Stories)
  • [ ] Video ads have captions (85% watch without sound)
  • [ ] ASC budget = $92/day (allows ~35+ purchases/week at $18 CPA)
  • [ ] No changes for 5-7 day learning phase after launch

Guidelines

  • Don't use Traffic when you want Sales. This is the most common Meta Ads mistake. Traffic finds clickers. Sales finds buyers. Always optimize for the deepest meaningful conversion.
  • Don't launch with weak tracking. If EMQ is below 6.0 or CAPI isn't active, fix tracking first. Use the meta-ads-tracking-setup skill before building campaigns.
  • Don't hyper-segment audiences. 2025 Meta algorithm prefers fewer, larger ad sets with 50+ conversions/week. Combine warm + cold in single ad sets. Let the algorithm sort.
  • Don't skip customer exclusions. Detailed targeting exclusions were removed March 2025, but customer list exclusions still work and are essential. Upload your buyer list and exclude from prospecting.
  • Don't ignore learning phase. 5-7 days minimum. Changes during learning (budget >20%, audience, creative swaps, bid changes) reset the clock. Advertisers under 20% learning spend see up to 68% lower CPA.
  • Don't forget captions on video. 85% of Facebook videos are watched without sound. No captions = invisible message.
  • Don't exceed 150 ads in ASC. Hard ceiling. Curate your best creative, don't dump everything in.
  • Instant Experience change (June 2025): No longer counts as a landing page view. Don't include in funnel metrics.
  • Offline Conversions API (May 2025): Discontinued. If someone needs offline conversion tracking, direct to CAPI with action_source: physical_store or system_generated.
  • Cross-references: For tracking setup, use the meta-ads-tracking-setup skill. For creative testing, use the meta-ads-creative-engine skill. For ongoing optimization after launch, use the meta-ads-optimizer skill. For Google campaign building, use the google-ads-campaign-builder skill. For audience deep-dives, read references/audience-strategy.md.
  • Confidence: Budget sufficiency estimates are based on target CPA math. Actual performance depends on creative quality, offer strength, and tracking accuracy. ASC's ~12% CPA improvement is Meta's benchmark — results vary.

04 同仓库更多 Skills

narayan-metaflow
narayan-metaflow/metaflow-marketing-skills

technical-seo-ai-crawler-audit

Diagnose and prioritize the technical issues blocking a site from being crawled, indexed, ranked, and cited. Modern technical SEO has two audiences: Google's crawler/indexer and AI answer engines (GPTBot, PerplexityBot, ClaudeBot, etc.). Both need HTML-visible content and a clean crawl path.

市场营销营销推广
★ 20
narayan-metaflow
narayan-metaflow/metaflow-marketing-skills

seo-reporting-measurement

Build reports that stakeholders trust, rooted in data sources that are actually reliable. Modern SEO measurement has to reckon with a very specific constraint: **the primary data source (GSC) is ~75% incomplete**, and the AI visibility layer is **probabilistic**, not deterministic. Building reports that ignore these realities produces false confidence.

数据分析师市场营销营销推广
★ 20
narayan-metaflow
narayan-metaflow/metaflow-marketing-skills

seo-max-kt

Long-form SEO and AEO methodology bundle; read SEO_AND_AEO_KT.md for the full corpus.

市场营销营销推广
★ 20
narayan-metaflow
narayan-metaflow/metaflow-marketing-skills

seo-keyword-research-intent-mapping

Build a keyword universe that reflects how both humans *and* AI answer engines search. Traditional keyword research looks at standalone terms; modern research also maps the **fan-out sub-queries** AI tools generate when breaking down a user prompt.

市场营销营销推广资料检索与研究
★ 20
narayan-metaflow
narayan-metaflow/metaflow-marketing-skills

seo-aeo-content-strategy

Design content systems that earn both Google rankings *and* AI citations. The core insight: AI answer engines read content differently than humans — they extract the highest-information-gain sentences and treat H2 headings as prompts. The "ski ramp" structure optimizes for both.

市场营销内容创作者视觉设计营销推广
★ 20
narayan-metaflow
narayan-metaflow/metaflow-marketing-skills

prompt-picker

Build a prompt library that reveals how a brand shows up across the full buyer journey in AI search systems. The library is designed for actual LLM visibility tracking, not theoretical SEO keyword research.

Skill与智能体开发
★ 20