====== Why backend revenue is invisible in ad dashboards ======

===== The front-end bias =====

Ad dashboards are built for acquisition.

They measure:

  
  * Clicks
  * Impressions
  * Conversions
  * Cost per acquisition
  * Immediate return on ad spend

They are not built for lifecycle revenue.

Backend performance exists outside their line of sight.

CBSplit was designed to restore that visibility.

===== Ad platforms track events, not lifecycles =====

Advertising systems optimize for:

  
  * Fast feedback loops
  * Immediate conversion signals
  * Short attribution windows
  * Predictable optimization cycles

Backend revenue unfolds through:

  
  * Upsells
  * Refund windows
  * Subscription rebills
  * Repeat purchases
  * Delayed confirmations

These events occur beyond typical ad attribution windows.

===== Attribution windows cut off revenue truth =====

Most ad platforms rely on:

  
  * 1-day click windows
  * 7-day click windows
  * Limited view-through attribution
  * Pixel-based tracking

Backend revenue often appears:

  
  * After refund periods close
  * During later billing cycles
  * Weeks after the first purchase

The ad dashboard cannot connect this revenue back to acquisition accurately.

===== Refund-adjusted revenue is rarely integrated =====

Ad dashboards typically report:

  
  * Gross conversion value
  * Front-end purchase totals
  * Pixel-reported revenue

They do not automatically subtract:

  
  * Refund amounts
  * Chargebacks
  * Subscription cancellations
  * Failed rebills

Gross revenue creates the illusion of performance.

Net revenue determines real profitability.

===== Upsell and subscription layers are external =====

Many backend flows occur:

  
  * On external checkout systems
  * Through third-party billing platforms
  * Via server-to-server callbacks
  * Across multiple domains

Ad pixels often cannot:

  
  * Persist identity across these systems
  * Attribute later revenue accurately
  * Stitch lifecycle events together

Backend revenue becomes disconnected from traffic source data.

===== Blended ROAS hides lifecycle fragility =====

Ad dashboards may show:

  
  * Strong day-one ROAS
  * Stable cost per conversion
  * Positive acquisition efficiency

They do not show:

  
  * Cohort-level churn
  * Refund clustering
  * Rebill survival rates
  * Processor risk signals

Scaling based solely on front-end ROAS can amplify backend instability.

===== Backend revenue requires infrastructure visibility =====

True performance measurement requires:

  
  * Server-side event stitching
  * Refund-adjusted calculations
  * Rebill-aware attribution
  * Cohort-based LTV tracking

Ad dashboards are not architected for this depth.

CBSplit operates at the lifecycle layer.

===== Acquisition without lifecycle visibility is incomplete =====

Advertising drives entry.

Backend systems determine durability.

Without integrating backend revenue into evaluation:

  
  * Profitable campaigns may be paused prematurely
  * Fragile campaigns may be scaled aggressively
  * Long-term ROI is miscalculated

CBSplit bridges the gap between acquisition and retention economics.