Attribution & Tracking Implementation

Attribution & Tracking Implementation

The Non-Negotiable Requirement

Every sale must be attributable to its original traffic source.

This is not optional. This is not “nice to have.” This is fundamental to running a marketing operation.

---

The Three-Stage Attribution Model

Overview: From Anonymous Visitor to Paying Customer

Everyone starts at Stage 1. Each subsequent stage is a subset of the previous.

Stage 1: Landing (UTM Parameters)
         ├─ Have: UTM params, session data, page views
         └─ Don't have: Contact information

         ↓ (Subset converts)

Stage 2: Lead (UTM + Contact)
         ├─ Have: UTM params + Email/Phone
         └─ Don't have: Purchase/transaction

         ↓ (Subset converts)

Stage 3: Sale (UTM + Contact + Transaction)
         ├─ Have: UTM params + Email/Phone + Card + Billing + Invoice
         └─ DEFINITIVE - Multiple system records

The Drop-off Groups:

• Dropped at Stage 1: UTM parameters, NO contact → Anonymous bounces
• Dropped at Stage 2: UTM parameters + Contact, NO sale → Leads who didn’t convert
• Reached Stage 3: UTM parameters + Contact + Sale → Customers

The Key Insight: The difference between groups is simply whether they have contact and/or sale data. The UTM parameters persist through all stages.

---

The Attribution Chain

Step 1: Landing - Capture UTM Parameters

User lands on your site with URL:

https://landing.fh.co/medication-tirzepatide-discount-01?
  utm_source=fb&
  utm_medium=cpc&
  utm_campaign=tirzepatide-q4&
  utm_term=weight-loss-medication&
  utm_content=video-ad-v2&
  campaign_id=120236311971890504&
  ad_id=120233281654680682&
  adset_id=120233281060550682&
  affid=1000

Capture and persist:

• All UTM parameters (utm_source, utm_medium, utm_campaign, utm_term, utm_content)
• Campaign identifiers (campaign_id, ad_id, adset_id)
• Affiliate/channel identifiers (affid, fbp, etc.)
• Landing page URL
• Referrer
• Timestamp

Storage:

• Browser cookies (first-party)
• Local storage
• Session storage
• Server-side session
• Analytics tool (PostHog, GA, etc.)

Step 2: Funnel Journey - Maintain Context

As user navigates through funnel:

• Page 1: Landing page
• Page 2: Quiz questions
• Page 3: Qualify/achieve screen
• Page 4: Contact form
• Page 5: Payment/conversion

All UTM parameters persist across:

• Page navigations
• Form submissions
• Redirects
• Subdomain changes
• Days/weeks (if user returns)

Technical implementation:

• Pass parameters in hidden form fields
• Maintain in session/cookies
• Backend session persistence
• User profile attributes

Step 3: Lead Capture - Connect Anonymous to Identified

User provides PII (email/phone):

{
  "email": "john@example.com",
  "phone": "+1-555-0123",
  "first_name": "John",
  "last_name": "Doe",

  // Attribution data (from Step 1)
  "utm_source": "fb",
  "utm_medium": "cpc",
  "utm_campaign": "tirzepatide-q4",
  "utm_term": "weight-loss-medication",
  "utm_content": "video-ad-v2",
  "campaign_id": "120236311971890504",
  "ad_id": "120233281654680682",
  "adset_id": "120233281060550682",
  "affid": "1000",
  "landing_page": "/medication-tirzepatide-discount-01",
  "first_seen": "2025-11-25T14:34:25Z"
}

Now you have a Lead with full attribution.

Step 4: Conversion - Close the Loop

When lead converts to customer/sale, you have MULTIPLE definitive records:

Sale Record (Billing System):

{
  "transaction_id": "txn_xyz789",
  "customer_id": "cust_abc123",
  "email": "john@example.com",
  "purchase_date": "2025-11-28T10:23:45Z",
  "amount": "$499.00",
  "product": "Tirzepatide Monthly Subscription",
  "status": "completed"
}

Billing Details:

{
  "card_last4": "4242",
  "card_brand": "Visa",
  "billing_address": {
    "street": "123 Main St",
    "city": "San Francisco",
    "state": "CA",
    "zip": "94102"
  },
  "billing_email": "john@example.com"
}

Invoice Record:

{
  "invoice_number": "INV-2025-001234",
  "invoice_date": "2025-11-28",
  "amount": "$499.00",
  "customer_email": "john@example.com",
  "status": "paid"
}

Attribution Record (Connected via Email):

{
  "email": "john@example.com",  // THE CONNECTOR

  // Original attribution (from Step 1, via Step 3)
  "utm_source": "fb",
  "utm_campaign": "tirzepatide-q4",
  "campaign_id": "120236311971890504",
  "ad_id": "120233281654680682",
  "landing_page": "/medication-tirzepatide-discount-01",
  "first_seen": "2025-11-25T14:34:25Z",

  // Journey metrics
  "days_to_convert": 3,
  "touchpoints": 7,
  "sessions": 4
}

The Connection is Trivial:

1. Sale record has email: john@example.com
2. Lead record has email: john@example.com + UTM parameters
3. Match on email → Sale is attributed to original UTM parameters

Sales are SUPER DEFINITIVE because they have:

• Transaction ID (unique identifier)
• Card information (person paid real money)
• Billing address (physical location)
• Invoice number (accounting record)
• Email/phone (contact identifier)
• Multiple system records (billing system, CRM, invoicing, etc.)

This is not speculation. This is hard data across multiple systems.

Now you can answer:

• Which campaigns drive sales? (campaign_id)
• What’s the ROI of each ad? (ad_id → revenue)
• Which landing pages convert to sales? (landing_page)
• What’s the cost per acquisition? (ad spend ÷ conversions by campaign)
• Which UTM terms have highest LTV? (utm_term → customer value)
• What’s the complete customer journey? (UTM → Lead → Sale with timestamps)

---

Why This is Trivial to Implement

Technical Reality

Capturing UTM parameters: 10 lines of JavaScript

// Parse URL parameters
const urlParams = new URLSearchParams(window.location.search);
const utmData = {
  utm_source: urlParams.get('utm_source'),
  utm_medium: urlParams.get('utm_medium'),
  utm_campaign: urlParams.get('utm_campaign'),
  utm_term: urlParams.get('utm_term'),
  utm_content: urlParams.get('utm_content'),
  campaign_id: urlParams.get('campaign_id'),
  ad_id: urlParams.get('ad_id')
};

// Store in localStorage/cookie
localStorage.setItem('attribution', JSON.stringify(utmData));

Persisting through funnel: Already done if using any modern analytics tool

Connecting to lead/customer: One database field per UTM parameter

Total implementation time: 1 day for a competent developer

• Morning: Implement UTM capture
• Afternoon: Add fields to lead/customer database
• Done.

Tool Support

Every major tool has this built-in:

• PostHog: Automatically captures UTM parameters, persists through session, connects to identified users
• Google Analytics: GA4 does this natively with User-ID
• Mixpanel: User profiles with UTM properties
• Segment: Track page with context, identify user connects everything
• HubSpot: Contact properties for UTM parameters
• Salesforce: Campaign attribution fields

If your tool doesn’t do this, get a better tool.

---

Common Excuses (And Why They’re Invalid)

❌ “Our funnel spans multiple domains/technologies”

Reality: Doesn’t matter. Pass UTM parameters via:

• URL query strings between domains
• Server-side session with shared database
• Third-party cookies (if same-party)
• Server-to-server API calls

❌ “We use a third-party funnel builder”

Reality: Every funnel builder (ClickFunnels, Unbounce, Leadpages, etc.) supports:

• Hidden form fields for UTM parameters
• Webhook pass-through of data
• Integration with CRM/analytics

❌ “Users come back days later, cookies expire”

Reality:

• First-party cookies can last 1-2 years
• Once user provides email, store attribution in database permanently
• Use server-side storage, not just cookies

❌ “It’s technically complex”

Reality: No, it’s not. This is basic web development. If your team can’t do this, they’re incompetent.

❌ “Our analytics tool doesn’t support it”

Reality: Get a better tool. This is 2025, not 2005.

---

What You Get With Proper Attribution

Campaign Performance Analysis

Campaign: 120236311971890504
- Ad Spend: $10,000
- Clicks: 5,000
- Leads: 423
- Sales: 187
- Revenue: $93,313 (187 × $499)
- ROAS: 9.3x
- CPA: $53.48
- Decision: SCALE THIS

vs.

Campaign: 120236311971234567
- Ad Spend: $10,000
- Clicks: 8,000
- Leads: 150
- Sales: 15
- Revenue: $7,485 (15 × $499)
- ROAS: 0.75x
- CPA: $666.67
- Decision: PAUSE THIS

Landing Page ROI

Landing Page: /medication-tirzepatide-discount-01
- Traffic: 33,804
- Sales: 1,960 (5.8% conversion)
- Revenue: $978,040

Landing Page: /medication-tirzepatide-pre-quiz-tt
- Traffic: 1,132
- Sales: 249 (22% conversion)
- Revenue: $124,251
- Insight: Lower traffic but 4x conversion - drive more traffic here

Attribution Reporting

Every sale shows:

• Which ad they clicked
• Which landing page they saw
• Which quiz answers they gave
• How many days to convert
• How many sessions/touchpoints
• Complete customer journey

This is what “sales are not mysterious” means.

---

Implementation Checklist

Phase 1: Capture (Day 1)

• Capture all UTM parameters on landing
• Capture campaign_id, ad_id, adset_id
• Capture landing page URL and referrer
• Store in cookies/localStorage
• Send to analytics tool (PostHog, GA, etc.)

Phase 2: Persist (Day 1)

• Maintain UTM data across page navigations
• Pass UTM data through form submissions
• Store in hidden form fields if needed
• Configure analytics tool to persist user properties

Phase 3: Connect (Day 1)

• When user provides email/phone, connect to UTM data
• Save attribution fields in CRM/database
• Create lead record with full attribution
• Set up PostHog user identification

Phase 4: Report (Day 1)

• Create attribution report: leads by campaign
• Create attribution report: sales by campaign
• Calculate ROAS/CPA by campaign
• Build dashboard showing end-to-end funnel

Total Time: 1 Day

Even if your system is:

• Multiple components
• Different technologies for landing pages, forms, billing
• Complex multi-step funnel
• Third-party payment processor
• Separate CRM and billing systems

It’s still 1 day.

The connection is simple:

1. Capture UTM on landing → Store with session
2. Capture email on lead form → Connect UTM to email
3. Sale happens with email → Match sale to lead → Get UTM attribution

Snap. Finger snap. Done.

No excuses.

---

---

Understanding the Funnel Stages

Stage 1: Landing with UTM Parameters

Who: Everyone who clicks your ad/link Data captured:

• UTM parameters (source, medium, campaign, term, content)
• Campaign/Ad IDs
• Landing page URL
• Session data
• Device/browser info
• Timestamp

No PII yet - These are anonymous visitors

Count: 55,775 people in your example

---

Stage 2: Lead (Subset of Stage 1)

Who: Visitors who provided contact information Data captured:

• Everything from Stage 1 (UTM parameters)
• PLUS: Email and/or phone number
• Name
• Any form fields they filled

Now you have PII - Anonymous visitor becomes identified lead

Count: 5,509 people in your example (9.88% of Stage 1)

The difference: Stage 2 has contact, Stage 1 doesn’t. That’s it.

---

Stage 3: Sale (Subset of Stage 2)

Who: Leads who purchased Data captured:

• Everything from Stage 1 (UTM parameters)
• Everything from Stage 2 (Email/phone)
• PLUS: Transaction data:
  • Transaction ID
  • Card information
  • Billing address
  • Purchase amount
  • Invoice number
  • Payment status
  • Product purchased

Super definitive - Multiple system records (billing, invoicing, CRM, payment processor)

Count: 187 people in your example (3.37% of Stage 2, 0.34% of Stage 1)

The difference: Stage 3 has transaction records, Stage 2 doesn’t.

---

The Drop-Off Analysis

Group A: Dropped at Landing (Stage 1 only)

• Count: 50,266 people (55,775 - 5,509)
• Have: UTM parameters, session data
• Don’t have: Contact information
• Analysis: Never engaged enough to provide email
• Attribution: Can’t attribute individual sales, but can see which campaigns drive more engagement

Group B: Lead but No Sale (Stage 2 only)

• Count: 5,322 people (5,509 - 187)
• Have: UTM parameters + email/phone
• Don’t have: Transaction/purchase
• Analysis: Provided contact but didn’t convert to customer
• Attribution: Know which campaigns generate leads (even if not sales yet)
• Action: Can nurture via email/SMS since you have contact info

Group C: Sale (Stage 3)

• Count: 187 people
• Have: UTM parameters + email/phone + transaction data
• Don’t have: Nothing - this is complete data
• Analysis: Full attribution from ad click → landing → lead → sale
• Attribution: Definitive ROI per campaign/ad/landing page

---

Why Sales Are Not Mysterious

Sales have multiple definitive records across multiple systems:

1. Billing System: Transaction ID, amount, card info, billing address
2. Payment Processor: Payment confirmation, receipt, transaction log
3. CRM/Database: Customer record with email/phone
4. Invoicing System: Invoice number, invoice PDF, accounting record
5. Analytics: Conversion event with UTM parameters

All connected by email/phone.

When you see:

• Transaction txn_xyz789 for john@example.com
• Lead record for john@example.com with campaign_id: 120236311971890504

You instantly know: This $499 sale came from Facebook campaign 120236311971890504.

It’s not rocket science. It’s basic data matching.

---

Key Principle

Sales are not mysterious. Every sale traces back to a specific user who landed with specific UTM parameters.

• Stage 1: UTM parameters (everyone who lands)
• Stage 2: UTM + Contact (subset who become leads)
• Stage 3: UTM + Contact + Transaction (subset who become customers)

The connection is trivial: Match on email/phone.

You must capture this. It’s not optional. It’s not hard. Just do it.

---

Multi-Touch Attribution: The Full Journey

The Reality: Users Don’t Convert in a Straight Line

Initial Traffic Source → Retargeting Channels → Final Conversion

Example Customer Journey

Day 1: Initial Landing

• User clicks Facebook ad (campaign 120236311971890504, affid 1000)
• Lands on /medication-tirzepatide-discount-01
• Provides email: sarah@example.com
• Doesn’t purchase (becomes Lead, not Sale)
• System stores: email + UTM parameters + campaign ID + affid

Day 3: SMS Retargeting

• Marketing sends SMS to sarah@example.com (actually phone number)
• User clicks SMS link → visits site
• New session, but SAME person (same email/phone)

Day 5: Email Retargeting

• Marketing sends email to sarah@example.com
• User clicks email link → visits site again
• Still doesn’t purchase

Day 7: Direct Visit → Purchase

• User remembers site, visits directly (no UTM parameters in URL)
• Makes purchase: $499
• Sale record: Transaction ID, card info, email: sarah@example.com

The Attribution Question

Which channel gets credit for the sale?

❌ Wrong Answer: “Direct visit” or “SMS” or “Email”

• This ignores how the user originally found you

✅ Correct Answer: Facebook campaign 120236311971890504

• This was the FIRST TOUCH that brought them in
• Without that initial ad, they never would have existed in your system

How Proper Attribution Works

When you query sales attribution:

SELECT
  sales.email,
  sales.amount,
  sales.purchase_date,
  leads.utm_source,
  leads.utm_campaign,
  leads.campaign_id,
  leads.affid,
  leads.first_seen_date
FROM sales
JOIN leads ON sales.email = leads.email

Result:

email: sarah@example.com
amount: $499
purchase_date: 2025-11-28
utm_source: fb
utm_campaign: tirzepatide-q4
campaign_id: 120236311971890504
affid: 1000
first_seen: 2025-11-21
days_to_convert: 7

The sale is attributed to the original Facebook campaign, even though the final session came from direct visit after SMS and email touches.

Multi-Touch Visibility

With proper system setup, you can see:

1. First Touch: Facebook ad (Day 1)
2. Middle Touches: SMS (Day 3), Email (Day 5)
3. Last Touch: Direct visit (Day 7)
4. Full Journey: FB → SMS → Email → Direct → Purchase

Attribution models you can use:

• First-Touch: Credit Facebook (most common for acquisition)
• Last-Touch: Credit Direct visit (less useful)
• Multi-Touch: Split credit across FB, SMS, Email
• Data-Driven: Algorithm determines contribution of each touch

For marketing analysis, first-touch is usually most important because:

• It shows which campaigns ACQUIRE customers
• SMS/Email retargeting works on already-acquired leads
• You need to know which paid ads are worth the investment

Why This Works

The system had the initial data stored from when user first landed (Day 1):

• Email captured with form submission
• UTM parameters persisted to database
• Lead record created with full attribution

When sale happens 7 days later:

• Sale record has email: sarah@example.com
• Lead record has email: sarah@example.com + original UTM data
• Match on email → Sale attributed to original source

It doesn’t matter that they came back via SMS/email/direct - the original attribution is preserved.

Real-World Example from Your Data

Scenario: You check sales attribution in your system (Moon, PostHog, etc.)

What you see:

• Sales attributed to specific campaigns
• Sales attributed to specific affids (traffic sources)
• Sales attributed to original landing pages

Some sales show:

• Original source: Facebook campaign X, affid 1000
• Final touchpoint: SMS or Email retargeting
• Attribution: Facebook campaign X (because that’s first touch)

Why you can see this instantly:

• Not “magical tools”
• The data was stored from Day 1 when they first landed
• You just need to query: sales → leads (via email) → UTM parameters
• The connections are already in the system

What You Can Analyze

Campaign Performance (Full Funnel):

Campaign 120236311971890504:
├─ Landings: 33,804
├─ Leads: 3,429 (10.1%)
└─ Sales: 187 (5.5% of leads, 0.55% of landings)
   ├─ Direct conversion: 45 sales (24%)
   ├─ Via SMS retargeting: 82 sales (44%)
   ├─ Via Email retargeting: 48 sales (26%)
   └─ Via other touchpoints: 12 sales (6%)

Key insights:

• Campaign drove 3,429 leads
• 187 became customers (even if took 7+ days and retargeting)
• Most sales (70%) required retargeting via SMS/email
• But ALL sales attributed back to original campaign

Traffic Source Performance:

Affid 1000 (Facebook):
- Leads: 5,200
- Sales: 420
- Revenue: $209,580
- ROAS: 10.5x
- Most sales came via retargeting, but all originated from FB

The Key Principle

Original attribution is sacred. Retargeting is amplification, not replacement.

• User finds you via paid ad → First-touch attribution
• You retarget via SMS/email → Conversion optimization
• User converts via any channel → Credit goes to first touch

Without the original ad, they wouldn’t exist in your retargeting list.

Implementation Requirements

To enable multi-touch attribution:

1. Capture first-touch data (UTM parameters at landing)
2. Store with PII (email/phone when captured)
3. Persist forever (never overwrite original attribution)
4. Track all touchpoints (SMS clicks, email clicks, direct visits)
5. Connect via identifier (email/phone matches across systems)

With PostHog, Moon, or similar tools: This is automatic

• First-touch stored on user profile
• All sessions/events tagged with user ID
• Sale event connected to user → Gets original attribution

Not magical. Just proper data architecture.