Objectives

Purchase Prediction

  • The model samples out your creative broadly to hundreds of thousands of devices; once the model sees a number of attributed purchases occur it can begin to accurately predict purchase

Billing

  • All campaigns are charged on a dynamic CPM basis

Core Objectives

  • Cost Per Purchase (CPP)
  • Defined as your ad spend divided by the number of purchases
  • The campaign will optimize to drive maximum purchases on your site
  • Return On Ad Spend (ROAS)
  • Defined as your revenue generated divided by ad spend
  • The campaign will optimize to drive maximum return on ad spend on your site
Optimal campaign structure at launch - 50/50 structure, splitting your initial test budget evenly between CPP and ROAS campaigns in order to maximize scaling potential

Limited Access Objectives

  • Cost Per Event (CPE)
  • Defined as your ad spend divided by the number of target events
  • Currently only generate_lead is available
  • Under development and only available by request
  •  More information here 

Optimization Window

  • Depending on your objective, you can optimize CPP/ROAS for either Day 0 (D0) or Day 7 (D7), with one optimization window allowed per campaign.
  • D0 Optimization Window: Optimizes for conversions that occur on the same day as the ad interaction.
  • D7 Optimization Window: Optimizes for conversions that occur within 7 days of the ad interaction, capturing delayed conversion value.