---
name: marketing-attribution-engine
description: 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

1. Map all marketing touchpoints: paid search, organic, social, email, display, referrals, direct.
2. Implement tracking infrastructure: UTM parameters, pixel tracking, CRM integration, call tracking.
3. Select attribution model: last-click, first-click, linear, time-decay, position-based, data-driven.
4. Configure attribution windows: click-through (1–90 days), view-through (1–30 days).
5. Collect and normalize journey data from all platforms and touchpoints.
6. Calculate credit allocation per touchpoint based on selected model.
7. Generate attribution reports: channel contribution, touchpoint value, customer journey paths.
8. Identify budget optimization opportunities: shift spend from low-attribution to high-attribution channels.
9. Build customer journey visualization: common paths, drop-off points, conversion timelines.
10. 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.