graph TD;
%% ---- Adobe Brand Colors (Enhanced Contrast) ----
classDef userAction fill:#f8f8f8,stroke:#2D2E2D,color:#2D2E2D,stroke-width:2px;
classDef serverCall fill:#666,stroke:#2D2E2D,color:#fff,stroke-width:2px;
classDef hitPacket fill:#ED2224,stroke:#fff,color:#fff,stroke-width:3px;
classDef warehouse fill:#2D2E2D,stroke:#ED2224,color:#f9f9f9,stroke-width:2px;
classDef painful fill:#8B0000,stroke:#ED2224,color:#fff,stroke-width:3px;
classDef titleBg fill:#1a1a1a,stroke:#ED2224,color:#f9f9f9;
%% ---- 1. User Actions (Clean & Distinct) ----
subgraph UserActions [" "]
style UserActions fill:none,stroke:none
subgraph UserTitle[ 🎯 TWO DISTINCT USER ACTIONS ]
style UserTitle fill:#1a1a1a,stroke:#ED2224,color:#f9f9f9
direction LR
Action1("🛒<br>Purchase<br>Clicked<br>12:34:18")
Action2("✉️<br>Newsletter<br>Signup<br>12:34:24")
end
end
%% ---- 2. The Fatal Flaw: Bundling ----
subgraph BundleSection [" "]
style BundleSection fill:none,stroke:none
Action1 & Action2 --> invisible( );
style invisible fill:none,stroke:none;
invisible --> |❌ FATAL FLAW: All actions bundled| ServerCall("📡<br>Single Hit<br>to Adobe");
subgraph BundleTitle[ ⚠️ THE BUNDLING PROBLEM ]
style BundleTitle fill:#1a1a1a,stroke:#ED2224,color:#f9f9f9
ServerCall --> HitData["
🔒 **LOCKED IN ONE TIMESTAMP**<br><br>
hit_time_gmt: 1734567858<br>
page_name: 'confirmation'<br>
eVar5: 'ABC-123' (Product)<br>
eVar12: 'SummerSale' (Campaign)<br>
event_list: 'purchase,event27'<br><br>
❌ **NO WAY TO SEPARATE ACTIONS**
"];
end
end
%% ---- 3. Rigid Storage ----
subgraph StorageSection [" "]
style StorageSection fill:none,stroke:none
HitData --> Storage{📦 Data Warehouse};
subgraph StorageTitle[ 🏗️ RIGID PRE-BUILT SCHEMA ]
style StorageTitle fill:#1a1a1a,stroke:#ED2224,color:#f9f9f9
Storage --> FactTable("Fact Table: Hits<br>📊 ONE ROW PER HIT")
FactTable --> Dim1("eVar5 Dimension<br>🏷️ Product Data")
FactTable --> Dim2("eVar12 Dimension<br>🏷️ Campaign Data")
FactTable --> Dim3("❓ event27 Dimension<br>🤷 Unknown Purpose")
end
end
%% ---- 4. The Painful Reality ----
subgraph PainSection [" "]
style PainSection fill:none,stroke:none
Dim1 & Dim2 & Dim3 --> PainPoints;
subgraph PainTitle[ 💀 PAINFUL LIMITATIONS IN 2025 ]
style PainTitle fill:#1a1a1a,stroke:#ED2224,color:#f9f9f9
PainPoints["🚫 IMPOSSIBLE IN ENTERPRISE 2025:<br><br>❓ Which promo drove this purchase?<br> eVar expired, attribution lost<br><br>❓ Cross-platform customer journey?<br> Separate silos, no unification<br><br>❓ Item-level attribution tracking?<br> Merchandising eVars complex setup<br><br>❓ Retroactive cohort analysis?<br> Impossible - data locked at collection<br><br>❓ Dynamic session definitions?<br> Client-side logic cannot change history<br><br>💸 BUSINESS IMPACT:<br> Marketing ROI unmeasurable<br> Customer experience fragmented<br> Data team bottlenecked on eVar mgmt"];
end
end
%% ---- Apply Enhanced Styling ----
class UserTitle,BundleTitle,StorageTitle,PainTitle titleBg;
class Action1,Action2 userAction;
class ServerCall serverCall;
class HitData hitPacket;
class Storage,FactTable,Dim1,Dim2,Dim3 warehouse;
class PainPoints painful;
Last active
July 28, 2025 16:03
-
-
Save ak--47/09854a3b7e03d3a8ca41a5be5bbba841 to your computer and use it in GitHub Desktop.
Mixpanel + Adobe Data Model Comparisons
graph TD;
%% ---- Mixpanel Brand Colors (Enhanced) ----
classDef userAction fill:#E0FAFA,stroke:#7856FF,color:#0a101a,stroke-width:2px;
classDef eventStream fill:#1A2433,stroke:#7856FF,color:#E0FAFA,stroke-width:3px;
classDef queryEngine fill:#07B096,stroke:#80E1D9,color:#0a101a,stroke-width:3px;
classDef enrichment fill:#1A2433,stroke:#80E1D9,color:#E0FAFA,stroke-width:2px;
classDef superpower fill:#7856FF,stroke:#F8BC3B,color:#E0FAFA,stroke-width:4px;
classDef titleDark fill:#0a101a,stroke:#7856FF,color:#E0FAFA;
%% ---- 1. Same User Actions (But Better Handling) ----
subgraph UserSection [" "]
style UserSection fill:none,stroke:none
subgraph UserTitle[🎯 SAME TWO USER ACTIONS]
style UserTitle fill:#0a101a,stroke:#7856FF,color:#E0FAFA
direction LR
Action1("🛒<br>Purchase<br>Clicked<br>12:34:18")
Action2("✉️<br>Newsletter<br>Signup<br>12:34:24")
end
end
%% ---- 2. Individual Event Streams ----
subgraph EventSection [" "]
style EventSection fill:none,stroke:none
subgraph EventTitle[⚡ ATOMIC EVENT CAPTURE]
style EventTitle fill:#0a101a,stroke:#7856FF,color:#E0FAFA
direction TB
Action1 --> Event1["
✨ **'Product Purchased'**<br>
🕐 time: 1734567858<br>
👤 distinct_id: 'user-xyz'<br><br>
📊 **properties**:<br>
• product_sku: 'ABC-123'<br>
• price: 99.99<br>
• campaign: 'SummerSale'<br>
• discount_code: 'SAVE20'
"]
Action2 --> Event2["
✨ **'Newsletter Signup'**<br>
🕐 time: 1734567864<br>
👤 distinct_id: 'user-xyz'<br><br>
📊 **properties**:<br>
• source: 'Confirmation Page'<br>
• campaign: 'SummerSale'<br>
• list_type: 'Weekly Updates'
"]
end
end
%% ---- 3. Live Query Intelligence ----
subgraph QuerySection [" "]
style QuerySection fill:none,stroke:none
Event1 & Event2 --> QueryEngine("🧠<br>LIVE QUERY<br>ENGINE");
subgraph QueryTitle[🚀 QUERY-TIME INTELLIGENCE]
style QueryTitle fill:#0a101a,stroke:#7856FF,color:#E0FAFA
direction LR
UserProfile["
👤 **User Profile (Cross-Device)**<br>
loyalty_tier: 'Gold'<br>
first_purchase: '2023-12-01'<br>
lifetime_orders: 47<br>
preferred_location: 'Downtown'<br>
**🔄 Retroactive Enrichment**
"]
GroupProfile["
🏢 **Behavioral Cohorts**<br>
'Heavy App Users'<br>
'Promo Sensitive Customers'<br>
'First Time Feature Users'<br>
**📊 Dynamic Query-Time**
"]
LookupTable["
📊 **Attribution Models**<br>
Multi-touch: Time Decay<br>
Hold Constant: product_id<br>
Session Props: utm_campaign<br>
**⚡ No Pre-Config Needed**
"]
QueryEngine -.-> |Borrows Properties| UserProfile;
QueryEngine -.-> |Enriches with| GroupProfile;
QueryEngine -.-> |Joins with| LookupTable;
end
end
%% ---- 4. Advanced Analytics Superpowers ----
subgraph SuperSection [" "]
style SuperSection fill:none,stroke:none
QueryEngine --> SuperPowers;
subgraph SuperTitle[🎪 ANALYTICS SUPERPOWERS UNLOCKED]
style SuperTitle fill:#0a101a,stroke:#7856FF,color:#E0FAFA
SuperPowers["
🔮 **RETROACTIVE ANALYSIS:**<br>
✅ Aug 15 push notification → Aug 16 purchase<br>
✅ Analyze 2023 events with 2025 user profiles<br><br>
🎯 **CROSS-PLATFORM ATTRIBUTION:**<br>
✅ Email → App → Kiosk → Purchase journey<br>
✅ Unified customer identity across touchpoints<br><br>
🛒 **HOLD PROPERTY CONSTANT:**<br>
✅ Item X: Promo View → Add Cart → Purchase<br>
✅ Per-item attribution without complex setup<br><br>
🧠 **BEHAVIORAL COHORTS:**<br>
✅ Users who never used Feature Y (dynamic)<br>
✅ First-time users of Feature Z (query-time)<br><br>
📊 **MULTI-TOUCH ATTRIBUTION:**<br>
✅ Time decay, participation, linear models<br>
✅ Campaign influence across 100+ touchpoints<br><br>
⚡ **QUERY-TIME FLEXIBILITY:**<br>
✅ Redefine sessions retroactively<br>
✅ Custom attribution windows on-demand<br>
✅ Zero implementation planning required
"];
end
end
%% ---- Apply Enhanced Styling ----
class UserTitle,EventTitle,QueryTitle,SuperTitle titleDark;
class Action1,Action2 userAction;
class Event1,Event2 eventStream;
class QueryEngine queryEngine;
class UserProfile,GroupProfile,LookupTable enrichment;
class SuperPowers superpower;
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment