Skip to content

Encyclopedia - Topics Index

Encyclopedia - Topics Index

Purpose: Quick-reference index of all major concepts. Use when you know the basics and want to look up something specific.

For structured learning: See Curriculum for step-by-step lessons.

How to Use This Index

This is a lookup layer, not a tutorial. Each entry gives you:

Concepts: What it is (simple facts, conclusions)
Source docs: Where to read the full detailed explanation

When to use this:

“I need to remember how UTM parameters work” → Look up Event Schema & Parameters
“What’s the consent gap thing again?” → Look up Cross-Validation
“How do I isolate landing page vs traffic quality?” → Look up Thinking Flows

When NOT to use this:

“I’m new and need to learn from scratch” → Use Curriculum
“I want exercises and hands-on practice” → Use Curriculum

Deep Source Documents

These are the full, detailed, opinionated knowledge bases:

Core Principles - Fundamental truths and rules (monitoring, data analysis, attribution, validation)
Thinking Flows - Step-by-step investigation methodologies
Practical Examples - Real-world case studies and walkthroughs
Attribution & Tracking - End-to-end attribution implementation (non-negotiable)
Event Schema & Parameters - Understanding event footprints (your attribution Bible)
Cross-Validation - Never trust a single system (triangulate data sources)

Quick Lookup by Topic

Attribution

The 3-Stage Funnel - Landing → Lead → Sale
The Attribution Chain - UTM → PII → Sale
First-Touch vs Last-Touch
Implementation (1 Day)

Event Schema & Parameters

What is an Event Schema?
Standard UTM Parameters
Platform-Specific IDs
Reading Event Footprints
Parameter Evolution

Cross-Validation

Three-System Cross-Check
Facebook Consent Gap
Unattributed Sales (5-20%)
Daily Cross-Validation
Micro vs Macro Analysis

Thinking Flows

Aggregate → Individual → Pattern
Multi-Dimensional Segmentation
Isolating Landing Page vs Traffic

Monitoring & Analytics

Self-Sufficient Monitoring
Discover, Don’t Define
Exploration-Driven Learning

Funnels & Conversion

Events → Funnel → Questions
Funnel Drop-off Investigation
Segmentation Analysis

Data Validation

Never Trust Charts Without Raw Events
Manual Funnel Walkthrough

Practical Skills

Using ChatGPT for Unknown Parameters
The “Better Hunter” Concept
Raw Data is Not “Developer Shit”

Attribution

The 3-Stage Funnel

Concepts:

Stage 1 – Landing: Anonymous visitor, UTM parameters captured (no contact info)
Stage 2 – Lead: Identified visitor, email/phone captured (UTM + PII)
Stage 3 – Sale: Customer, transaction completed (UTM + PII + transaction data)
Each stage is a subset of the previous (Stage 3 ⊂ Stage 2 ⊂ Stage 1)
Attribution is trivial: Sale email → Lead record → Original UTM parameters

Why this matters: Sales aren’t mysterious—every sale traces back to a specific landing with specific UTM parameters.

Source: Attribution & Tracking - Three-Stage Model

The Attribution Chain

Concepts:

Step 1: User lands with UTM parameters
Step 2: User navigates funnel
Step 3: User provides PII (email/phone)—becomes lead
Step 4: Lead converts to sale
PII is the bridge connecting anonymous → identified → customer

Source: Attribution & Tracking - The Attribution Chain

First-Touch vs Last-Touch Attribution

Concepts:

First-touch: Attribute to original traffic source (correct for acquisition)
Last-touch: Attribute to final source before conversion (misleading)
Retargeting is amplification, not acquisition

Source: Core Principles - First-Touch Attribution

Implementation (1 Day of Dev Work)

Concepts:

Capturing UTM: 10 lines of JavaScript
Persisting: Already done in modern tools
Total time: 1 day maximum
No excuses

Source: Attribution & Tracking - Implementation

Event Schema & Parameters

What is an Event Schema?

Concepts:

Event = name + user ID + timestamp + properties + parameters
Parameters are how you see attribution
Must understand your schema to read data correctly

Source: Event Schema - What is an Event Schema?

Standard UTM Parameters

Concepts:

utm_source: Traffic source (fb, google, email)
utm_medium: Medium (cpc, social, email)
utm_campaign: Campaign name
utm_term: Keyword/targeting
utm_content: Ad variation

Source: Event Schema - Standard UTM Parameters

Platform-Specific IDs

Concepts:

Facebook: campaign_id, ad_id, adset_id, fbp, fbclid
Google: gclid, g_campaignid
Custom: affid, aff_sub, tid

Source: Event Schema - Platform-Specific IDs

Reading Event Footprints

Concepts:

Every event leaves a “footprint” (parameter combination)
Surface labels lie. Parameters tell truth.
Example: Label says “Google Ads” but parameters show SMS retargeting

Source: Event Schema - Reading Event Footprints

Parameter Evolution (Schema Changes)

Concepts:

Schema changes tell story of system changes
Track evolution in changelog
Understanding evolution = better hunter

Source: Event Schema - Parameter Evolution

Cross-Validation

The Three-System Cross-Check

Concepts:

System 1: Ad Platforms (undercount due to consent)
System 2: Attribution/Analytics (first-party tracking)
System 3: Billing (ultimate truth)
Typical: FB 80, PostHog 100, Billing 95

Source: Cross-Validation - Three-System Cross-Check

Concepts:

Facebook sees 50-80% of actual conversions
iOS ATT, GDPR, ad blockers cause undercounting
Calculate gap: (Your System - FB) / Your System
Don’t panic and stop campaigns based on FB data alone

Source: Cross-Validation - Facebook Consent Gap

Unattributed Sales (Normal 5-20%)

Concepts:

5-20% unattributed is normal
20-30% is high but manageable
30% means tracking is broken
Don’t try to eliminate all unattributed

Source: Cross-Validation - Unattributed Sales

Daily Cross-Validation Workflow

Concepts:

Morning: Check ad platform vs attribution (campaign decisions)
Weekly: Check attribution vs billing (validate tracking)

Source: Cross-Validation - Daily Workflow

Micro vs Macro Analysis

Concepts:

Micro: Event-level footprints
Macro: System totals and gaps
Need both: Macro finds problems, micro fixes them

Source: Cross-Validation - Micro-Macro Balance

Thinking Flows

Core Flow: Aggregate → Individual → Pattern

Concepts:

Step 1: See aggregate number
Step 2: Understand it’s made of individuals
Step 3: Drill down to the ones
Step 4: Random sampling
Step 5: Segment to find patterns
Step 6: Form hypotheses and validate

Source: Thinking Flows - Core Flow

Multi-Dimensional Segmentation

Concepts:

Layer filters: landing page + affid + campaign
Single dimension misses winners
Example: 5.80% → 44.21% with multi-dimensional filtering

Source: Thinking Flows - Multi-Dimensional Segmentation

Isolating Landing Page vs Traffic Quality

Concepts:

Hold page constant, vary traffic → Find traffic quality issues
Hold traffic constant, vary page → Find page quality issues
Multi-step validation confirms diagnosis

Source: Thinking Flows - Advanced Isolation

Monitoring & Analytics

Self-Sufficient Monitoring

Concepts:

Don’t wait for alerts
System shows ALL landing pages automatically
Discover changes yourself through observation

Source: Core Principles - Self-Sufficient Monitoring

Discover, Don’t Define

Concepts:

Auto-discover patterns > manually define tracking
Catches unexpected traffic without manual setup

Source: Core Principles - Discover, Don’t Define

Exploration-Driven Learning

Concepts:

Learn by clicking around, not reading docs
Good tools enable natural discovery
Learning by doing > learning by reading

Source: Core Principles - Exploration-Driven Learning

Funnels & Conversion

Events → Funnel → Questions

Concepts:

After discovering events, build funnels immediately
Funnel visualization screams problems at you
90% drop-off = obvious red flag

Source: Core Principles - Events → Funnel → Questions

Funnel Drop-off Investigation

Concepts:

Numbers = real users you can investigate
Click through to see WHO the drop-offs are
Segment by campaign, device, source

Source: Practical Examples - Funnel Drop-off Investigation

Segmentation Analysis

Concepts:

Break down by dimensions (landing page, campaign, etc.)
Without segmentation: vague, not actionable
With segmentation: specific, actionable intelligence

Source: Practical Examples - Segmentation Analysis

Data Validation

Never Trust Charts Without Checking Raw Events

Concepts:

Charts = abstractions. Raw events = reality.
Always spot-check 5-10 events manually
Only then trust aggregated charts

Source: Core Principles - Never Trust Charts

The Manual Funnel Walkthrough

Concepts:

Go through your own funnel monthly
Watch events fire in real-time
Verify footprints at each step
10-15 min = complete understanding of data flow

Source: Event Schema - Manual Funnel Walkthrough

Practical Skills

Using ChatGPT/Google for Unknown Parameters

Concepts:

Unknown parameter? Look it up. Takes 1 minute.
Build parameter vocabulary over time
No shame in not knowing

Source: Core Principles - Use ChatGPT/Google

The “Better Hunter” Concept

Concepts:

Bad hunter: Only dashboards, doesn’t understand raw data
Good hunter: Checks raw events, understands parameters
Great hunter: Knows schema evolution, debugs attribution, owns data

How to become better: Look at raw data daily, research unknown parameters, walk funnel manually, track schema changes

Source: Event Schema - The “Better Hunter” Concept

Raw Data is Not “Developer Shit”

Concepts:

Understanding ugly data = core marketing analytics work
IDs, affids, UTMs are THE CORE
Don’t delegate to developers—own it yourself

Source: Core Principles - Raw Data

Summary: Most Important Concepts

Sales aren’t mysterious: Email → lead → original UTM. Done.
Parameters are everything: Learn your event footprints.
Never trust one system: Cross-validate always.
Numbers are people: Drill down to individuals.
First-touch is sacred: Attribute to acquisition.
5-20% unattributed is normal: Billing is truth.
Facebook undercounts 20-50%: Check your system first.
Traffic quality persists: Fix targeting before pages.
Multi-dimensional segmentation: Layer your filters.
Own the raw data: Your job, not developer’s.