Marketing AI Skill

Marketing Attribution Engine

Track and attribute conversions across all marketing touchpoints using multi-touch attribution models. Use when setting up marketing attribution, building attribution models, tracking cross-channel conversions, analyzing customer journey touchpoints, allocating marketing budget based on attribution data, or measuring multi-channel campaign effectiveness. Triggers on phrases like "marketing attribution", "attribution model", "multi-touch attribution", "touchpoint tracking", "cross-channel attribution", "journey attribution", "first-touch last-touch", "data-driven attribution", "attribution reporting".

Marketing Attribution Engine

Accurately credit each marketing touchpoint along the customer journey with sophisticated multi-touch attribution models.

Workflow

Map all marketing touchpoints: paid search, organic, social, email, display, referrals, direct.
Implement tracking infrastructure: UTM parameters, pixel tracking, CRM integration, call tracking.
Select attribution model: last-click, first-click, linear, time-decay, position-based, data-driven.
Configure attribution windows: click-through (1–90 days), view-through (1–30 days).
Collect and normalize journey data from all platforms and touchpoints.
Calculate credit allocation per touchpoint based on selected model.
Generate attribution reports: channel contribution, touchpoint value, customer journey paths.
Identify budget optimization opportunities: shift spend from low-attribution to high-attribution channels.
Build customer journey visualization: common paths, drop-off points, conversion timelines.
Continuously refine model: compare model outputs, validate against revenue data, iterate.

Attribution Models

ATTRIBUTION MODEL COMPARISON
===============================

LAST-CLICK (Single-Touch):
  → How it works: 100% credit to final touchpoint before conversion
  → Example journey: Organic Search (Day 1) → Facebook Ad (Day 3)
     → Email (Day 5) → Google Ads (Day 7, conversion)
     → Credit: Google Ads = 100%, all others = 0%
  → Pros: Simple to implement, clear accountability, default in most platforms
  → Cons: Ignores all upstream touchpoints, over-values bottom-funnel channels
  → Best for: Short sales cycles, transactional purchases, simple funnels
  → Used by: 70% of businesses (default in Google Analytics)

FIRST-CLICK (Single-Touch):
  → How it works: 100% credit to first touchpoint in journey
  → Example journey: Facebook Ad (Day 1) → Organic Search (Day 3)
     → Email (Day 5) → Direct (Day 7, conversion)
     → Credit: Facebook Ad = 100%, all others = 0%
  → Pros: Values awareness channels, identifies top-of-funnel drivers
  → Cons: Ignores nurturing and conversion touchpoints
  → Best for: Brand-building campaigns, awareness-focused organizations

LINEAR (Multi-Touch):
  → How it works: Equal credit to every touchpoint in the journey
  → Example journey (4 touchpoints): Each gets 25% credit
  → Pros: Acknowledges all channels, simple to understand
  → Cons: Over-credits irrelevant touchpoints, doesn't weight importance
  → Best for: Organizations valuing full-funnel visibility equally

TIME-DECAY (Multi-Touch):
  → How it works: More credit to touchpoints closer to conversion
  → Formula: Credit = 1 / (1 + e^(k × (days_from_conversion - midpoint)))
  → Example journey:
     Day 1 (Social): 6% credit
     Day 5 (Email): 15% credit
     Day 10 (Content): 20% credit
     Day 15 (PPC): 30% credit
     Day 20 (Direct, conversion): 29% credit
  → Pros: Rewards touchpoints that drive action, reflects urgency
  → Cons: Under-credits early awareness touchpoints
  → Best for: Medium-length sales cycles, action-oriented marketing

POSITION-BASED / U-SHAPED (Multi-Touch):
  → How it works: 40% first touchpoint, 40% last touchpoint, 20% middle
  → Example journey (3 touchpoints):
     First (Social): 40% credit
     Middle (Email): 20% credit
     Last (PPC, conversion): 40% credit
  → Pros: Values both awareness and conversion, balanced view
  → Cons: Middle touchpoints all share small credit equally
  → Best for: B2B with defined awareness → consideration → decision stages
  → Industry standard: Most popular multi-touch model (used by 40% of businesses)

DATA-DRIVEN (Algorithmic):
  → How it works: Machine learning assigns credit based on actual contribution
  → Analyzes: All historical journeys, identifies patterns, weights touchpoints
  → Factors: Channel, time to conversion, frequency, sequence, audience segment
  → Requires: Minimum 100+ conversions over 30 days for meaningful output
  → Pros: Most accurate, adapts to actual business patterns, no arbitrary weights
  → Cons: Requires significant data, black-box algorithm (less transparent)
  → Best for: High-volume conversion businesses with mature tracking
  → Used by: Google Ads, Google Analytics 4 (default for GA360)

MODEL COMPARISON BY SALES CYCLE:

  SHORT CYCLE (< 7 days): Last-Click or Time-Decay
    → Touchpoints: 2–4 average
    → Recommendation: Time-Decay (captures urgency without complexity)

  MEDIUM CYCLE (7–30 days): Position-Based or Data-Driven
    → Touchpoints: 5–10 average
    → Recommendation: Position-Based (balanced awareness + conversion value)

  LONG CYCLE (30–90+ days): Data-Driven or Custom
    → Touchpoints: 10–20+ average
    → Recommendation: Data-Driven (ML needed to weight complex journeys)

  MIXED CYCLES: Multiple models in parallel
    → Run 2–3 models simultaneously
    → Compare outputs to understand model sensitivity
    → Use averaged attribution for budget allocation decisions

Tracking Infrastructure

TRACKING IMPLEMENTATION FRAMEWORK
====================================

UTM PARAMETER STANDARDS:

  REQUIRED UTM PARAMETERS (every campaign link):
    → utm_source: Platform/channel (google, facebook, linkedin, email, twitter)
    → utm_medium: Marketing medium (cpc, cpm, email, social, referral, organic)
    → utm_campaign: Campaign name (spring_sale_2025, product_launch_q1)
    → utm_content: Specific element (banner_top, link_bottom, cta_button)
    → utm_term: Paid keyword (for paid search — auto-populated by some platforms)

  NAMING CONVENTIONS:
    → All lowercase, underscores (no spaces)
    → Consistent across all team members and platforms
    → Date format: YYYY-MM-DD (2025-03-15, not 3/15/25)
    → Campaign structure: [objective]_[channel]_[audience]_[date]
      Example: leadgen_search_b2b_enterprise_2025-03-01

  UTM MANAGEMENT:
    → Use Google's Campaign URL Builder (free tool)
    → Maintain UTM glossary document shared across team
    → Validate UTMs before publishing (check with URL inspector)
    → Audit monthly: find broken, missing, or inconsistent UTMs

PIXEL AND TAG IMPLEMENTATION:

  META PIXEL (Facebook/Instagram):
    → Base pixel code on every page
    → Standard events: ViewContent, AddToCart, InitiateCheckout, Purchase, Lead
    → Custom events: Specific actions not covered by standard events
    → Conversions API: Server-side tracking (complements browser pixel)
    → Testing: Meta Pixel Helper browser extension

  GOOGLE ADS TAG:
    → Global site tag (gtag.js) on every page
    → Event snippets on conversion pages
    → Enhanced Conversions: Hashed first-party data for better matching
    → Testing: Google Tag Assistant browser extension

  GOOGLE ANALYTICS 4:
    → GA4 measurement protocol (gtag.js or GTM)
    → Enhanced measurement: Auto-captures scrolls, clicks, video engagement
    → Custom events: Specific business actions
    → Link to Google Ads for cross-platform attribution

  GOOGLE TAG MANAGER:
    → Container deployed on site (manages all tags from one place)
    → Triggers: Page views, clicks, form submissions, scrolls
    → Variables: URL, click text, form data, custom dimensions
    → Debug: Preview mode for testing before publishing

CALL TRACKING:

  → Provider: CallRail, Invoca, PhoneBurner ($49–$299/month)
    → Dynamic number replacement (swaps number based on traffic source)
    → Call recording and transcription
    → Call-to-lead conversion tracking
    → CRM integration (push calls as leads)
    → Attribution: Maps calls back to marketing source

  → Implementation:
     * Landing pages: Swap phone number based on UTM/referrer
     * Google Ads: Use call extensions with tracked numbers
     * Offline: Use unique codes ("Mention code SAVE20 for discount")

Attribution Analysis and Reporting

ATTRIBUTION REPORTING DASHBOARD
=================================

CHANNEL CONTRIBUTION REPORT:

  ATTRIBUTION BY CHANNEL (Position-Based Model, Last 30 Days):

  ┌────────────────────┬──────────┬──────────┬────────────┬──────────┐
  │ Channel            │ Touches  │ Revenue  │ Credit %   │ CPA      │
  ├────────────────────┼──────────┼──────────┼────────────┼──────────┤
  │ Google Search      │ 1,200    │ $125,000 │ 28%        │ $45      │
  │ Facebook/Instagram │ 850      │ $85,000  │ 18%        │ $65      │
  │ Email              │ 620      │ $62,000  │ 14%        │ $25      │
  │ Organic Content    │ 450      │ $45,000  │ 12%        │ $30      │
  │ LinkedIn           │ 200      │ $35,000  │ 9%         │ $120     │
  │ Direct             │ 380      │ $30,000  │ 8%         │ $20      │
  │ Display/Retargeting│ 150      │ $20,000  │ 6%         │ $85      │
  │ Referral           │ 100      │ $15,000  │ 4%         │ $50      │
  │ Twitter/X          │ 50       │ $8,000   │ 1%         │ $95      │
  └────────────────────┴──────────┴──────────┴────────────┴──────────┘

  INSIGHTS:
    → Google Search drives highest revenue but Facebook has best awareness role
    → Email has lowest CPA — optimize deliverability and segmentation
    → LinkedIn has highest CPA — evaluate if lead quality justifies cost
    → Retargeting contributes 6% credit but essential for closing

CUSTOMER JOURNEY PATH ANALYSIS:

  TOP CONVERSION PATHS (Last 30 Days):

    PATH 1: Organic Search → Email → Google Ads → Conversion (18% of conversions)
      → Average time: 12 days
      → Average revenue: $250
      → Insight: Content marketing + paid search combo is strongest path

    PATH 2: Facebook → Google Search → Email → Conversion (15% of conversions)
      → Average time: 18 days
      → Average revenue: $180
      → Insight: Social awareness → search intent → email nurturing

    PATH 3: Direct → Conversion (12% of conversions)
      → Average time: 0 days (same-day)
      → Average revenue: $320
      → Insight: Brand recognition driving high-intent direct traffic

    PATH 4: LinkedIn → Email → Google Ads → Conversion (8% of conversions)
      → Average time: 25 days
      → Average revenue: $500
      → Insight: B2B long-cycle path with highest revenue per conversion

    PATH 5: Display Retargeting → Google Ads → Conversion (7% of conversions)
      → Average time: 8 days
      → Average revenue: $150
      → Insight: Retargeting effective when paired with paid search

TOUCHPOINT ANALYSIS:

  TOUCHPOINTS PER CONVERSION:
    → Average: 6.5 touchpoints before conversion
    → Median: 5 touchpoints
    → Minimum: 1 (direct conversion)
    → Maximum: 28 (enterprise B2B deal)
    → Trend: Decreasing (was 8.2 touchpoints 6 months ago)

  TOUCHPOINT DISTRIBUTION:
    → 1 touchpoint: 12% of conversions (branded/direct)
    → 2–3 touchpoints: 28% of conversions (consideration)
    → 4–6 touchpoints: 35% of conversions (standard journey)
    → 7+ touchpoints: 25% of conversions (research-heavy)

  TIME TO CONVERSION:
    → Average: 14.5 days
    → Median: 9 days
    → < 3 days: 30% of conversions
    → 3–14 days: 40% of conversions
    → 15–30 days: 20% of conversions
    → 30+ days: 10% of conversions

Budget Optimization Based on Attribution

ATTRIBUTION-DRIVEN BUDGET ALLOCATION
=======================================

BUDGET REALLOCATION FRAMEWORK:

  STEP 1: CALCULATE CHANNEL EFFICIENCY

    Efficiency Score = (Attributed Revenue × Credit %) / Channel Spend

    ┌────────────────────┬────────────┬──────────┬────────────┬──────────┐
    │ Channel            │ Spend      │ Revenue  │ Credit %   │ Efficiency│
    ├────────────────────┼────────────┼──────────┼────────────┼──────────┤
    │ Google Search      │ $50,000    │ $35,000  │ 28%        │ 0.20     │
    │ Facebook           │ $30,000    │ $15,300  │ 18%        │ 0.10     │
    │ Email              │ $5,000     │ $8,680   │ 14%        │ 0.17     │
    │ LinkedIn           │ $15,000    │ $3,150   │ 9%         │ 0.05     │
    │ Retargeting        │ $10,000    │ $1,200   │ 6%         │ 0.06     │
    └────────────────────┴────────────┴──────────┴────────────┴──────────┘

  STEP 2: IDENTIFY OVER- AND UNDER-SPEND

    → OVER-SPEND: LinkedIn (efficiency 0.05), Retargeting (0.06)
       * Action: Reduce budget by 20–30%, reallocate to higher-efficiency channels
    → UNDER-SPEND: Google Search (0.20), Email (0.17)
       * Action: Increase budget by 15–25%, scale proven channels
    → MONITOR: Facebook (0.10) — acceptable but test optimizations

  STEP 3: CALCULATE REALLOCATED BUDGET

    CURRENT BUDGET: $110,000/month

    PROPOSED REALLOCATION:
      → Google Search: $50,000 → $65,000 (+$15,000, +30%)
      → Email: $5,000 → $8,000 (+$3,000, +60%)
      → Facebook: $30,000 → $28,000 (-$2,000, -7%)
      → LinkedIn: $15,000 → $10,000 (-$5,000, -33%)
      → Retargeting: $10,000 → $9,000 (-$1,000, -10%)
      → New: Content Marketing: $0 → $10,000 (new investment)

    EXPECTED OUTCOME:
      → Projected revenue increase: 15–25% (based on efficiency differentials)
      → Projected CPA decrease: 10–15% (shift from high-CPA to low-CPA channels)
      → Timeline: 60–90 days to see full impact

  STEP 4: IMPLEMENT AND MONITOR

    → Implement changes gradually (10–20% shifts, not 50%+)
    → Monitor for 30 days before next adjustment
    → Compare pre/post attribution data
    → Adjust attribution model if needed (data may reveal different patterns)

Integration Points

Google Analytics 4: Data-driven attribution, cross-channel tracking
Google Ads: Conversion tracking, attribution reports, campaign manager
Meta Ads Manager: Attribution window settings, conversion reporting
HubSpot: Marketing attribution reporting, deal stage tracking
Salesforce: Revenue attribution, closed-won deal touchpoint tracking
Segment / mParticle: Customer data platform for unified journey tracking
Braze / Adobe Analytics: Advanced attribution and journey analytics
CallRail / Invoca: Call tracking attribution
Tableau / Looker: Custom attribution dashboards
Impact / Refersion: Affiliate and partner attribution tracking

Edge Cases

Cross-device journeys: Users switch between mobile, tablet, and desktop. Google Analytics 4 uses device graph to stitch cross-device journeys when users are signed in. For anonymous users, use probabilistic matching (IP + fingerprinting). Limitation: 30–50% of cross-device journeys may not be fully tracked.
Offline conversions: Physical store visits, phone calls, in-person sales. Use: call tracking numbers, promo codes by channel, QR codes with channel attribution, post-purchase surveys ("How did you hear about us?"). Upload offline conversions back to ad platforms for closed-loop attribution.
B2B team-based buying: Multiple decision-makers touch content before one converts. Solution: Use group attribution (credit all team members' touchpoints), implement account-level attribution, track marketing-qualified accounts (not just leads), use CRM deal stages to map content consumption to revenue.
Attribution window selection: Shorter windows (1-day click) under-count; longer windows (90-day click) over-attribute. Recommendation: Use platform defaults but understand the impact. Google Ads: 30-day click, 1-day view. Meta: 7-day click, 1-day view. GA4: 30-day default. Document and standardize across platforms.
Model comparison reporting: Never rely on a single attribution model. Run 3 models in parallel (last-click, position-based, data-driven). Report all three. Use variance between models as a confidence indicator (high variance = low confidence in any single model). Present range, not point estimate, to stakeholders.

Disclaimer: All rights reserved by Circulos AI. These skills are specifically designed for Claude Code, Claude Cowork, Codex, and OpenClaw. When using or referencing any skill, please provide proper attribution to Circulos AI.