====== Why vendors misjudge affiliate quality ======

===== The surface metric trap =====

Vendors often evaluate affiliates using simple metrics.

Sales volume.
EPC.
Gross revenue.
Traffic scale.

Affiliates generating the most sales are assumed to be the highest quality.

This assumption is incomplete.

CBSplit was built to expose what surface metrics cannot show.

===== Volume does not equal quality =====

An affiliate can generate high volume through:

  
  * Aggressive angles
  * Broad targeting
  * Heavy discount framing
  * Retry-dependent traffic

High volume can coexist with:

  
  * High refunds
  * Early churn
  * Increased support burden
  * Processor risk

CBSplit separates revenue volume from revenue durability.

===== Early revenue hides delayed damage =====

Affiliate quality cannot be measured at checkout.

It must account for:

  
  * Refund rates over time
  * Chargeback frequency
  * Subscription survival
  * Cohort-based LTV

Vendors who judge quality too early misinterpret short-term performance.

CBSplit measures quality after outcomes settle.

===== Traffic intent matters more than traffic scale =====

Two affiliates can drive similar sales numbers.

One may bring:

  
  * High-intent buyers
  * Clear expectation alignment
  * Stable subscription behavior

The other may bring:

  
  * Curiosity-driven clicks
  * Misaligned messaging
  * High post-purchase regret

Surface stats treat them as equal.

CBSplit does not.

===== Affiliate messaging affects downstream outcomes =====

Affiliate traffic is shaped by messaging.

Messaging influences:

  
  * Buyer expectations
  * Refund likelihood
  * Support interactions
  * Long-term satisfaction

Vendors often see the sale.
They rarely see the messaging that shaped it.

CBSplit connects downstream behavior back to traffic source quality.

===== Aggregated reporting hides affiliate variance =====

Marketplace reports aggregate performance.

Aggregation smooths over:

  
  * Cohort differences
  * Refund clustering
  * Subscription retention gaps
  * Payment behavior patterns

An affiliate can look profitable in aggregate while degrading vendor stability.

CBSplit preserves granularity.

===== Processor health reflects affiliate quality =====

Affiliate quality also impacts:

  
  * Refund thresholds
  * Chargeback ratios
  * Risk scoring
  * Account stability

These risks accumulate silently.

CBSplit monitors refund-adjusted and rebill-adjusted performance by affiliate source.

===== Misjudging quality leads to wrong incentives =====

When vendors reward only volume, they encourage:

  
  * Short-term tactics
  * Aggressive positioning
  * Fragile revenue
  * Increased refund exposure

True affiliate quality reflects:

  
  * Revenue durability
  * Subscription retention
  * Low refund behavior
  * Aligned buyer expectations

CBSplit supports outcome-based evaluation instead of event-based evaluation.