菜单
Post-cookie first-party data collection, Enhanced Conversions, server-side tracking, consent mode, and privacy-preserving measurement. Use when setting up Enhanced Conversions, server-side tagging, consent mode v2, CDP integration, or planning for cookie deprecation.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Add conversion measurement to a live ChatGPT Ads campaign - choose between the OpenAI Conversions API (server-side) and the JavaScript Pixel (client-side), define the events that matter, and roll out without delaying launch. Use when setting up ChatGPT Ads tracking, CAPI vs pixel decisions, measuring leads/signups/purchases from ChatGPT Ads, or moving from CTR optimization to conversion-based optimization.
Build, review, and scale a ChatGPT Ads campaign from your existing Google/Meta/Microsoft text ads and website. Use when launching ChatGPT Ads, building a ChatGPT Ads campaign workbook, repurposing existing search/social ads for ChatGPT, writing context hints (not keywords), or scaling to 100+ upload-ready ads. Covers the full 9-step golden-path workflow with copy-paste prompts.
Comparing ad-platform performance with CRM; attribution gaps; 'ad platform shows conversions but CRM shows zero'; UTM/pixel issues; last-click vs data-driven. Covers data-source hierarchy and gap diagnosis.
Defining GTM metrics, pipeline efficiency, AI cost metrics, TTFV, CAC/LTV/NRR, magic number, attribution models, weekly review cadence.
Multi-platform product launches: Product Hunt, Hacker News, BetaList, AppSumo, waitlist, launch day, multi-channel rollout — pre-launch through post-launch.
SPEAR loop for any non-trivial agent task — Scope, Plan, Execute, Assess, Resolve. Prevents the strong-start-weak-finish failure mode by forcing an explicit MECE rubric and a Plan→Execute→Assess inner loop until the rubric passes.
| name | first-party-data-strategy |
| description | Post-cookie first-party data collection, Enhanced Conversions, server-side tracking, consent mode, and privacy-preserving measurement. Use when setting up Enhanced Conversions, server-side tagging, consent mode v2, CDP integration, or planning for cookie deprecation. |
Build a comprehensive first-party data strategy for the post-cookie era. Covers Enhanced Conversions setup for Google and Meta, server-side tracking architecture, Customer Data Platform (CDP) integration, Google Consent Mode v2 configuration, first-party cookie strategies, privacy-preserving measurement, and data clean rooms.
1. Inventory current data collection points:
□ Website forms (signup, contact, download, purchase)
□ Email/newsletter subscriptions
□ Account registrations
□ Customer support interactions
□ In-app behavior and events
□ CRM records (Salesforce, HubSpot, etc.)
□ Loyalty/rewards program data
□ Survey and feedback responses
□ Offline purchases (POS)
2. Assess data quality:
□ Email match rate (% valid, deliverable emails)
□ Phone number coverage (% with valid phone)
□ Address data completeness
□ Data freshness (% updated in last 90 days)
□ Consent status (% with valid consent)
3. Identify gaps:
□ Anonymous visitors with no identifier (typically 95-98%)
□ Cross-device identity resolution
□ Offline-to-online attribution
□ Consent coverage by geography (GDPR, CCPA, etc.)
4. Score readiness:
- Level 1 (Basic): Email list only, no consent management
- Level 2 (Developing): CRM + consent, basic pixel tracking
- Level 3 (Mature): CDP, server-side tracking, consent mode
- Level 4 (Advanced): Data clean rooms, privacy-preserving ML
Google Enhanced Conversions:
Purpose: Match conversion data using hashed PII (email, phone, address)
Result: 5-15% more attributed conversions
Setup via GTM:
1. Create conversion action in Google Ads
2. In GTM: Edit conversion tag → Enable Enhanced Conversions
3. Map user-provided data:
- email → sha256(lowercase(trim(email)))
- phone → sha256(normalize(phone)) — E.164 format
- first_name → sha256(lowercase(trim(first_name)))
- last_name → sha256(lowercase(trim(last_name)))
- street → sha256(lowercase(trim(street)))
- city → sha256(lowercase(trim(city)))
- region → sha256(lowercase(trim(region)))
- postal_code → sha256(trim(postal_code))
- country → sha256(lowercase(trim(country)))
4. Source: Data layer, CSS selector, or JavaScript variable
5. Test with Tag Assistant → verify hashed data sent
Meta Conversions API (CAPI):
Purpose: Server-side event tracking, bypasses ad blockers
Result: 15-30% more attributed conversions
Setup:
1. Generate System User access token in Business Manager
2. Server sends POST to: https://graph.facebook.com/v21.0/{pixel_id}/events
3. Include user data parameters:
- em: sha256(email)
- ph: sha256(phone)
- fn: sha256(first_name)
- ln: sha256(last_name)
- external_id: sha256(user_id)
- fbc: _fbc cookie value
- fbp: _fbp cookie value
- client_ip_address: user IP
- client_user_agent: user agent string
4. Deduplicate with browser pixel using event_id
5. Test with Events Manager → Test Events tab
Architecture:
Browser → First-Party Domain → Server-Side GTM → Platform APIs
↓
Your Server
↓
[Google Ads CAPI]
[Meta CAPI]
[LinkedIn CAPI]
[TikTok Events API]
[Reddit CAPI]
Components:
1. First-Party Domain Proxy:
- Subdomain: track.yourdomain.com
- Proxies to GTM server container
- First-party context = longer cookie lifespan
2. GTM Server Container:
- Cloud Run, App Engine, or Stape.io
- Receives client-side events
- Enriches with server-side data (CRM, order details)
- Forwards to platform APIs
3. Platform API Clients:
- Google: Measurement Protocol (GA4) + Enhanced Conversions API
- Meta: Conversions API
- LinkedIn: Conversions API
- TikTok: Events API
- Reddit: Conversions API
Benefits:
- Bypasses ad blockers (first-party domain)
- Longer cookie lifespan (first-party cookies = 1-2 years vs 7 days ITP)
- Data enrichment before sending (add LTV, customer segment)
- Single implementation, multi-platform distribution
- Reduced page weight (fewer client-side tags)
Cost:
- Stape.io: $20-100/month (managed hosting)
- Cloud Run: $50-200/month (self-hosted, pay per request)
- App Engine: $30-150/month (Google-managed)
Consent Mode v2 is REQUIRED for:
- EEA (European Economic Area) users since March 2024
- Remarketing lists, conversion measurement, personalization
Two modes:
Basic Mode:
- Tags don't fire until consent is granted
- No data collection without consent
- Simplest to implement
- Loses all unconsented user data
Advanced Mode (Recommended):
- Cookieless pings sent even WITHOUT consent
- Google uses ML modeling to estimate conversions
- Recovers 50-70% of lost conversion data
- Requires proper implementation
Implementation:
1. Install a Consent Management Platform (CMP):
- Cookiebot, OneTrust, Termly, or custom
- CMP must be Google-certified
2. Set default consent state (before CMP loads):
gtag('consent', 'default', {
'ad_storage': 'denied',
'ad_user_data': 'denied',
'ad_personalization': 'denied',
'analytics_storage': 'denied',
'wait_for_update': 500 // ms to wait for CMP
});
3. Update consent when user interacts with CMP:
gtag('consent', 'update', {
'ad_storage': 'granted',
'ad_user_data': 'granted',
'ad_personalization': 'granted',
'analytics_storage': 'granted'
});
4. Consent parameters:
| Parameter | Controls |
|-----------|----------|
| ad_storage | Cookies for advertising (Google Ads, Floodlight) |
| ad_user_data | Sending user data to Google for ads |
| ad_personalization | Personalized advertising (remarketing) |
| analytics_storage | Cookies for analytics (GA4) |
| functionality_storage | Cookies for functionality (language, region) |
| personalization_storage | Cookies for personalization (video recs) |
| security_storage | Cookies for security (authentication) |
5. Verify in Google Ads:
Tools → Diagnostics → check consent mode status
GA4 → Admin → Data Streams → check consent signals
CDP-to-Ad Platform workflow:
CRM/CDP Ad Platforms
┌─────────┐ ┌──────────────┐
│ Customer │──audience──→ │ Google Ads │ (Customer Match)
│ Data │ sync │ Meta │ (Custom Audiences)
│ Platform │ │ LinkedIn │ (Matched Audiences)
│ │──events───→ │ TikTok │ (Custom Audiences)
│ Segment, │ stream │ Reddit │ (Custom Audiences)
│ mParticle│ └──────────────┘
│ Rudder- │
│ Stack │──suppress──→ Exclusion lists
└─────────┘ (purchasers, churned)
Integration patterns:
1. Audience Sync (batch, daily):
- CDP segments → platform audience lists
- Refresh daily for fresh suppression
- Use hashed PII (email, phone) for matching
2. Event Streaming (real-time):
- Purchase events → Google Enhanced Conversions
- Lead events → Meta CAPI
- Sign-up events → LinkedIn CAPI
- Enrich events with CRM data (LTV, segment)
3. Identity Resolution:
- CDP resolves cross-device identity
- Single user profile across touchpoints
- Deterministic (email/login) + probabilistic (device graph)
Popular CDPs for advertising:
| CDP | Best For | Ad Integrations | Price |
|-----|----------|-----------------|-------|
| Segment | Developer-first | All major platforms | $120+/mo |
| mParticle | Enterprise | All major platforms | Custom |
| RudderStack | Open-source, cost-effective | All major platforms | Free-$500+/mo |
| Hightouch | Reverse ETL from warehouse | All major platforms | $350+/mo |
| Census | Reverse ETL from warehouse | All major platforms | $400+/mo |
| Browser | Third-Party Cookies | First-Party Cookies (JS) | First-Party Cookies (Server) |
|---|---|---|---|
| Chrome | Blocked (Privacy Sandbox) | 7 days (ITP-like) | 1-2 years |
| Safari | Blocked since 2020 | 7 days (ITP) | 1-2 years |
| Firefox | Blocked (ETP) | 7 days | 1-2 years |
| Edge | Following Chrome | 7 days | 1-2 years |
| Brave | Blocked | Varies | 1-2 years |
| Identifier | Google Match Rate | Meta Match Rate | LinkedIn Match Rate |
|---|---|---|---|
| Email (hashed) | 40-60% | 50-70% | 30-50% |
| Phone (hashed) | 30-50% | 40-60% | 20-35% |
| Email + Phone | 55-75% | 65-80% | 40-60% |
| Full PII (name, address, email) | 60-80% | 70-85% | 45-65% |
| Mobile Ad ID (IDFA/GAID) | 20-40% | 30-50% | N/A |
| Regulation | Geography | Consent Required | Penalties |
|---|---|---|---|
| GDPR | EU/EEA | Opt-in (explicit) | Up to 4% of global revenue |
| UK GDPR | United Kingdom | Opt-in (explicit) | Up to £17.5M or 4% revenue |
| CCPA/CPRA | California | Opt-out (right to refuse sale) | $2,500-7,500 per violation |
| LGPD | Brazil | Opt-in (explicit) | Up to 2% of Brazil revenue |
| POPIA | South Africa | Opt-in (explicit) | Up to R10M or imprisonment |
| PDPA | Thailand | Opt-in (consent) | Up to THB 5M |
| PIPL | China | Opt-in (consent) | Up to 5% of annual revenue |
| Platform | Name | Best For | Access |
|---|---|---|---|
| Ads Data Hub | Google Ads data analysis | BigQuery integration | |
| Meta | Advanced Analytics | Meta campaign analysis | Via Meta Business Suite |
| Amazon | Amazon Marketing Cloud | Amazon Ads analysis | AWS Clean Rooms |
| LiveRamp | Safe Haven | Cross-platform matching | Enterprise agreement |
| Snowflake | Data Clean Rooms | Custom analysis | Snowflake account |
| InfoSum | InfoSum Platform | Privacy-first matching | Enterprise agreement |
User: "We're losing conversion data due to cookie restrictions. Fix our tracking."
Implementation plan:
1. Google Enhanced Conversions:
- Add data layer to checkout confirmation page:
dataLayer.push({
'event': 'purchase',
'enhanced_conversion_data': {
'email': customer.email // unhashed, GTM hashes it
},
'transaction_id': order.id,
'value': order.total,
'currency': 'USD'
});
- Enable Enhanced Conversions in GTM conversion tag
- Expected recovery: +12% attributed conversions
2. Meta CAPI (server-side):
- Send purchase event from server on order completion
- Include: email (hashed), phone (hashed), fbc cookie, fbp cookie
- Deduplicate with browser pixel via event_id
- Expected recovery: +25% attributed conversions
3. Server-side GTM:
- Deploy on track.yourdomain.com (Cloud Run)
- Route GA4 and Google Ads tags through server
- First-party cookies extend from 7 days → 2 years
- Expected recovery: +15% returning user attribution
Total expected improvement: +30-40% conversion attribution
User: "We need to comply with GDPR consent requirements for our Google Ads."
Setup:
1. Install Cookiebot CMP (Google-certified)
- Auto-scans cookies on your domain
- Generates consent banner with accept/reject/customize
2. Set consent defaults (pre-banner):
- All EEA users: denied by default
- US users: granted by default (CCPA = opt-out model)
3. Configure Advanced Mode:
- Cookieless pings sent when denied
- Google models conversions from consented cohort
- Recovers ~60% of unconsented conversions
4. GTM implementation:
- All Google tags fire based on consent state
- Non-Google tags: require ad_storage = granted
5. Verification:
- Google Ads → Tools → Diagnostics → consent signal check
- GA4 → Admin → Data Collection → consent mode enabled
- Test with EU VPN: verify banner appears, tags behave correctly
Expected impact:
- Consent rate: 60-75% (well-designed banner)
- Modeled conversions recover: 50-70% of denied users
- Net conversion loss: <15% vs no consent implementation
User: "We're a B2B SaaS company. How do we build a first-party data strategy?"
Strategy:
1. Identify first-party data sources:
- Free trial signups: email, company, job title, company size
- Demo requests: email, phone, company, use case
- Content downloads: email, company (gated content)
- Product usage: in-app events, feature adoption, login frequency
- Support tickets: issues, satisfaction scores
- Sales CRM: deal stage, revenue potential, close date
2. Build unified customer profile:
- CDP: RudderStack (cost-effective, open-source)
- Identity: email as primary key, user_id as secondary
- Segments: trial_active, trial_expired, customer, churned, enterprise_lead
3. Activate for advertising:
- Google Customer Match: upload email segments weekly
- LinkedIn Matched Audiences: company list for ABM
- Meta Custom Audiences: email list for lookalike expansion
- Suppression: always exclude active customers from acquisition campaigns
4. Event streaming for optimization:
- Trial signup → Google Enhanced Conversion
- Demo booked → Meta CAPI (Lead event)
- Paid conversion → all platforms (Purchase event, value = ARR)
- Pass LTV data for value-based bidding
5. Privacy compliance:
- Consent mode v2 for EU visitors
- Consent at signup: "We may use your data for advertising" checkbox
- Data retention: 24 months, auto-delete after
- Data subject requests: automated via CDP
Part of the Synter free skills collection — open-source agent skills for advertising, PPC, and marketing automation.
Want this skill (and 40+ more) running on autopilot against your live ad accounts? Try Synter — AI Agent Media Buyers that connect to Google, Meta, LinkedIn, TikTok, Reddit, Amazon, and 7+ more platforms.