Skip to main content
AdMesh implements two distinct pricing models designed specifically for AI-native advertising environments: CPX (Cost Per Exposure) and CPC (Cost Per Click). These models differ fundamentally in their billing behavior, making them suitable for different campaign objectives.
Key Innovation: Unlike traditional advertising models, AdMesh uses a cascading attribution system that ensures brands are never charged twice for the same user engagement. The system automatically selects the highest-value engagement and charges only for that, while releasing or waiving lower-value charges.

​
CPX Model: Awareness-Focused Pricing

​
Overview

The CPX Model is designed for brand awareness campaigns where the primary goal is visibility and reach, not immediate clicks or conversions.

​
Billing Behavior

  • CPX is charged immediately when the recommendation is exposed to the user
  • Amount: Based on contextual relevance score (CRS) and campaign settings
  • Timing: Charged after 30-minute attribution window if no click occurs
  • Click is tracked for analytics (stored in clicks collection)
  • CPC is set to 0 - no additional charge for clicks
  • CPX charge remains (already charged on exposure)
  • should_bill_cpc = False
  • billing_reason = "CPX_model_no_cpc_charge"
  • If conversion occurs, CPA is charged
  • CPX charge is waived/released (conversion is higher value)
  • Only CPA is charged

​
Use Cases

  • Brand Awareness Campaigns: Maximize visibility without click risk
  • Predictable Budgeting: Fixed cost per exposure, no variable click costs
  • Low-Risk Testing: Test new products or markets with controlled costs
  • Top-of-Funnel: Build brand recognition before driving conversions

​
Advantages

✅ Predictable Costs: Know exactly what you’ll pay per exposure
✅ No Click Risk: Clicks don’t increase costs
✅ Analytics Included: Click tracking still provides valuable insights
✅ Lower Barrier: Easier to justify spend for awareness goals

​
Example Scenario

User Query: “What are the best project management tools?” 
T+0:    Recommendation shown → CPX charged ($0.05)
T+5min: User clicks link → Tracked, but no CPC charge
T+2hrs: User signs up → CPA charged ($10.00), CPX waived

Result: Only $10.00 charged (CPA)

​
CPC Model: Performance-Focused Pricing

​
Overview

The CPC Model is designed for performance campaigns where ROI and conversion tracking are priorities. This model follows a more traditional CPC approach but with AI-native enhancements.

​
Billing Behavior

  • CPX budget is reserved (not charged yet)
  • Budget is held in pending state
  • If no click occurs within 30 minutes, CPX is charged
  • CPX is released (waived, returned to available budget)
  • CPC is charged instead
  • Amount: Based on agent trust score (typically 0.10−0.10-0.30)
  • should_bill_cpc = True (unless duplicate click in session)
  • billing.components.cpx_released = cpx_value
  • If conversion occurs, CPA is charged
  • CPC charge is waived/released (conversion is higher value)
  • Only CPA is charged

​
Use Cases

  • Performance Campaigns: Focus on clicks and conversions
  • ROI Optimization: Pay only when users engage
  • Conversion-Focused: Maximize qualified traffic
  • Budget Efficiency: Reserve budget but only charge on engagement

​
Advantages

✅ Pay for Engagement: Only charged when users click
✅ Budget Protection: CPX reserved but not wasted if click occurs
✅ Performance Tracking: Clear correlation between spend and engagement
✅ Industry Standard: Familiar CPC model with AI enhancements

​
Example Scenario

User Query: “Best CRM for small teams” 
T+0:    Recommendation shown → CPX reserved ($0.05, not charged)
T+5min: User clicks link → CPX released, CPC charged ($0.20)
T+2hrs: User signs up → CPA charged ($10.00), CPC waived

Result: Only $10.00 charged (CPA)

​
Cascading Attribution Model

​
Core Principle: Single Billing Guarantee

AdMesh guarantees that brands are never charged twice for the same user engagement. The system automatically selects the highest-value engagement and charges only for that.

​
Attribution Hierarchy

CPA (Highest Value) > CPC (Medium Value) > CPX (Base Value)

​
Attribution Windows

  1. Exposure → Click: 30 minutes
    • If click occurs within 30 min: Proceed to click billing
    • If no click: Charge CPX
  2. Click → Conversion: 24 hours
    • If conversion occurs within 24 hrs: Charge CPA (waive CPC)
    • If no conversion: Charge CPC (waive CPX if applicable)
  3. Conversion: Immediate
    • Verified via webhook/postback
    • Highest priority billing event

​
Billing Flow

​
Comparison with Industry Standards

​
Traditional Advertising Models

ModelWhen ChargedIndustry Standard
CPMPer 1,000 impressionsCharge on impression, clicks free
CPCPer clickCharge only on click, no impression charge
CPAPer conversionCharge only on conversion, no click/impression charge

​
AdMesh Innovation

ModelExposureClickConversionInnovation
CPX Model✅ Charged❌ Free (tracked)✅ Charged (waives CPX)Awareness-focused, predictable costs
CPC Model⏸️ Reserved✅ Charged (releases CPX)✅ Charged (waives CPC)Performance-focused, budget protection

​
Key Differences

  1. CPX Model ≠ Traditional CPM
    • CPM: Charges per 1,000 impressions (volume-based)
    • CPX: Charges per verified exposure with intent (quality-based)
    • CPX: Tracks clicks for analytics (CPM typically doesn’t)
  2. CPC Model ≠ Traditional CPC
    • Traditional CPC: No exposure charge/reservation
    • AdMesh CPC: Reserves CPX on exposure, releases on click
    • Both: Charge on click, but AdMesh prevents double-billing
  3. Cascading Attribution
    • Traditional: Separate billing for each event type
    • AdMesh: Automatic waiver/release of lower-value charges
    • Result: Single billing guarantee, never charged twice

​
Examples and Scenarios

​
Scenario 1: CPX Model - Exposure Only

Timeline:
  • T+0: Recommendation exposed → CPX charged ($0.05)
  • T+5min: User clicks → Tracked, no CPC charge
  • T+30min: No conversion → Final billing: CPX only
Result: $0.05 charged

​
Scenario 2: CPX Model - Exposure + Conversion

Timeline:
  • T+0: Recommendation exposed → CPX charged ($0.05)
  • T+5min: User clicks → Tracked, no CPC charge
  • T+2hrs: User converts → CPA charged ($10.00), CPX waived
Result: $10.00 charged (CPA)

​
Scenario 3: CPC Model - Click Only

Timeline:
  • T+0: Recommendation exposed → CPX reserved ($0.05, not charged)
  • T+5min: User clicks → CPX released, CPC charged ($0.20)
  • T+24hrs: No conversion → Final billing: CPC only
Result: $0.20 charged (CPC)

​
Scenario 4: CPC Model - Click + Conversion

Timeline:
  • T+0: Recommendation exposed → CPX reserved ($0.05, not charged)
  • T+5min: User clicks → CPX released, CPC charged ($0.20)
  • T+2hrs: User converts → CPA charged ($10.00), CPC waived
Result: $10.00 charged (CPA)

​
Scenario 5: CPC Model - No Click

Timeline:
  • T+0: Recommendation exposed → CPX reserved ($0.05, not charged)
  • T+30min: No click → CPX charged ($0.05)
  • No conversion → Final billing: CPX only
Result: $0.05 charged (CPX)

​
How AdMesh Selects the Model

Your brand-agent’s AI (unified reasoning layer) automatically selects the optimal pricing model based on:
  • Query context — User intent and conversation flow
  • Campaign goals — Awareness vs performance objectives
  • Budget availability — Wallet balance and pricing model affordability
  • Market conditions — Competitive landscape and relevance scores
The selected model is stored in the recommendation as preferred_pricing_model and determines billing behavior.

​
Fallback Behavior

When wallet balance is insufficient for the preferred pricing model, the brand-agent automatically falls back:
  1. Preferred Model: Try preferred_pricing_model first
  2. Fallback Order: CPX → CPC (prefers CPX when balance is low)
  3. No-Bid: If all models fail, send no-bid response
This ensures campaigns can continue even with limited budget, prioritizing lower-cost CPX when possible.

​
Summary

AdMesh’s CPX/CPC billing models represent an AI-native innovation that:
  1. Prevents Double-Billing: Cascading attribution ensures single charge
  2. Supports Multiple Goals: Awareness (CPX) vs Performance (CPC)
  3. Protects Budget: Reservation system prevents waste
  4. Provides Flexibility: LLM selects optimal model per context
  5. Maintains Standards: Familiar CPC model with enhancements
This approach is not traditional but is purpose-built for conversational AI environments where intent, context, and verified exposure matter more than raw impressions.

Pricing Models

Detailed guide to CPX and CPC models for advertisers.

Cost Per Exposure

Deep dive into CPX calculation and mechanics.

Budget & Performance

Learn how to optimize your ROI across pricing models.

Contextual Relevance Score

Understand how CRS affects CPX pricing.